Upload folder using huggingface_hub

.gitattributes

@@ -215,3 +215,7 @@ datasets/bdcharm50/dept_061/GEO050K_HARM_061_S_FGEOL_2154.lyr filter=lfs diff=lf
 datasets/bdcharm50/dept_061/GEO050K_HARM_061_S_FGEOL_2154.shp filter=lfs diff=lfs merge=lfs -text
 docs/images/eure_reach_graph.png filter=lfs diff=lfs merge=lfs -text
 docs/images/explore_view.png filter=lfs diff=lfs merge=lfs -text
+datasets/c3s_seasonal/hindcast_month08.nc filter=lfs diff=lfs merge=lfs -text
+datasets/meteofrance_precipitation/daily_precipitation.csv filter=lfs diff=lfs merge=lfs -text
+datasets/single_trace_H605641201_2026-01-01.png filter=lfs diff=lfs merge=lfs -text
+datasets/spatiotemporal_gnn/selected_subgraph.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -18,6 +18,12 @@ braided splits/rejoins, ~4,500 nodes per basin including virtual infill
 points), not a single chain of gauges, with covariates pulled from nine
 independent data sources.
 
+The model itself is a spatiotemporal GNN trained with quantile regression
+(not a single point estimate) and four graph-tied physics losses, evaluated
+directly against real historical flood events rather than only against a
+naive baseline. See [§4](#4-the-model) for the architecture and real,
+current evaluation results.
+
 ```mermaid
 flowchart LR
     hubeau["Hub'Eau<br/>discharge · water level<br/>catchment area"]
@@ -28,6 +34,8 @@ flowchart LR
     bdtopo["IGN BD TOPO<br/>real reach topology + catchment polygons"]
     bdcav["Géorisques<br/>BDCavités (sinkholes)"]
     wc["ESA WorldCover<br/>landcover · NDVI"]
+    mf["Météo-France<br/>real historical precipitation"]
+    ecmwf["ECMWF (Open-Meteo)<br/>real forecast archive"]
 
     bdtopo --> brg["build_reach_graph.py<br/>real confluences, splits/rejoins,<br/>gauge snapping"]
     brg --> brgs["build_reach_graphs.py<br/>~4,500 nodes/basin"]
@@ -45,7 +53,13 @@ flowchart LR
     cc --> nf
 
     nf --> pyg["build_pyg_graph<br/>x_static / x_dynamic split"]
-    pyg --> phys["physics_losses.py<br/>confluence · split-rejoin ·<br/>routing · water balance"]
+    pyg --> phys["physics_losses.py<br/>confluence · split-rejoin ·<br/>routing · manning · water balance"]
+
+    mf --> train
+    ecmwf --> train
+    phys --> train["train_spatiotemporal_gnn.py<br/>quantile regression + physics losses,<br/>100-node subgraph"]
+    train --> model["model.pt<br/>trained checkpoint"]
+    model --> eval["evaluate_flood_detection.py<br/>vs. real historical flood events"]
 
     pyg --> app["src/app.py<br/>Streamlit explorer +<br/>network validation view"]
     pyg --> testsuite["test_build_graph.py<br/>validation"]
@@ -57,26 +71,43 @@ flowchart LR
 
 ```
 PoC_v1/
-├── scripts/                     # one-off download / extraction / build scripts
-│   ├── download_hubeau.py           # discharge + water level, Hub'Eau API v2
-│   ├── download_elevation.py        # point elevations, Open Topo Data
-│   ├── download_era5_sample.py      # ERA5 sanity-check pull (Jan 2020 only)
-│   ├── download_era5_full.py        # ERA5 1960–2026, split instant/accum vars
-│   ├── extract_era5.py              # unzips CDS API's zipped NetCDF output
-│   ├── download_catchment.py        # Hub'Eau referentiel/sites -> surface_bv
-│   ├── download_bdtopo_hydro.py     # IGN WFS -> tronçons, surfaces, catchments
-│   ├── analyze_bdtopo_hydro.py      # centerline export + karst check
-│   ├── run_bdtopo_checks.py         # karst + catchment cross-check, one shot
-│   ├── cross_check_catchments.py    # spatial join: station -> containing polygon
-│   ├── build_reach_graphs.py        # real reach-based topology, both basins
-│   ├── enrich_reach_graph.py        # runs node_features.py against the reach graph
-│   ├── compute_cumulative_catchment.py  # graph-wide catchment area from BD TOPO polygons
-│   ├── diagnose_confluences.py      # verify real vs. artifact confluences
-│   ├── build_dynamic_tensors.py     # genuine [n_nodes, T] tensors, wired into physics_losses.py
-│   ├── download_bdcavites.py        # Géorisques BDCavités (sinkhole/cavity inventory)
-│   ├── download_bdcharm.py          # BRGM BD Charm-50 harmonized geology, per department
-│   ├── fetch_landcover.py           # ESA WorldCover landcover class, real gauges
-│   └── fetch_worldcover_ndvi.py     # ESA WorldCover NDVI percentile composite
+├── scripts/                     # organized by purpose, not a flat list — see the
+│   │                             #   category note below the tree
+│   ├── data_acquisition/            # everything that talks to an external API/WFS/S3
+│   │   ├── download_hubeau.py           # discharge + water level, Hub'Eau API v2
+│   │   ├── download_elevation.py        # point elevations, Open Topo Data
+│   │   ├── download_era5_sample.py      # ERA5 sanity-check pull (Jan 2020 only)
+│   │   ├── download_era5_full.py        # ERA5 1960–2026, split instant/accum vars
+│   │   ├── extract_era5.py              # unzips CDS API's zipped NetCDF output
+│   │   ├── download_catchment.py        # Hub'Eau referentiel/sites -> surface_bv
+│   │   ├── download_bdtopo_hydro.py     # IGN WFS -> tronçons, surfaces, catchments
+│   │   ├── download_bdcavites.py        # Géorisques BDCavités (sinkhole/cavity inventory)
+│   │   ├── download_bdcharm.py          # BRGM BD Charm-50 harmonized geology, per department
+│   │   ├── fetch_landcover.py           # ESA WorldCover landcover class, real gauges
+│   │   ├── fetch_worldcover_ndvi.py     # ESA WorldCover NDVI percentile composite
+│   │   ├── fetch_ecmwf_forecast.py      # live ECMWF forecast, Open-Meteo (current conditions only)
+│   │   ├── fetch_previous_runs.py       # historical ECMWF forecast archive, Open-Meteo Previous Runs API
+│   │   └── fetch_meteofrance_precipitation.py  # real historical precipitation, Météo-France public API
+│   │
+│   ├── graph_building/              # everything that builds or enriches the real reach graph
+│   │   ├── build_reach_graphs.py        # real reach-based topology, both basins
+│   │   ├── enrich_reach_graph.py        # runs node_features.py against the reach graph
+│   │   ├── compute_cumulative_catchment.py  # graph-wide catchment area from BD TOPO polygons
+│   │   ├── compute_edge_width.py        # real channel width from BD TOPO polygons, per edge
+│   │   ├── diagnose_confluences.py      # verify real vs. artifact confluences
+│   │   └── build_dynamic_tensors.py     # genuine [n_nodes, T] tensors, wired into physics_losses.py
+│   │
+│   ├── analysis/                    # one-off diagnostics, not part of the main pipeline
+│   │   ├── analyze_bdtopo_hydro.py      # centerline export + karst check
+│   │   ├── run_bdtopo_checks.py         # karst + catchment cross-check, one shot
+│   │   └── cross_check_catchments.py    # spatial join: station -> containing polygon
+│   │
+│   ├── training/
+│   │   └── train_spatiotemporal_gnn.py  # main model: quantile regression + physics losses (§4)
+│   │
+│   └── evaluation/
+│       ├── identify_flood_events.py     # real historical high-flow events — ground truth
+│       └── evaluate_flood_detection.py  # precision/recall of the trained model vs. real historical floods
 │
 ├── src/
 │   ├── app.py                       # Streamlit river explorer + network validation view
@@ -104,14 +135,22 @@ PoC_v1/
 │       ├── build_graph.py           # PyG conversion: x_static/x_dynamic split, structural columns
 │       ├── build_reach_graph.py     # real reach topology: confluences, splits/rejoins, MultiDiGraph
 │       ├── node_features.py         # pulls every loader into one feature table (static, one row/node)
-│       ├── dynamic_features.py      # genuine [n_nodes, T] series: discharge, groundwater, climate
-│       └── physics_losses.py        # confluence/split-rejoin/routing/water-balance loss terms, NaN-masked
+│       ├── dynamic_features.py      # genuine [n_nodes, T] series: discharge, groundwater, climate,
+│       │                            #   precipitation, forecast (§4.4, §4.5)
+│       ├── physics_losses.py        # confluence/split-rejoin/routing/manning/water-balance loss terms,
+│       │                            #   NaN-masked (§2.5, §4.3)
+│       ├── spatiotemporal_gnn.py    # the model — TypeAwareInputEncoder, RiverMessagePassing (spatial),
+│       │                            #   per-node GRU (temporal), quantile decoder (§4.1)
+│       ├── pinball_loss.py          # quantile regression loss, asymmetric and NaN-masked (§4.2)
+│       └── subgraph_selection.py    # real-gauge + confluence/braid subgraph reduction for
+│                                     #   tractable training (§4.6)
 │
 ├── datasets/                    # not checked in; populated by the scripts above
 │   ├── station_list.csv             # raw station roster (X, Y, names, INSEE, etc.)
 │   ├── station_elevations.csv       # station_code, lat, lon, elevation_m
 │   ├── idpr.csv
 │   ├── catchment_area.csv
+│   ├── edge_widths.csv              # real channel width per edge, from compute_edge_width.py
 │   ├── ades/
 │   ├── hydrometric/
 │   ├── safran/
@@ -119,15 +158,27 @@ PoC_v1/
 │   ├── bdcavites/
 │   ├── bdcharm50/
 │   ├── centerlines/
-│   └── reach_graph/                 # {eure,risle}_{nodes,edges}.csv, _nodes_enriched.csv
+│   ├── meteofrance_precipitation/    # real historical precipitation, Météo-France (§3.12)
+│   │   └── daily_precipitation.csv
+│   ├── previous_runs_forecast.csv    # real ECMWF forecast archive, partial coverage (§3.13)
+│   ├── flood_events_discharge.csv    # real historical high-flow events — ground truth (§4.7)
+│   ├── flood_events_waterlevel.csv
+│   ├── flood_detection_evaluation.csv  # confusion matrix by lead time, real held-out test data
+│   ├── reach_graph/                 # {eure,risle}_{nodes,edges}.csv, _nodes_enriched.csv
+│   └── spatiotemporal_gnn/
+│       └── model.pt                 # trained model checkpoint
 ```
 
-`scripts/` talks to the outside world (APIs, WFS, S3);
-`src/` doesn't — nothing under `src/` makes a network call, and a script
-under `src/` that wants one is a bug. Most of `src/data/loaders/` predates
-the graph work — general-purpose readers/plotters for each dataset, with
-`node_features.py` stitching them together afterward rather than the other
-way around.
+`scripts/` talks to the outside world (APIs, WFS, S3) or trains/evaluates a
+model against real data; `src/` doesn't — nothing under `src/` makes a
+network call, and a script under `src/` that wants one is a bug. Most of
+`src/data/loaders/` predates the graph work — general-purpose
+readers/plotters for each dataset, with `node_features.py` stitching them
+together afterward rather than the other way around. `scripts/` itself is
+organized into subfolders by purpose (`data_acquisition/`, `graph_building/`,
+`analysis/`, `training/`, `evaluation/`) — every category listed in the tree
+above needs its own `__init__.py` for `python -m scripts.category.script_name`
+to resolve correctly.
 
 ---
 
@@ -205,7 +256,7 @@ flowchart TD
     xdynamic --> x
 
     edges["Edge table<br/>(build_reach_graph_tables)"] --> eattr{"numeric edge<br/>attribute?"}
-    eattr -->|"distance_km,<br/>elevation_drop_m,<br/>verified_continuous"| edgeattr["data.edge_attr<br/>[n_edges, 3]"]
+    eattr -->|"distance_km,<br/>elevation_drop_m,<br/>verified_continuous,<br/>width_m"| edgeattr["data.edge_attr<br/>[n_edges, 4]"]
     eattr -->|"toponym, cleabs<br/>(diagnostic metadata)"| meta["not used by build_pyg_graph —<br/>stays in edges_df only"]
 ```
 
@@ -225,14 +276,16 @@ flowchart TD
 | `climate_*` (temp/wind/solar/precip/evap/snow/runoff) | ERA5, vectorized station interpolation | every node (needs `safran_path`) |
 | `{col}__was_missing` | auto-generated | any feature column with real gaps |
 
-**Edge features** — three numeric attributes per edge, from
-`build_reach_graph.py`'s `build_reach_graph_tables`:
+**Edge features** — four numeric attributes per edge, from
+`build_reach_graph.py`'s `build_reach_graph_tables` (three original plus real
+channel width, added by `compute_edge_width.py`):
 
 | Feature | Meaning |
 |---|---|
 | `distance_km` | along-river distance between the two endpoint nodes |
 | `elevation_drop_m` | elevation difference, upstream minus downstream — negated on the reverse edge when `bidirectional=True` |
 | `verified_continuous` | `False` for any edge deliberately flagged via `known_losing_reaches` (the bétoire stretch — §3.7) |
+| `width_m` (+ `width_missing_flag`) | real channel width, derived from BD TOPO polygon area ÷ tronçon length; `NaN` (not zero) where no real polygon match exists, with an explicit missingness flag rather than a silent fill — real coverage is partial, not universal |
 
 `toponym` and `cleabs` also live on the real edges table (the tronçon's river
 name and unique BD TOPO ID) but are diagnostic metadata, not model input —
@@ -280,11 +333,12 @@ left-merge on `station_code` can never mislabel a confluence or virtual node).
 
 **A genuine `[n_nodes, T]` series exists separately**, in
 `src/graph/dynamic_features.py` — `build_discharge_timeseries`,
-`build_groundwater_timeseries`, `build_climate_timeseries` — built
-specifically because `physics_losses.py`'s `routing_consistency_loss` needs a
-real time dimension and had nothing to consume before this existed. Same
-loaders as everywhere else, no re-fetching; the only difference is that these
-functions pivot to wide `[date x station_code]` form instead of collapsing to
+`build_waterlevel_timeseries`, `build_groundwater_timeseries`,
+`build_climate_timeseries`, and (added for model training, §4)
+`build_precipitation_timeseries` and `build_forecast_tensor`. Same
+loaders as everywhere else, no re-fetching; the difference is that these
+functions pivot to wide `[date x station_code]` form (or, for the forecast
+archive, a real per-lead-time lookup — see §3.13) instead of collapsing to
 one aggregate the way `node_features.py`'s `add_*_features` do.
 
 Two real challenges, not incidental engineering:
@@ -308,6 +362,9 @@ Two real challenges, not incidental engineering:
 for a basin, saves them, and feeds discharge directly into
 `routing_consistency_loss` alongside `build_routing_index` — the real
 integration point, not just parallel unconnected pieces.
+`train_spatiotemporal_gnn.py` (§4) is the fuller integration point, wiring
+discharge, water level, groundwater, precipitation, and (partially) forecast
+data together into one training loop.
 
 **Climate is untested against real data** — no real ERA5/`safran_path` files
 were available to validate `build_climate_timeseries` against in this
@@ -338,40 +395,49 @@ improvement.
 
 ### 2.5 Physics-informed loss terms (`physics_losses.py`)
 
-Four constraints, each tied to real graph structure, not generic:
+Five constraints, each tied to real graph structure, not generic — four
+original, plus Manning's equation added alongside the model (§4.3):
 
 | Term | Constraint | Applies to |
 |---|---|---|
 | `confluence_mass_balance_loss` | `Q_confluence ≈ sum(Q_upstream_branches)` — new mass genuinely enters | `is_confluence` nodes |
 | `split_rejoin_conservation_loss` | `Q_split ≈ Q_rejoin` — same water, no new mass | paired `braid_id` nodes |
 | `routing_consistency_loss` | `Q_downstream[t] ≈ Q_upstream[t - lag]`, lag from real `distance_km`/slope | every edge, real `[n_nodes, T]` via `dynamic_features.py` (§2.3) |
-| `water_balance_loss` | `P - ET - Q - ΔS ≈ 0` in volume terms | nodes with `cumulative_catchment_area_km2` |
+| `manning_consistency_loss` | `Q ≈ (1/n)·A·R^(2/3)·S^(1/2)`, learned roughness `n`, with a bank-height relaxation gate so overbank flow isn't penalized against in-channel physics | every edge with real width data |
+| `water_balance_loss` | `P - ET - Q - ΔS ≈ 0` in volume terms, normalized to a relative (not absolute) imbalance | nodes with `cumulative_catchment_area_km2` and real precipitation |
 
 Confluence and split/rejoin are deliberately different constraints, not one
 generic "conserve mass everywhere" rule — a model that only learned "sum the
 inflows" would get a split/rejoin wrong, since a rejoin's two branches
 together should equal the *split's* value, not add something new on top.
-All four apply graph-wide, not just at the 27 labeled gauges — that's the
-actual mechanism by which sparse supervision generalizes to the ~4,500
+All four original terms apply graph-wide, not just at the 27 labeled gauges
+— that's the actual mechanism by which sparse supervision generalizes to the
 ungauged nodes, not an incidental detail. `ΔS` (storage change) defaults to
 zero, a named steady-state approximation — this project has no direct
 basin-wide storage measurement, only sparse well *levels*, which aren't the
-same thing.
+same thing; likewise `ET` defaults to `0.0`, since real evapotranspiration
+data exists (from the ERA5 climate work) but lives on a different grid than
+this graph's nodes, and mapping between them is unstarted work.
 
-**All four are NaN-masked, not just tolerant of complete data.** Real ground-
+**All five are NaN-masked, not just tolerant of complete data.** Real ground-
 truth Q is ~93.5% `NaN` by construction (only real gauges with real
 observations ever have a value — confirmed against the real discharge
 tensor) — that's the normal shape of the data, not a rare edge case. The
 shared `_mse` helper every loss function uses previously computed a plain
 mean, so a single `NaN` anywhere in a residual silently poisoned the *entire*
-loss to `NaN` — confirmed as a real, not hypothetical, failure: calling
-`routing_consistency_loss` directly on the real discharge tensor returned
-`NaN` before this was fixed. `_mse` now masks `NaN` out before averaging
-(returning `NaN` only if truly nothing usable exists at all, which is a
-real "no data" signal worth keeping, not silently averaging to a misleading
-`0`) — verified with the exact real scenario that first exposed the bug:
-`routing_consistency_loss` on the real discharge tensor now returns a real
-number instead of `NaN`.
+loss to `NaN` — confirmed as a real, not hypothetical, failure. `_mse` now
+masks `NaN` out before averaging (returning `NaN` only if truly nothing
+usable exists at all, which is a real "no data" signal worth keeping, not
+silently averaging to a misleading `0`).
+
+Masking the *value* isn't sufficient on its own, either — a second, separate
+real bug found during model training (§4.3): a masked-out `NaN` upstream
+input can still poison a *gradient*, since `0 × NaN = NaN` in IEEE754
+arithmetic even when the outer loss correctly excludes that position. The
+real fix, applied in `water_balance_loss`, is to replace `NaN` inputs with a
+safe finite placeholder *before* any arithmetic touches them, and apply the
+real `valid` mask only at the very end, for loss selection — not to rely on
+masking alone to keep `NaN` out of the graph.
 
 This also means these functions are directly usable as a diagnostic against
 real historical data alone, independent of any trained model — e.g.
@@ -443,7 +509,12 @@ about:
 - Everything else with no data is unexplained from the name alone and worth a
   direct check on Hub'Eau's site before assuming it's just a gap.
 
- 
+14 of the 27 stations have neither real discharge nor real water-level data
+at all. These aren't excluded from the graph — they still get real static
+attributes (landcover, geology, catchment area) and, since §4.4, a real
+precipitation-based discharge *estimate* as an input feature (never a
+target), plus whatever signal reaches them through message passing from
+nearby gauged nodes.
 
 ### 3.2 Hydrometric data (`hydrometric/`, via `scripts/download_hubeau.py`)
 
@@ -521,6 +592,10 @@ reach graph's ~4,500 nodes (~193,000 calls). Vectorized indexing with
 DataArray indexers sharing a `station` dimension does every station in one
 call per file instead.
 
+Note: this is real historical *reanalysis* — a model-based reconstruction of
+what already happened, not a forecast. §3.13 covers the separate, genuine
+forecast archive used for real advance-warning signal in the model (§4.5).
+
 ### 3.5 IDPR (`idpr.csv`)
 
 BRGM's *Indice de Développement et de Persistance des Réseaux* — an
@@ -573,6 +648,15 @@ full `download_bdtopo_hydro.py → build_reach_graphs.py → enrich_reach_graph.
 box to actually resolve this, which had not yet happened as of the last
 verified run in this project.
 
+**In practice, the model (§4) currently uses a per-node fallback between the
+two sources**, since `cumulative_catchment_area_km2` isn't present at all on
+the real enriched node tables used for training as of the last verified run —
+`catchment_area_km2` (Hub'Eau, real but gauge-only) is used wherever the
+cumulative BD TOPO value is missing. This recovers real catchment area for
+real gauges but not for confluences or braid points without a Hub'Eau
+delineation; those still have no real catchment area at all until the
+BD TOPO re-pull above is actually run.
+
 ### 3.7 BD TOPO hydrography (`bdtopo_hydro/`, via `scripts/download_bdtopo_hydro.py`)
 
 IGN's BD TOPO / BD TOPAGE hydrographic network, pulled from the Geoplateforme
@@ -587,8 +671,10 @@ that bbox:
   altitude data doubles as a fine-grained elevation profile, denser than
   anything derivable from the 27 gauge points alone.
 - `surface_hydrographique.geojson` — hydrographic surfaces, including a
-  `Nature` attribute that's supposed to flag karst-influenced reaches. IGN
-  documents this attribute as **provisional and incomplete**.
+  `Nature` attribute that's supposed to flag karst-influenced reaches (IGN
+  documents this attribute as **provisional and incomplete**), and the real
+  source for `compute_edge_width.py`'s channel-width derivation (§2.2):
+  polygon area ÷ tronçon length, joined by `cleabs`.
 - `bassin_versant_topographique.geojson` — catchment polygons, incremental
   (see §3.6).
 
@@ -699,9 +785,251 @@ km average snap distance rather than 0.05 km. It's the fallback when a real
 vector source isn't available, not the method used for the current
 `centerlines/` files.
 
+### 3.12 Real historical precipitation (`meteofrance_precipitation/`)
+
+Real, directly-measured rain-gauge observations from Météo-France's public
+"Données climatologiques de base — horaires" dataset — deliberately not ERA5
+reanalysis (§3.4) for this specific variable: precipitation varies sharply
+over short distances in ways a ~9 km reanalysis grid cell can miss, so a real
+nearby gauge is more locally authoritative where one exists. Official source,
+Open Licence 2.0, no authentication — `fetch_meteofrance_precipitation.py`
+fetches only the departments this project needs (27, 28, 61) and only the
+decades overlapping the real study period, aggregating real hourly `RR1`
+readings to daily sums per station.
+
+Each graph node reads from its nearest real Météo-France gauge by haversine
+distance (`map_stations_to_precipitation_gauges`) — several nodes can share
+one real gauge if they're close together, which is expected, not a bug, the
+same way several climate stations already share one ERA5 grid cell elsewhere
+in this project.
+
+**Also used to estimate discharge for stations with no real gauge data at
+all** (§3.1): `fit_specific_discharge_coefficient` fits a simple, standard
+hydrological scaling relationship (`Q ≈ k · precipitation · catchment_area`,
+a real regionalization technique) using only real (discharge, precipitation,
+catchment area) triples from gauged stations during the training period, then
+applies it to every node with real precipitation and catchment area,
+including the 14 with no real discharge history. This is wired in as a model
+**input feature**, never a target — it's a real, physically-grounded
+estimate, not a measurement, and reaches the model exactly the way every
+other imperfect-but-real input does.
+
+### 3.13 Real ECMWF forecast archive (`previous_runs_forecast.csv`)
+
+Real historical forecasts — not the live current forecast, and a genuinely
+different thing from ERA5 reanalysis (§3.4): reanalysis is a best estimate of
+what already happened, with zero forecast skill baked in; this archive is
+what was actually predicted *in advance*, which is what a model needs to
+learn genuine multi-day-ahead skill from. Fetched via
+`fetch_previous_runs.py` (Open-Meteo's Previous Runs API).
+
+Real, confirmed schema: one row per real observation date, with several
+`precipitation_previous_dayN` columns — each holding what was forecasted `N`
+days before that date, *for* that date. For an anchor date `t` and a horizon
+`h`, the real forecast value lives in the row for `date = t + h`, column
+`precipitation_previous_day{h}` — not the row for `t` itself, since that
+forecast was issued on `(t+h) - h = t`, the anchor. `station_code` in this
+file already matches this project's real graph node codes directly — no
+nearest-neighbor mapping needed here, unlike §3.12's precipitation gauges.
+
+**Real, significant coverage limits, not implementation gaps**: only lead
+times `[1, 2, 3, 5, 7]` days have any real data at all — none of this
+project's other horizons (4, 10, 30, 60, 90, 180) have real forecast data,
+ever. Coverage also only starts **2024-03-01** — the large majority of this
+project's 2013+ training window has no real forecast data regardless of lead
+time. Both gaps are handled with an explicit real/missing flag per
+`(example, node, horizon)`, the same value-plus-missingness pattern used
+throughout this project, not a silent fill.
+
+**A real, confirmed gotcha worth flagging directly**: with this project's
+default `--train-end`/`--val-end` split, the *entire* real forecast archive
+falls after the validation cutoff — meaning the model can see 0% real
+forecast coverage during training and validation, then 100% at evaluation
+time. That mismatch produced a genuine, measured collapse in evaluation
+performance at exactly the horizons where real forecast coverage is densest,
+since the model's forecast-conditioning pathway was never trained on any
+real (non-missing) input. The fix is a `--train-end`/`--val-end` shift so all
+three splits genuinely overlap the real forecast-covered period — see §4.5
+and §7.
+
+---
+
+## 4. The model
+
+### 4.1 Architecture (`spatiotemporal_gnn.py`)
+
+`SpatiotemporalGNN` predicts discharge and water level at any node — gauged
+or ungauged — for a requested lead time, trained end-to-end against the real
+graph and physics substrate above.
+
+- **Input encoding** (`TypeAwareInputEncoder`): gauged nodes get a dedicated
+  encoder path over real static features plus dynamic history; non-gauged
+  nodes get a learned placeholder, rather than forcing every node through
+  the same encoder as if the ~14 uninstrumented stations (§3.1) had real
+  time series to offer.
+- **Spatial** (`RiverMessagePassing`, two stacked layers): directed message
+  passing over the real reach graph, respecting real flow direction, with
+  **sum aggregation, not mean** — deliberate, not a default left
+  unconsidered: a confluence's downstream discharge is the *sum* of its
+  upstream branches (§2.5), so the aggregation itself reflects that physical
+  reality rather than averaging it away.
+- **Temporal** (a per-node GRU): runs over the real 30-day lookback window,
+  fed by the spatial layers' output *at each timestep* — spatial happens
+  first, then the resulting sequence of embeddings goes through the GRU, so
+  the recurrent step learns genuine spatiotemporal patterns rather than each
+  node's isolated history.
+- **Decoder**: combines the GRU's final hidden state with a basin embedding,
+  a horizon embedding (letting one trained model answer at ten different
+  lead times without retraining separately for each), and — when available —
+  a real forecast-precipitation embedding (§4.5), then outputs quantile
+  predictions for discharge and water level.
+
+### 4.2 Quantile regression, not a point estimate (`pinball_loss.py`)
+
+The model predicts several quantiles (`[0.5, 0.9, 0.95, 0.99]` by default)
+per target, not a single value — a real, evidence-driven architecture
+decision, not a default choice. Real discharge is heavily right-skewed
+(confirmed against real data: the 99th percentile runs 4–12x the median), and
+a model trained on plain MSE structurally minimizes *average* error, which
+means predicting close to the mean is a genuinely good strategy for MSE and a
+genuinely bad one for flood risk — confirmed directly, not just argued: an
+earlier MSE-trained version of this model caught **0 of 203** real historical
+flood events at every tested lead time (§4.7), despite showing real average-
+case skill.
+
+Pinball loss is asymmetric by construction: for a high quantile (e.g. 0.95),
+under-predicting is penalized far more than over-predicting by the same
+amount, which is the actual mechanism that pushes the model to cover the
+real tail rather than hide near the mean.
+
+**Quantile crossing is architecturally impossible, not just discouraged.** A
+well-known failure mode of naive quantile regression is the predicted 90th
+percentile coming out *lower* than the predicted median — physically
+nonsensical. The lowest quantile is predicted directly; every subsequent one
+is the previous quantile plus a `softplus`-transformed (always non-negative)
+offset, so `quantile[k] ≥ quantile[k-1]` holds for every real input,
+including deliberately adversarial ones.
+
+The naive baseline this model is compared against is also quantile-aware —
+each quantile's naive prediction is that quantile's own real empirical value
+from the training-period distribution (the direct quantile-regression
+equivalent of climatology as a baseline in the earlier climate work), not
+zero or the mean for every quantile alike.
+
+### 4.3 Physics losses, wired into training
+
+All five terms from §2.5 are combined per training example, horizon-weighted
+rather than with one fixed weight applied uniformly — routing and water
+balance specifically only apply to the short contiguous-horizon block (the
+first 5 daily steps), since both need a real, physically comparable period,
+not a mismatched one.
+
+Two real numerical-stability bugs were found and fixed during training,
+worth naming since both generalize beyond this specific project:
+
+- **Manning's equation** (`1/n`, `hydraulic_radius^(2/3)`) has genuinely
+  steep local derivatives near its natural boundaries — an untrained model's
+  raw output is close to unconstrained noise, and a single bad step could
+  push `n` toward zero, producing a real, not hypothetical, `inf` loss
+  (confirmed: two real training runs failed exactly this way before the
+  fix). `n`, bank height, hydraulic radius, and depth are all explicitly
+  clamped to physically real ranges before use.
+- **A gradient-masking gotcha in `water_balance_loss`** (§2.5): a value
+  correctly excluded from the *loss* can still corrupt its *gradient* if
+  `NaN` entered the computation before masking, since `0 × NaN = NaN`. Fixed
+  by replacing `NaN` with a safe finite placeholder before any arithmetic,
+  not by relying on the final mask alone.
+
+**A real, current limitation worth being direct about**: Manning's learned
+roughness `n` has landed on implausible values across real training runs
+(0.38–0.92, against a real natural-stream range of roughly 0.025–0.1), and
+varies meaningfully run to run rather than converging to a stable estimate.
+The likely mechanism is clamp saturation — the value used *inside* the loss
+computation is clamped to `[0.01, 0.3]` for numerical safety, and once the
+underlying parameter drifts past that ceiling, the clamp's own gradient is
+zero there, removing further corrective signal. This means Manning's
+equation has likely been running at its artificial ceiling for meaningful
+stretches of training rather than converging to a physically informed value
+— worth investigating (a wider clamp range, or a gradient that doesn't
+vanish at the boundary) before trusting this specific learned parameter.
+
+### 4.4 Real precipitation and a specific-discharge estimate
+
+Covered fully in §3.12. In short: real Météo-France precipitation feeds the
+temporal encoder directly as a dynamic input channel, and — for the 14
+stations with no real discharge history — a real, physically-grounded
+specific-discharge estimate (fit only on real gauged data, applied uniformly
+to every node) gives the model a genuine signal to work with instead of
+nothing, without ever fabricating a value the supervised loss is held
+accountable to match.
+
+### 4.5 Real forecast integration
+
+Covered fully in §3.13. The model's decoder accepts real forecasted
+precipitation per `(node, horizon)`, defaulting to fully-missing when no real
+forecast dataset is wired in — backward compatible with every training run
+before this was added. Given the real, confirmed train/test coverage
+mismatch documented in §3.13, using this correctly requires shifting
+`--train-end`/`--val-end` so training and validation genuinely overlap the
+real 2024-03-01+ forecast-covered period (see §7 for the actual command) —
+using the project's older default split dates with forecast data wired in
+will reproduce the coverage-mismatch failure, not avoid it.
+
+### 4.6 Subgraph reduction (`subgraph_selection.py`)
+
+Training runs against a real, reduced 100-node subgraph (50 nodes per basin:
+real gauges plus real confluences/braid points only — **virtual infill
+nodes are deliberately excluded**, not because they're uninteresting, but
+because training on the full ~4,500-node graph per basin was not tractable
+within this project's real compute constraints. Chains through dropped nodes
+are collapsed into single edges (summing distance/elevation, distance-
+weighting average width) rather than simply deleted, so real connectivity
+between kept nodes is preserved exactly. A full-graph retrain, once the
+100-node architecture is validated, remains real, unstarted future work.
+
+### 4.7 Real evaluation
+
+**`identify_flood_events.py`** establishes real ground truth independent of
+any model: real historical high-flow events, defined per station by that
+station's own real historical percentile (95th by default — a
+station-relative, not a shared absolute, threshold, since these rivers have
+very different real scales), with consecutive above-threshold days grouped
+into discrete events rather than counted individually.
+
+**`evaluate_flood_detection.py`** checks the trained model against this real
+ground truth on the held-out real test split — flagging a real event when
+the model's predicted upper quantile (0.95 by default, matching the same
+percentile `identify_flood_events.py` uses) exceeds that station's real
+threshold, not the median, since flood risk is a tail question. Precision
+and recall are reported per lead time, since catching a flood a day out and
+catching it a week out are different questions that would be hidden by one
+combined number.
+
+Real, current results (subject to change as the model continues to be
+refined — treat these as a snapshot, not a final claim):
+
+- The naive-baseline comparison (§4.2) shows real, meaningful skill: the
+  quantile model beats the real empirical-quantile baseline by roughly
+  30–50% on both discharge and water level across real training runs.
+- Real flood-detection recall at short lead times (1–5 days) has reached
+  roughly 65–70% in real evaluation runs — a genuine, validated improvement
+  over the earlier MSE-trained model's `0/203`.
+- **Precision remains low in absolute terms** (under 10% at every tested
+  horizon in the most recent full evaluation) — real, meaningful skill above
+  the base rate at short horizons, but still many false alarms; not yet
+  something to treat as deployment-ready without further tuning (e.g. a
+  higher `--flag-quantile`, trading recall for precision).
+- **Long-horizon "skill" needs a base-rate check, not just a recall number**:
+  at the longest tested horizon (180 days), precision has landed *below* the
+  real base rate in at least one evaluation run — a sign the model is likely
+  hedging with a wide quantile spread rather than genuinely predicting that
+  far out, not a sign of real long-range foresight. `evaluate_flood_
+  detection.py`'s output should always be read against the real base rate
+  (`(tp+fn)/total`), not the recall number alone.
+
 ---
 
-## 4. Applications
+## 5. Applications
 
 `src/app.py` (Streamlit) has two views, selected by a radio at the top:
 
@@ -740,7 +1068,7 @@ development, not just a theoretical risk.
 
 ---
 
-## 5. Testing (`src/test_build_graph.py`)
+## 6. Testing (`src/test_build_graph.py`)
 
 Not a unit test suite in the pytest sense — a script with two independent
 sections, both run from `main()`.
@@ -764,8 +1092,8 @@ they can't silently reintroduce themselves:
   `feature_names`, but remain accessible as their own `Data` attributes
 - target values never attach to a non-gauge node, and target coverage never
   exceeds the real gauge count
-- `edge_attr` stays exactly 3 columns despite extra edge metadata
-  (`toponym`, `cleabs`) sitting on the real edges table
+- `edge_attr` stays exactly the real numeric edge columns despite extra edge
+  metadata (`toponym`, `cleabs`) sitting on the real edges table
 - `physics_losses.py`'s `build_confluence_index`/`build_braid_index` produce
   counts matching `is_confluence`/`is_rejoin_point` sums, with every index
   within node bounds and every confluence having ≥ 2 upstream branches
@@ -776,25 +1104,49 @@ they can't silently reintroduce themselves:
 Exits 0 on a clean pass across both sections, 1 otherwise — usable as a
 pre-commit or CI gate if that's ever set up.
 
+`identify_flood_events.py` and `physics_losses.py`'s own functions (§2.5,
+§4.7) double as an informal, model-independent validation layer too — real
+historical data checked directly against real physics and real event
+definitions, without needing a trained model in the loop at all.
+
 ---
 
-## 6. Known limitations and open questions
+## 7. Known limitations and open questions
 
 - **`cumulative_catchment_area_km2` underestimates the largest catchments**
   by up to ~25%, traced to the BD TOPO pull's original bounding box having an
   insufficient southern margin. The bbox has been widened in
   `download_bdtopo_hydro.py`; the full re-pull-and-rebuild chain needs
-  re-running for this to actually resolve. Treat the largest catchments'
-  values as approximate until then.
+  re-running for this to actually resolve — and, in the meantime, this
+  column is entirely absent from the real enriched node tables the model
+  currently trains against (§3.6), which is why the model falls back to the
+  gauge-only Hub'Eau value.
 - **Climate's genuine time series (`build_climate_timeseries`) is untested
   against real data** — discharge and groundwater's equivalents are; verify
   climate's real output before relying on it.
 - **Landcover and NDVI only cover the 27 real gauges**, not the full reach
   graph — same scope `catchment_area_km2` had before its cumulative-BD-TOPO
   extension.
-- **No model exists yet.** This repo builds the graph and the physics-loss
-  substrate a model would train against; there is no architecture, forward
-  pass, or training loop here.
+- **The model trains and evaluates against a 100-node subgraph**, not the
+  full ~4,500-node-per-basin graph — a real, deliberate scope limit for
+  tractable compute, not a claim the full graph wouldn't matter. A full-graph
+  retrain is real, unstarted future work.
+- **Manning's learned roughness `n` is currently not trustworthy as a
+  physical finding** — real evidence (§4.3) points to clamp saturation, not
+  genuine convergence. Needs either a wider clamp range or a
+  non-vanishing-at-the-boundary gradient before the learned value means
+  anything physically.
+- **Real forecast precipitation coverage is genuinely partial** — five lead
+  times, roughly 2.5 years — and using it correctly requires a real,
+  documented train/val/test split shift (§3.13, §4.5); the project's
+  original default split dates predate the entire forecast archive.
+- **Flood-detection precision is low in absolute terms** even where recall is
+  real and meaningful (§4.7) — worth tuning the flag quantile and continuing
+  to validate before treating this as deployment-ready.
+- **Real evapotranspiration is not wired into `water_balance_loss`** — passed
+  as an explicit `0.0` placeholder; real ERA5 evaporation data exists but
+  lives on a different grid than this graph's nodes, and building that
+  mapping is real, unstarted work.
 - **The karst losing-reach flag rests on naming evidence and a BDCavités
   cross-check** (§3.7–3.8), not a fully confirmed BD TOPO classification.
 - **The groundwater-well BDLISA aquifer-unit field is unused.** First place
@@ -803,51 +1155,76 @@ pre-commit or CI gate if that's ever set up.
 
 ---
 
-## 7. Running things
+## 8. Running things
 
 Data acquisition (from repo root, in roughly dependency order):
 
 ```bash
-python -m scripts.download_hubeau
-python -m scripts.download_elevation
-python -m scripts.download_era5_full        # slow; download_era5_sample.py first if just testing
-python -m scripts.extract_era5
-python -m scripts.download_catchment
-python -m scripts.download_bdtopo_hydro --check   # verify typeNames before the real pull
-python -m scripts.download_bdtopo_hydro
-python -m scripts.download_bdcavites --check
-python -m scripts.download_bdcavites
-python -m scripts.download_bdcharm
+python -m scripts.data_acquisition.download_hubeau
+python -m scripts.data_acquisition.download_elevation
+python -m scripts.data_acquisition.download_era5_full        # slow; download_era5_sample.py first if just testing
+python -m scripts.data_acquisition.extract_era5
+python -m scripts.data_acquisition.download_catchment
+python -m scripts.data_acquisition.download_bdtopo_hydro --check   # verify typeNames before the real pull
+python -m scripts.data_acquisition.download_bdtopo_hydro
+python -m scripts.data_acquisition.download_bdcavites --check
+python -m scripts.data_acquisition.download_bdcavites
+python -m scripts.data_acquisition.download_bdcharm
+python -m scripts.data_acquisition.fetch_meteofrance_precipitation --check --department 27
+python -m scripts.data_acquisition.fetch_meteofrance_precipitation --data-root datasets
+python -m scripts.data_acquisition.fetch_previous_runs        # real ECMWF forecast archive
 ```
 
 Build and validate the reach graph:
 
 ```bash
-python -m scripts.build_reach_graphs --data-root datasets
-python -m scripts.enrich_reach_graph --data-root datasets   # --skip-climate if that step hangs
-python -m scripts.compute_cumulative_catchment --data-root datasets
-python -m scripts.diagnose_confluences --data-root datasets --basin eure
-python -m scripts.diagnose_confluences --data-root datasets --basin risle
+python -m scripts.graph_building.build_reach_graphs --data-root datasets
+python -m scripts.graph_building.enrich_reach_graph --data-root datasets   # --skip-climate if that step hangs
+python -m scripts.graph_building.compute_cumulative_catchment --data-root datasets
+python -m scripts.graph_building.compute_edge_width --data-root datasets
+python -m scripts.graph_building.diagnose_confluences --data-root datasets --basin eure
+python -m scripts.graph_building.diagnose_confluences --data-root datasets --basin risle
 ```
 
 Build genuine `[n_nodes, T]` dynamic tensors and verify the physics-loss wiring:
 
 ```bash
-python -m scripts.build_dynamic_tensors --data-root datasets --basin risle
-python -m scripts.build_dynamic_tensors --data-root datasets --basin eure
+python -m scripts.graph_building.build_dynamic_tensors --data-root datasets --basin risle
+python -m scripts.graph_building.build_dynamic_tensors --data-root datasets --basin eure
 ```
 
 Landcover / NDVI, real gauges only (needs `rasterio`, and `geopandas` for NDVI's
 tile lookup):
 
 ```bash
-python -m scripts.fetch_landcover --check
-python -m scripts.fetch_landcover
-python -m scripts.fetch_worldcover_ndvi --check
-python -m scripts.fetch_worldcover_ndvi
+python -m scripts.data_acquisition.fetch_landcover --check
+python -m scripts.data_acquisition.fetch_landcover
+python -m scripts.data_acquisition.fetch_worldcover_ndvi --check
+python -m scripts.data_acquisition.fetch_worldcover_ndvi
+```
+
+Real historical flood events (ground truth, needed before evaluation):
+
+```bash
+python -m scripts.evaluation.identify_flood_events --data-root datasets
+```
+
+Train the model — **with real forecast data, use the shifted split dates**
+(§3.13, §4.5) so training and validation actually overlap real forecast
+coverage:
+
+```bash
+python -m scripts.training.train_spatiotemporal_gnn --data-root datasets \
+    --train-end 2025-06-30 --val-end 2025-12-31
+```
+
+Evaluate the trained model against real historical flood events:
+
+```bash
+python -m scripts.evaluation.evaluate_flood_detection --data-root datasets
 ```
 
-Validate everything against whatever's actually in `datasets/`:
+Validate the graph itself against whatever's actually in `datasets/`:
 
 ```bash
 python -m src.test_build_graph --data-root datasets

datasets/c3s_seasonal/hindcast_month08.nc

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9c49bb3f1e6c982a0d8ccc40885f63b8354df868bb4bf25e5a4c5aac2294b644
+size 1638995

datasets/cavity_proximity_check.csv

@@ -0,0 +1,28 @@
+station_code,distance_to_nearest_cavity_km,n_cavities_within_20km
+H420021011,0.06734330830433045,2287
+H420021010,0.08011680750680154,2293
+H420021012,0.0956934778700636,2300
+H426021201,0.12426673679090168,2668
+H426021010,0.12530881046311096,2663
+H606000101,0.21627269892203413,4921
+H603021010,0.22211681771610073,1385
+H605021010,0.24516891878964753,3516
+H605641101,0.2695660117395612,3927
+H623022010,0.3230080907463569,5209
+H431021010,0.32400268298867924,2716
+H404021010,0.3493730085326225,837
+H605641401,0.3968900448567439,4298
+H438021010,0.41373353591738987,3058
+H404021101,0.41563171923773506,811
+H602021010,0.46718998412477725,830
+H416021010,0.6053264261702067,817
+H620021010,0.6184076860160554,5578
+H414000201,0.6776727899980145,309
+H605641201,0.6996777298184973,4547
+H605022010,0.7668131359487362,4277
+H600021010,0.8089347801714617,406
+H432021010,0.8598446846110105,2457
+H623021010,0.8948607263402845,6105
+H403301101,4.8485620480579215,858
+H402021010,6.098887227694296,104
+H403021010,6.141950442395451,765

datasets/cavity_proximity_check_naturelle.csv

@@ -0,0 +1,28 @@
+station_code,distance_to_nearest_cavity_km,n_cavities_within_20km
+H623022010,0.3230080907463569,1493
+H431021010,0.32400268298867924,225
+H605641401,0.3968900448567439,1225
+H602021010,0.46718998412477725,462
+H605021010,0.5271133764669997,1144
+H620021010,0.6184076860160554,1500
+H404021101,0.6311460233156068,39
+H605641101,0.7603711475520486,1171
+H605022010,0.7668131359487362,1200
+H404021010,0.7915529843965451,64
+H605641201,0.845518793923525,1279
+H438021010,1.2862986368668676,1131
+H603021010,1.3117713525800223,637
+H432021010,1.499596307034988,434
+H600021010,1.8306899198118753,292
+H606000101,1.8604021674075912,1424
+H623021010,1.8642663958164236,1567
+H420021011,2.419255283585594,188
+H420021010,2.419581430251758,188
+H420021012,2.443729561571243,188
+H426021201,3.4018664278664312,212
+H426021010,3.4155339402279665,211
+H416021010,3.4260842668004514,85
+H403301101,4.8485620480579215,86
+H414000201,6.833629159522594,23
+H403021010,6.879985125697823,16
+H402021010,13.02124359582901,7

datasets/climate_anomaly_models/cross_validation/cross_validation_summary.csv

@@ -0,0 +1,91 @@
+variable,regime,horizon_native,n_folds,n_folds_beat_climatology,mean_rmse_margin_pct,std_rmse_margin_pct,fraction_folds_beat,verdict
+evap_mm,daily,1,4,4,15.433514989628518,5.348543414069798,1.0,CONSISTENT WIN
+evap_mm,daily,5,4,4,1.5941790604402408,0.8738870160044205,1.0,CONSISTENT WIN
+evap_mm,daily,10,4,4,0.6855695594151723,0.5058997897142183,1.0,CONSISTENT WIN
+evap_mm,daily,14,4,4,0.31703221488214517,0.10421165892974571,1.0,CONSISTENT WIN
+evap_mm,monthly,3,4,2,-9.142908958079303,33.73357889735841,0.5,UNSTABLE
+evap_mm,monthly,6,4,2,-2.605919560987279,29.582925646143334,0.5,UNSTABLE
+evap_mm,weekly,3,4,2,-0.4781177859291126,4.091387621523481,0.5,UNSTABLE
+evap_mm,weekly,6,4,2,-1.252800527841585,5.013289959528252,0.5,UNSTABLE
+evap_mm,weekly,9,4,2,-1.2027990617044706,5.424053622684331,0.5,UNSTABLE
+evap_mm,weekly,13,4,2,-1.4548025160472695,5.947546001336505,0.5,UNSTABLE
+precip_mm,daily,1,4,4,0.398346029372596,0.35931526061206814,1.0,CONSISTENT WIN
+precip_mm,daily,5,4,2,0.05693663929914346,0.14641059041695165,0.5,UNSTABLE
+precip_mm,daily,10,4,1,-0.07197716314093104,0.08575379979797601,0.25,UNSTABLE
+precip_mm,daily,14,4,2,-0.08587614053700632,0.11792945958011258,0.5,UNSTABLE
+precip_mm,monthly,3,4,2,-2.3791358481673317,11.06489055412864,0.5,UNSTABLE
+precip_mm,monthly,6,4,2,-0.9583540316061492,10.912475516103187,0.5,UNSTABLE
+precip_mm,weekly,3,4,2,-0.015214069249200879,1.149273008516348,0.5,UNSTABLE
+precip_mm,weekly,6,4,2,-0.04972394083088033,1.0399297273887897,0.5,UNSTABLE
+precip_mm,weekly,9,4,2,0.051414901527331025,1.037358514987821,0.5,UNSTABLE
+precip_mm,weekly,13,4,2,0.04422228789871496,1.0343175452259066,0.5,UNSTABLE
+runoff_mm,daily,1,4,3,4.877954803864734,39.979028236007835,0.75,UNSTABLE
+runoff_mm,daily,5,4,3,11.53760137241838,35.68889748590784,0.75,UNSTABLE
+runoff_mm,daily,10,4,4,12.647406802587517,20.170513319198708,1.0,CONSISTENT WIN
+runoff_mm,daily,14,4,3,7.690957745195579,13.457090046067707,0.75,UNSTABLE
+runoff_mm,monthly,3,4,2,-19.47792381951619,35.03362297433532,0.5,UNSTABLE
+runoff_mm,monthly,6,4,1,-53.75021188161636,48.01535561591501,0.25,UNSTABLE
+runoff_mm,weekly,3,4,2,-219.20442424909095,457.2589229773,0.5,UNSTABLE
+runoff_mm,weekly,6,4,2,-128.77067123736416,259.1345908605477,0.5,UNSTABLE
+runoff_mm,weekly,9,4,2,-48.45571601522676,96.01623745113712,0.5,UNSTABLE
+runoff_mm,weekly,13,4,1,-5.3300287516180065,9.188865977661955,0.25,UNSTABLE
+snow_mm,daily,1,4,1,-19.972510335125083,38.97547730423528,0.25,UNSTABLE
+snow_mm,daily,5,4,1,-7.17907940104464,24.55281983704293,0.25,UNSTABLE
+snow_mm,daily,10,4,1,-2.953434915675786,14.982349351868141,0.25,UNSTABLE
+snow_mm,daily,14,4,1,0.7458644892012884,7.491445927922103,0.25,UNSTABLE
+snow_mm,monthly,3,4,2,-4.265509211224067,21.69877156533628,0.5,UNSTABLE
+snow_mm,monthly,6,4,0,-99.74973771226338,66.67228584800623,0.0,CONSISTENT LOSS
+snow_mm,weekly,3,4,2,-7.151926922617358,20.09063564835978,0.5,UNSTABLE
+snow_mm,weekly,6,4,2,-4.6320344756553045,13.423756666597912,0.5,UNSTABLE
+snow_mm,weekly,9,4,1,-1.6645630009629744,6.414858145946501,0.25,UNSTABLE
+snow_mm,weekly,13,4,1,-1.4416367620038266,8.265961451641886,0.25,UNSTABLE
+solar_Wm2,daily,1,4,4,7.494190896684461,2.112615474319945,1.0,CONSISTENT WIN
+solar_Wm2,daily,5,4,3,0.6368114447161746,0.667452764839831,0.75,UNSTABLE
+solar_Wm2,daily,10,4,2,0.12611177008512928,0.34470089507591656,0.5,UNSTABLE
+solar_Wm2,daily,14,4,3,0.15097146474936807,0.1720920233122145,0.75,UNSTABLE
+solar_Wm2,monthly,3,4,3,1.9645999393333369,5.8952762743345675,0.75,UNSTABLE
+solar_Wm2,monthly,6,4,2,-2.2430165083916958,5.899020662126656,0.5,UNSTABLE
+solar_Wm2,weekly,3,4,0,-1.1531262450252477,0.6458618480667215,0.0,CONSISTENT LOSS
+solar_Wm2,weekly,6,4,0,-1.0188692244288324,0.5387379039757494,0.0,CONSISTENT LOSS
+solar_Wm2,weekly,9,4,2,0.03659436540832812,0.9212770495827473,0.5,UNSTABLE
+solar_Wm2,weekly,13,4,2,-0.23780402929158306,1.5433966577330025,0.5,UNSTABLE
+temp_C,daily,1,4,4,30.022315857033007,3.881342677328426,1.0,CONSISTENT WIN
+temp_C,daily,5,4,4,4.3798662093635965,2.0394131024283917,1.0,CONSISTENT WIN
+temp_C,daily,10,4,4,0.8635724471007284,0.6019941605242723,1.0,CONSISTENT WIN
+temp_C,daily,14,4,2,0.1062136867173081,0.6007895705700086,0.5,UNSTABLE
+temp_C,monthly,3,4,3,2.9646108671036138,3.178287043631982,0.75,UNSTABLE
+temp_C,monthly,6,4,2,-17.783839448067383,48.79274270360168,0.5,UNSTABLE
+temp_C,weekly,3,4,4,1.8297437524107794,1.2441291862764345,1.0,CONSISTENT WIN
+temp_C,weekly,6,4,4,2.1199368956334927,1.1740608784052555,1.0,CONSISTENT WIN
+temp_C,weekly,9,4,4,2.25291486752172,1.3097993338486495,1.0,CONSISTENT WIN
+temp_C,weekly,13,4,4,2.32855140914343,1.2820550187897695,1.0,CONSISTENT WIN
+wind_speed_ms,daily,1,4,4,8.110292330816359,3.0343494658345915,1.0,CONSISTENT WIN
+wind_speed_ms,daily,5,4,3,0.3347707536849176,0.5327603028429463,0.75,UNSTABLE
+wind_speed_ms,daily,10,4,1,0.015021426922881866,0.06289696631058019,0.25,UNSTABLE
+wind_speed_ms,daily,14,4,1,-0.04207062300455699,0.20427513206685075,0.25,UNSTABLE
+wind_speed_ms,monthly,3,4,1,-8.183872269480815,15.064771776907126,0.25,UNSTABLE
+wind_speed_ms,monthly,6,4,1,-1.6041295144038523,6.789908736004105,0.25,UNSTABLE
+wind_speed_ms,weekly,3,4,1,0.2742306713673833,1.329372444317103,0.25,UNSTABLE
+wind_speed_ms,weekly,6,4,2,0.006060444506114593,0.7274930140084717,0.5,UNSTABLE
+wind_speed_ms,weekly,9,4,3,0.5271060954048611,3.0150369917061477,0.75,UNSTABLE
+wind_speed_ms,weekly,13,4,2,0.2478600372184011,1.3342068832530227,0.5,UNSTABLE
+wind_u_ms,daily,1,4,4,16.769026898019185,3.2088757865950726,1.0,CONSISTENT WIN
+wind_u_ms,daily,5,4,4,0.980008964943553,0.40556019208605765,1.0,CONSISTENT WIN
+wind_u_ms,daily,10,4,1,-0.03345921924049753,0.19852298388169823,0.25,UNSTABLE
+wind_u_ms,daily,14,4,3,-0.04383528473376304,0.25837623408189836,0.75,UNSTABLE
+wind_u_ms,monthly,3,4,1,-3.568035503830248,4.242051580248167,0.25,UNSTABLE
+wind_u_ms,monthly,6,4,1,-1.4728956447857082,3.5434369207340533,0.25,UNSTABLE
+wind_u_ms,weekly,3,4,2,0.012038768089962987,0.5975149620546968,0.5,UNSTABLE
+wind_u_ms,weekly,6,4,2,-0.07086904111307483,0.8461159046113296,0.5,UNSTABLE
+wind_u_ms,weekly,9,4,2,0.11986003388474414,0.7117612240800026,0.5,UNSTABLE
+wind_u_ms,weekly,13,4,2,-0.07751618069339203,1.244023547366199,0.5,UNSTABLE
+wind_v_ms,daily,1,4,4,13.69755565805852,1.9976696042728612,1.0,CONSISTENT WIN
+wind_v_ms,daily,5,4,3,0.534129232788973,0.4873797436434326,0.75,UNSTABLE
+wind_v_ms,daily,10,4,2,-0.02875817079336135,0.11595203795682225,0.5,UNSTABLE
+wind_v_ms,daily,14,4,3,0.11651430428134665,0.4074932178939845,0.75,UNSTABLE
+wind_v_ms,monthly,3,4,1,-1.1359041475356935,2.2695141308167246,0.25,UNSTABLE
+wind_v_ms,monthly,6,4,1,-1.3578007824004086,2.5700897827163307,0.25,UNSTABLE
+wind_v_ms,weekly,3,4,2,-0.24031923720682136,0.9738638117861981,0.5,UNSTABLE
+wind_v_ms,weekly,6,4,1,-0.25925672289960094,0.4846777886552502,0.25,UNSTABLE
+wind_v_ms,weekly,9,4,2,-0.21047109882084625,0.7352878874865387,0.5,UNSTABLE
+wind_v_ms,weekly,13,4,2,-0.1689636176525115,0.6241016168050836,0.5,UNSTABLE

datasets/climate_anomaly_models/cross_validation/raw_fold_results.csv

@@ -0,0 +1,361 @@
+fold,variable,regime,horizon_native,model_rmse,climatology_only_rmse,model_beats_climatology,rmse_margin_pct
+1,wind_u_ms,daily,1,2.8739426136016846,3.3672165870666504,True,14.649309324491101
+1,wind_u_ms,daily,5,3.3327529430389404,3.3494937419891357,True,0.49980087260152917
+1,wind_u_ms,daily,10,3.3171987533569336,3.325258255004883,True,0.2423722017927111
+1,wind_u_ms,daily,14,3.2852590084075928,3.2890868186950684,True,0.11637911975197583
+1,wind_u_ms,weekly,3,2.233687400817871,2.2404446601867676,True,0.3016034936713495
+1,wind_u_ms,weekly,6,2.1911985874176025,2.210965633392334,True,0.8940458266826348
+1,wind_u_ms,weekly,9,2.2114953994750977,2.2111427783966064,False,-0.01594745856922507
+1,wind_u_ms,weekly,13,2.1828994750976562,2.21201491355896,True,1.3162406041132533
+1,wind_u_ms,monthly,3,1.5748039484024048,1.6138828992843628,True,2.4214241875471028
+1,wind_u_ms,monthly,6,0.9909133315086365,0.931471049785614,False,-6.381549027927772
+1,wind_v_ms,daily,1,2.832612991333008,3.1754367351531982,True,10.796113177913808
+1,wind_v_ms,daily,5,3.1707751750946045,3.170428991317749,False,-0.010919146203983638
+1,wind_v_ms,daily,10,3.1927731037139893,3.1901307106018066,False,-0.08283024590187213
+1,wind_v_ms,daily,14,3.164053201675415,3.18038272857666,True,0.5134453395976398
+1,wind_v_ms,weekly,3,1.7792274951934814,1.7929102182388306,True,0.763157179102343
+1,wind_v_ms,weekly,6,1.8972629308700562,1.8864867687225342,False,-0.571229140123746
+1,wind_v_ms,weekly,9,1.880104899406433,1.885765552520752,True,0.3001779890799312
+1,wind_v_ms,weekly,13,1.8449993133544922,1.835805058479309,False,-0.5008295860563308
+1,wind_v_ms,monthly,3,1.1433367729187012,1.094919204711914,False,-4.422022008420825
+1,wind_v_ms,monthly,6,1.0670667886734009,1.0154725313186646,False,-5.080812701820447
+1,temp_C,daily,1,2.8351054191589355,3.8317997455596924,True,26.01112773588262
+1,temp_C,daily,5,3.7103633880615234,3.830357551574707,True,3.1327144241104206
+1,temp_C,daily,10,3.814448356628418,3.8454227447509766,True,0.8054872033208522
+1,temp_C,daily,14,3.8655846118927,3.8470683097839355,False,-0.4813094184388056
+1,temp_C,weekly,3,2.7701127529144287,2.8661770820617676,True,3.3516536625935043
+1,temp_C,weekly,6,2.6957316398620605,2.7946975231170654,True,3.5412019524969307
+1,temp_C,weekly,9,2.6915500164031982,2.8015084266662598,True,3.924971605168787
+1,temp_C,weekly,13,2.696559429168701,2.8093483448028564,True,4.014771462670648
+1,temp_C,monthly,3,2.2958106994628906,2.4492390155792236,True,6.264325986169568
+1,temp_C,monthly,6,1.212746024131775,1.4310722351074219,True,15.25612793118431
+1,precip_mm,daily,1,0.2669251263141632,0.26778000593185425,True,0.31924699333548945
+1,precip_mm,daily,5,0.2669302821159363,0.2676481008529663,True,0.26819496747498545
+1,precip_mm,daily,10,0.26824191212654114,0.2679697871208191,False,-0.10155062951158496
+1,precip_mm,daily,14,0.2865450382232666,0.28626635670661926,False,-0.09735042561531122
+1,precip_mm,weekly,3,0.12012144923210144,0.12172988057136536,True,1.3213118518759714
+1,precip_mm,weekly,6,0.11948014795780182,0.12094420194625854,True,1.2105201943515054
+1,precip_mm,weekly,9,0.12284043431282043,0.12438424676656723,True,1.2411639688135743
+1,precip_mm,weekly,13,0.12692475318908691,0.1282644122838974,True,1.0444511232354274
+1,precip_mm,monthly,3,0.05582819879055023,0.060593485832214355,True,7.864355344829282
+1,precip_mm,monthly,6,0.05131503939628601,0.05966459959745407,True,13.994161123180222
+1,evap_mm,daily,1,0.10997074097394943,0.13233686983585358,True,16.900905159421086
+1,evap_mm,daily,5,0.12870867550373077,0.132272407412529,True,2.694236824225713
+1,evap_mm,daily,10,0.13091085851192474,0.1320226937532425,True,0.842154640016533
+1,evap_mm,daily,14,0.13138845562934875,0.13184316456317902,True,0.3448862406608615
+1,evap_mm,weekly,3,0.09260445088148117,0.0965011790394783,True,4.038010930833283
+1,evap_mm,weekly,6,0.09022482484579086,0.09441053867340088,True,4.433523933265402
+1,evap_mm,weekly,9,0.08938121795654297,0.09414008259773254,True,5.055088661356451
+1,evap_mm,weekly,13,0.08798616379499435,0.09332296997308731,True,5.71864159448847
+1,evap_mm,monthly,3,0.05771525204181671,0.07399329543113708,True,21.999349122745542
+1,evap_mm,monthly,6,0.04062597081065178,0.05735654756426811,True,29.16942784059604
+1,snow_mm,daily,1,0.003109490731731057,0.00312636187300086,True,0.5396413452806461
+1,snow_mm,daily,5,0.002658030018210411,0.0030897397082298994,True,13.97236436679688
+1,snow_mm,daily,10,0.0026892637833952904,0.003054675180464983,True,11.962365079159403
+1,snow_mm,daily,14,0.0027143952902406454,0.003036443144083023,True,10.606088721599727
+1,snow_mm,weekly,3,0.004058920778334141,0.0029661590233445168,False,-36.84096996787015
+1,snow_mm,weekly,6,0.003663738491013646,0.0029460384976118803,False,-24.361527997123876
+1,snow_mm,weekly,9,0.0032524289563298225,0.0029474864713847637,False,-10.345848501954048
+1,snow_mm,weekly,13,0.0033864439465105534,0.0030179170425981283,False,-12.211300002970255
+1,snow_mm,monthly,3,0.004496293608099222,0.0037231631577014923,False,-20.765419554565657
+1,snow_mm,monthly,6,0.0018420011037960649,0.0007720344001427293,False,-138.59054770817548
+1,runoff_mm,daily,1,0.020946914330124855,0.02228635363280773,True,6.01013214073337
+1,runoff_mm,daily,5,0.022135987877845764,0.022257430478930473,True,0.5456272286222358
+1,runoff_mm,daily,10,0.022578537464141846,0.022582419216632843,True,0.017189267694305534
+1,runoff_mm,daily,14,0.03948547691106796,0.03948530554771423,False,-0.0004339927255271502
+1,runoff_mm,weekly,3,0.15213415026664734,0.015139754861593246,False,-904.8653472757574
+1,runoff_mm,weekly,6,0.09193050861358643,0.014895575121045113,False,-517.1665603142978
+1,runoff_mm,weekly,9,0.04304801672697067,0.014731019735336304,False,-192.22699786158356
+1,runoff_mm,weekly,13,0.02409614436328411,0.020256971940398216,False,-18.95235099392858
+1,runoff_mm,monthly,3,0.017046168446540833,0.019594138488173485,True,13.003736005900647
+1,runoff_mm,monthly,6,0.010098992846906185,0.005454735364764929,False,-85.14175613616409
+1,solar_Wm2,daily,1,6.423816680908203,7.0748138427734375,True,9.201615425262318
+1,solar_Wm2,daily,5,6.964319705963135,7.067264556884766,True,1.4566435159321207
+1,solar_Wm2,daily,10,7.034942150115967,7.033107757568359,False,-0.02608224714932632
+1,solar_Wm2,daily,14,7.0255866050720215,7.02707052230835,True,0.021117153038627358
+1,solar_Wm2,weekly,3,4.5501909255981445,4.508223056793213,False,-0.9309181971751017
+1,solar_Wm2,weekly,6,4.4641828536987305,4.420793056488037,False,-0.9814935161240741
+1,solar_Wm2,weekly,9,4.347484111785889,4.398562908172607,True,1.161261017588573
+1,solar_Wm2,weekly,13,4.309525489807129,4.3671417236328125,True,1.3193122062857043
+1,solar_Wm2,monthly,3,2.5916008949279785,2.8515241146087646,True,9.115238350928278
+1,solar_Wm2,monthly,6,2.1289284229278564,2.0291895866394043,False,-4.915205407378043
+1,wind_speed_ms,daily,1,1.8817723989486694,2.084139585494995,True,9.709867225532417
+1,wind_speed_ms,daily,5,2.0680997371673584,2.076103687286377,True,0.38552747476116267
+1,wind_speed_ms,daily,10,2.068087577819824,2.0673975944519043,False,-0.03337448828283299
+1,wind_speed_ms,daily,14,2.049168348312378,2.048762798309326,False,-0.019794873441982868
+1,wind_speed_ms,weekly,3,1.1857573986053467,1.1841224431991577,False,-0.13807317102881578
+1,wind_speed_ms,weekly,6,1.2243714332580566,1.2338104248046875,True,0.7650277025439345
+1,wind_speed_ms,weekly,9,1.2273221015930176,1.2445919513702393,True,1.3875913112091363
+1,wind_speed_ms,weekly,13,1.2473136186599731,1.2533409595489502,True,0.48090193199671366
+1,wind_speed_ms,monthly,3,0.9937477111816406,0.9493570327758789,False,-4.675867652864517
+1,wind_speed_ms,monthly,6,0.6234548687934875,0.5675956606864929,False,-9.841373353600748
+2,wind_u_ms,daily,1,2.6418282985687256,3.2315168380737305,True,18.248041679910024
+2,wind_u_ms,daily,5,3.1940410137176514,3.231450080871582,True,1.1576557340424936
+2,wind_u_ms,daily,10,3.2238967418670654,3.2215042114257812,False,-0.07426749382473374
+2,wind_u_ms,daily,14,3.193862199783325,3.1952741146087646,True,0.04418759626863283
+2,wind_u_ms,weekly,3,2.2445132732391357,2.228588104248047,False,-0.7145855692549439
+2,wind_u_ms,weekly,6,2.2226250171661377,2.1988980770111084,False,-1.0790377418165995
+2,wind_u_ms,weekly,9,2.1677985191345215,2.1504316329956055,False,-0.8076000125948438
+2,wind_u_ms,weekly,13,2.2135322093963623,2.179701328277588,False,-1.5520879250694297
+2,wind_u_ms,monthly,3,1.6880786418914795,1.6302248239517212,False,-3.548824498912895
+2,wind_u_ms,monthly,6,0.8178536891937256,0.8347481489181519,True,2.0238990342562344
+2,wind_v_ms,daily,1,2.549076557159424,3.009953498840332,True,15.311762851435207
+2,wind_v_ms,daily,5,3.011502742767334,3.020596504211426,True,0.30105846416139803
+2,wind_v_ms,daily,10,3.022494077682495,3.017493724822998,False,-0.16571212123366996
+2,wind_v_ms,daily,14,3.0280661582946777,3.0147788524627686,False,-0.44073898889979274
+2,wind_v_ms,weekly,3,1.88368558883667,1.856187105178833,False,-1.4814499885876318
+2,wind_v_ms,weekly,6,1.896781086921692,1.8858832120895386,False,-0.5778658382603974
+2,wind_v_ms,weekly,9,1.8937994241714478,1.8805522918701172,False,-0.704427755537547
+2,wind_v_ms,weekly,13,1.953708529472351,1.9366624355316162,False,-0.8801788906519318
+2,wind_v_ms,monthly,3,0.8344284892082214,0.8305781483650208,False,-0.4635735783297472
+2,wind_v_ms,monthly,6,1.1044340133666992,1.1022324562072754,False,-0.19973619421435582
+2,temp_C,daily,1,2.407862424850464,3.5607821941375732,True,32.37827270607177
+2,temp_C,daily,5,3.4259023666381836,3.575531482696533,True,4.184807679151097
+2,temp_C,daily,10,3.5823936462402344,3.597421169281006,True,0.4177304333752765
+2,temp_C,daily,14,3.6087841987609863,3.602736711502075,False,-0.16785815182119643
+2,temp_C,weekly,3,2.514961004257202,2.534212827682495,True,0.7596766623148424
+2,temp_C,weekly,6,2.5919442176818848,2.6234023571014404,True,1.1991351358818367
+2,temp_C,weekly,9,2.5600292682647705,2.6129696369171143,True,2.0260613787615576
+2,temp_C,weekly,13,2.7696034908294678,2.816328763961792,True,1.6590844694777305
+2,temp_C,monthly,3,2.0094735622406006,2.08955454826355,True,3.832442952470309
+2,temp_C,monthly,6,1.8301355838775635,1.9805843830108643,True,7.596182239132374
+2,precip_mm,daily,1,0.31503328680992126,0.31563422083854675,True,0.19038937762482924
+2,precip_mm,daily,5,0.3163193464279175,0.316197007894516,False,-0.03869060438494145
+2,precip_mm,daily,10,0.3168396055698395,0.31681838631629944,False,-0.0066976079850537985
+2,precip_mm,daily,14,0.31729981303215027,0.31730714440345764,True,0.0023104967652575674
+2,precip_mm,weekly,3,0.12507475912570953,0.12552420794963837,True,0.3580574864962754
+2,precip_mm,weekly,6,0.13432995975017548,0.13413164019584656,False,-0.14785441677992656
+2,precip_mm,weekly,9,0.1362466663122177,0.1358945369720459,False,-0.2591195702327964
+2,precip_mm,weekly,13,0.128716379404068,0.1286322921514511,False,-0.06537025128797234
+2,precip_mm,monthly,3,0.05491117760539055,0.058735284954309464,True,6.510749631833252
+2,precip_mm,monthly,6,0.07407385110855103,0.07431261986494064,True,0.3213031068256883
+2,evap_mm,daily,1,0.1056857630610466,0.12314990907907486,True,14.18120902291084
+2,evap_mm,daily,5,0.12201866507530212,0.12303939461708069,True,0.8295957119711509
+2,evap_mm,daily,10,0.12280278652906418,0.12293122708797455,True,0.1044816373779891
+2,evap_mm,daily,14,0.12257875502109528,0.12282748520374298,True,0.20250368411851632
+2,evap_mm,weekly,3,0.08262113481760025,0.08342373371124268,True,0.9620750090381804
+2,evap_mm,weekly,6,0.08130913972854614,0.08230140060186386,True,1.2056427546328379
+2,evap_mm,weekly,9,0.08077941834926605,0.08197318762540817,True,1.4562923691552323
+2,evap_mm,weekly,13,0.07983306795358658,0.08075778931379318,True,1.145055316724308
+2,evap_mm,monthly,3,0.0606376938521862,0.07345578074455261,True,17.450072359780865
+2,evap_mm,monthly,6,0.061921149492263794,0.07167094200849533,True,13.603550118088123
+2,snow_mm,daily,1,0.008721661753952503,0.004888464696705341,False,-78.41310708105566
+2,snow_mm,daily,5,0.006938621401786804,0.004863916430622339,False,-42.655029146933884
+2,snow_mm,daily,10,0.0059899999760091305,0.004839463625103235,False,-23.774046878621842
+2,snow_mm,daily,14,0.005193319637328386,0.004825499374419451,False,-7.6224289833876
+2,snow_mm,weekly,3,0.0030265592504292727,0.003262108424678445,True,7.220764719749956
+2,snow_mm,weekly,6,0.0030508071649819613,0.0032240336295217276,True,5.372973251692283
+2,snow_mm,weekly,9,0.0030535596888512373,0.003217357210814953,True,5.091058009136195
+2,snow_mm,weekly,13,0.0029975948855280876,0.0032544971909374,True,7.8937633169477746
+2,snow_mm,monthly,3,0.003329826984554529,0.0038593532517552376,True,13.720595982238198
+2,snow_mm,monthly,6,0.0015523867914453149,0.0007244865992106497,False,-114.27405187848716
+2,runoff_mm,daily,1,0.013180810026824474,0.009204266592860222,False,-43.20326224632465
+2,runoff_mm,daily,5,0.010970048606395721,0.009195654653012753,False,-19.296004692842917
+2,runoff_mm,daily,10,0.008678611367940903,0.009186448529362679,True,5.5281119770994644
+2,runoff_mm,daily,14,0.009069954976439476,0.009186554700136185,True,1.2692432310339388
+2,runoff_mm,weekly,3,0.0050124963745474815,0.006684175692498684,True,25.009505956392566
+2,runoff_mm,weekly,6,0.006080025341361761,0.006825813092291355,True,10.92599139246634
+2,runoff_mm,weekly,9,0.006508655846118927,0.0067115058191120625,True,3.0224211743285454
+2,runoff_mm,weekly,13,0.006666881963610649,0.00661603594198823,False,-0.7685269860722563
+2,runoff_mm,monthly,3,0.008079104125499725,0.008321375586092472,True,2.9114352319063137
+2,runoff_mm,monthly,6,0.006043288391083479,0.0063767642714083195,True,5.2295469321338395
+2,solar_Wm2,daily,1,6.168939590454102,6.648454189300537,True,7.212422394639133
+2,solar_Wm2,daily,5,6.640743732452393,6.635233402252197,False,-0.08304651646956294
+2,solar_Wm2,daily,10,6.622495651245117,6.603960990905762,False,-0.28065974897306833
+2,solar_Wm2,daily,14,6.5959153175354,6.595357418060303,False,-0.00845897257319126
+2,solar_Wm2,weekly,3,3.923182964324951,3.8419766426086426,False,-2.113659953465276
+2,solar_Wm2,weekly,6,3.7402150630950928,3.69185471534729,False,-1.3099201208207218
+2,solar_Wm2,weekly,9,3.6102983951568604,3.6155154705047607,True,0.14429686141467504
+2,solar_Wm2,weekly,13,3.5327775478363037,3.5402140617370605,True,0.2100583120419559
+2,solar_Wm2,monthly,3,2.580043315887451,2.648366689682007,True,2.5798305824016894
+2,solar_Wm2,monthly,6,2.492854118347168,2.568377733230591,True,2.9405182075156344
+2,wind_speed_ms,daily,1,1.7070062160491943,1.8969701528549194,True,10.014070939378326
+2,wind_speed_ms,daily,5,1.8803834915161133,1.8922932147979736,True,0.6293804357974124
+2,wind_speed_ms,daily,10,1.8924022912979126,1.8944388628005981,True,0.10750262479702462
+2,wind_speed_ms,daily,14,1.8768984079360962,1.8813806772232056,True,0.23824361233075436
+2,wind_speed_ms,weekly,3,1.133108377456665,1.123690128326416,False,-0.8381535881494507
+2,wind_speed_ms,weekly,6,1.1608633995056152,1.153270959854126,False,-0.6583396197238509
+2,wind_speed_ms,weekly,9,1.1428027153015137,1.1464357376098633,True,0.3168971612769055
+2,wind_speed_ms,weekly,13,1.161649465560913,1.1590807437896729,False,-0.22161715523300554
+2,wind_speed_ms,monthly,3,0.4962719678878784,0.49306195974349976,False,-0.6510354491854471
+2,wind_speed_ms,monthly,6,0.45714855194091797,0.4569828510284424,False,-0.036259766007121524
+3,wind_u_ms,daily,1,2.723127841949463,3.1526782512664795,True,13.624936485176043
+3,wind_u_ms,daily,5,3.1262266635894775,3.152287721633911,True,0.8267347509422611
+3,wind_u_ms,daily,10,3.1570487022399902,3.1548163890838623,False,-0.07075889309603145
+3,wind_u_ms,daily,14,3.1233091354370117,3.1262121200561523,True,0.09285948962057258
+3,wind_u_ms,weekly,3,1.996690034866333,2.0098955631256104,True,0.657025593844353
+3,wind_u_ms,weekly,6,2.0099024772644043,2.015272378921509,True,0.2664603412059983
+3,wind_u_ms,weekly,9,1.9999929666519165,2.017063856124878,True,0.846323700716025
+3,wind_u_ms,weekly,13,1.9591344594955444,1.9684662818908691,True,0.4740656459891579
+3,wind_u_ms,monthly,3,1.2181833982467651,1.1449017524719238,False,-6.4006929517420215
+3,wind_u_ms,monthly,6,0.742477297782898,0.7398920655250549,False,-0.34940667406785075
+3,wind_v_ms,daily,1,2.7843217849731445,3.2412281036376953,True,14.096703596755676
+3,wind_v_ms,daily,5,3.1963932514190674,3.231754779815674,True,1.0941897144381518
+3,wind_v_ms,daily,10,3.232675790786743,3.2352981567382812,True,0.08105484640036599
+3,wind_v_ms,daily,14,3.215407609939575,3.2247416973114014,True,0.2894522491400902
+3,wind_v_ms,weekly,3,1.9473044872283936,1.9378350973129272,False,-0.48865819019362944
+3,wind_v_ms,weekly,6,1.9386684894561768,1.9321800470352173,False,-0.33580941025219113
+3,wind_v_ms,weekly,9,1.9925246238708496,1.9734578132629395,False,-0.9661625639914165
+3,wind_v_ms,weekly,13,1.8821989297866821,1.8900978565216064,True,0.4179109937440438
+3,wind_v_ms,monthly,3,1.239267110824585,1.2336450815200806,False,-0.4557250208120643
+3,wind_v_ms,monthly,6,0.6757810711860657,0.6698251962661743,False,-0.8891685402536902
+3,temp_C,daily,1,2.5399370193481445,3.504300832748413,True,27.519435671392422
+3,temp_C,daily,5,3.4045639038085938,3.5056092739105225,True,2.8823911111238063
+3,temp_C,daily,10,3.4806363582611084,3.4981226921081543,True,0.4998776597086052
+3,temp_C,daily,14,3.490513563156128,3.495943546295166,True,0.1553223919989361
+3,temp_C,weekly,3,2.769240379333496,2.793513059616089,True,0.8688944624418028
+3,temp_C,weekly,6,2.759183883666992,2.7902677059173584,True,1.1140086015562711
+3,temp_C,weekly,9,2.7689220905303955,2.7895116806030273,True,0.738107325945337
+3,temp_C,weekly,13,2.725949287414551,2.755459785461426,True,1.0709827159365786
+3,temp_C,monthly,3,1.255995750427246,1.2963720560073853,True,3.114561548363793
+3,temp_C,monthly,6,1.6863813400268555,0.8875446319580078,False,-90.0052436018387
+3,precip_mm,daily,1,0.38922742009162903,0.3898385465145111,True,0.15676398045962134
+3,precip_mm,daily,5,0.3898153603076935,0.38998299837112427,True,0.04298599275634439
+3,precip_mm,daily,10,0.38967365026474,0.3896752595901489,True,0.00041299142538025037
+3,precip_mm,daily,14,0.3896777629852295,0.3896777927875519,True,7.64793964123668e-06
+3,precip_mm,weekly,3,0.1623513251543045,0.16181111335754395,False,-0.33385333402093753
+3,precip_mm,weekly,6,0.16087333858013153,0.1609797328710556,True,0.06609173032315431
+3,precip_mm,weekly,9,0.1635473668575287,0.16424912214279175,True,0.4272505545892556
+3,precip_mm,weekly,13,0.1626420021057129,0.1635376662015915,True,0.5476806149199436
+3,precip_mm,monthly,3,0.07796118408441544,0.06940451264381409,False,-12.32869609576315
+3,precip_mm,monthly,6,0.09049215167760849,0.08267532289028168,False,-9.454851234993683
+3,evap_mm,daily,1,0.11175497621297836,0.12269963324069977,True,8.919877540506809
+3,evap_mm,daily,5,0.12034324556589127,0.12266843765974045,True,1.895509666715439
+3,evap_mm,daily,10,0.12100701779127121,0.12259463220834732,True,1.2950113626329
+3,evap_mm,daily,14,0.12182147055864334,0.12236817926168442,True,0.446773586352004
+3,evap_mm,weekly,3,0.08173413574695587,0.08074971288442612,False,-1.2191038548195463
+3,evap_mm,weekly,6,0.08371792733669281,0.0805465579032898,False,-3.9373121781452136
+3,evap_mm,weekly,9,0.08340877294540405,0.07974432408809662,False,-4.595247247013062
+3,evap_mm,weekly,13,0.07733358442783356,0.07290568202733994,False,-6.073466809943757
+3,evap_mm,monthly,3,0.06917297840118408,0.048059068620204926,False,-43.933247953337315
+3,evap_mm,monthly,6,0.07606972008943558,0.05567074939608574,False,-36.642170106630736
+3,snow_mm,daily,1,0.04865730553865433,0.048644475638866425,False,-0.026374834180866114
+3,snow_mm,daily,5,0.04864303395152092,0.04864303395152092,False,0.0
+3,snow_mm,daily,10,0.0486416257917881,0.0486416257917881,False,0.0
+3,snow_mm,daily,14,0.04864085838198662,0.04864085838198662,False,0.0
+3,snow_mm,weekly,3,0.0034804767929017544,0.0035579833202064037,True,2.1783836608922913
+3,snow_mm,weekly,6,0.00347675709053874,0.0035346040967851877,True,1.6365908221251961
+3,snow_mm,weekly,9,0.01997566781938076,0.01994440332055092,False,-0.15675825607490776
+3,snow_mm,weekly,13,0.019968092441558838,0.01995086297392845,False,-0.08635951062819495
+3,snow_mm,monthly,3,0.0024621703196316957,0.00290302955545485,True,15.186178004794034
+3,snow_mm,monthly,6,0.015892282128334045,0.015629617497324944,False,-1.6805570005411694
+3,runoff_mm,daily,1,0.0327693447470665,0.03346773236989975,True,2.0867491562152227
+3,runoff_mm,daily,5,0.03284818306565285,0.03346148133277893,True,1.8328485252244204
+3,runoff_mm,daily,10,0.03267849609255791,0.03345860540866852,True,2.3315655466874263
+3,runoff_mm,daily,14,0.03289739415049553,0.033447980880737305,True,1.6460985558588932
+3,runoff_mm,weekly,3,0.014113156124949455,0.015081881545484066,True,6.423107207234856
+3,runoff_mm,weekly,6,0.014320099726319313,0.014998653903603554,True,4.524100506920907
+3,runoff_mm,weekly,9,0.014340928755700588,0.014929494820535183,True,3.942303955422761
+3,runoff_mm,weekly,13,0.014765613712370396,0.014902004972100258,True,0.9152544237182569
+3,runoff_mm,monthly,3,0.014373633079230785,0.01111630629748106,False,-29.302240281808636
+3,runoff_mm,monthly,6,0.013938525691628456,0.010297472588717937,False,-35.3587064353996
+3,solar_Wm2,daily,1,6.7541632652282715,7.081015586853027,True,4.615896090267142
+3,solar_Wm2,daily,5,7.001336574554443,7.061535835266113,True,0.8524952944517815
+3,solar_Wm2,daily,10,7.015348434448242,7.049310684204102,True,0.4817811453815057
+3,solar_Wm2,daily,14,6.998568058013916,7.022741317749023,True,0.34421401332287144
+3,solar_Wm2,weekly,3,4.185779094696045,4.150809288024902,False,-0.8424816522414215
+3,solar_Wm2,weekly,6,4.154299259185791,4.092688083648682,False,-1.5053963135686181
+3,solar_Wm2,weekly,9,4.080623626708984,4.0368332862854,False,-1.084769603252029
+3,solar_Wm2,weekly,13,3.7359890937805176,3.6498241424560547,False,-2.360797341498224
+3,solar_Wm2,monthly,3,2.6004652976989746,2.470052480697632,False,-5.279758953320275
+3,solar_Wm2,monthly,6,2.5720105171203613,2.353238582611084,False,-9.296632144562855
+3,wind_speed_ms,daily,1,1.8754488229751587,2.0637600421905518,True,9.124666403343701
+3,wind_speed_ms,daily,5,2.0497143268585205,2.0653021335601807,True,0.7547470391070481
+3,wind_speed_ms,daily,10,2.0700366497039795,2.0698740482330322,False,-0.00785562150924458
+3,wind_speed_ms,daily,14,2.065276622772217,2.061025619506836,False,-0.20625669206373298
+3,wind_speed_ms,weekly,3,1.1828533411026,1.2095273733139038,True,2.2053268739359986
+3,wind_speed_ms,weekly,6,1.2314677238464355,1.2375398874282837,True,0.4906640701873967
+3,wind_speed_ms,weekly,9,1.2210297584533691,1.269612193107605,True,3.826557031980101
+3,wind_speed_ms,weekly,13,1.264724612236023,1.2899187803268433,True,1.9531592589447022
+3,wind_speed_ms,monthly,3,0.7322767972946167,0.5619924664497375,False,-30.300109167052106
+3,wind_speed_ms,monthly,6,0.4035535156726837,0.43167153000831604,True,6.513752328093224
+4,wind_u_ms,daily,1,2.540086269378662,3.1972415447235107,True,20.553820102499568
+4,wind_u_ms,daily,5,3.144023895263672,3.1898248195648193,True,1.4358445021879283
+4,wind_u_ms,daily,10,3.191394805908203,3.1840338706970215,False,-0.23118269183393603
+4,wind_u_ms,daily,14,3.2001404762268066,3.1864778995513916,False,-0.4287673445762334
+4,wind_u_ms,weekly,3,2.180706739425659,2.176443338394165,False,-0.19588844590090665
+4,wind_u_ms,weekly,6,2.271817207336426,2.263556480407715,False,-0.3649445905243329
+4,wind_u_ms,weekly,9,2.2254221439361572,2.2356314659118652,True,0.45666390598702045
+4,wind_u_ms,weekly,13,2.222440004348755,2.2103211879730225,False,-0.5482830478065496
+4,wind_u_ms,monthly,3,0.6874726414680481,0.6440383791923523,False,-6.7440487522131765
+4,wind_u_ms,monthly,6,0.9342527985572815,0.9233158826828003,False,-1.1845259114034448
+4,wind_v_ms,daily,1,2.5541727542877197,2.9903318881988525,True,14.585643006129388
+4,wind_v_ms,daily,5,2.963116407394409,2.9855735301971436,True,0.7521878987603258
+4,wind_v_ms,daily,10,2.976463556289673,2.9780256748199463,True,0.05245483756173071
+4,wind_v_ms,daily,14,2.9683499336242676,2.9714372158050537,True,0.10389861728744934
+4,wind_v_ms,weekly,3,1.9218381643295288,1.926571249961853,True,0.2456740508516327
+4,wind_v_ms,weekly,6,1.897179126739502,1.9057143926620483,True,0.44787749703793084
+4,wind_v_ms,weekly,9,1.906007170677185,1.9161344766616821,True,0.5285279351656473
+4,wind_v_ms,weekly,13,1.920538067817688,1.9260705709457397,True,0.2872430123541728
+4,wind_v_ms,monthly,3,0.8928779363632202,0.9000577330589294,True,0.7977040174198632
+4,wind_v_ms,monthly,6,1.2509294748306274,1.2602365016937256,True,0.738514306686859
+4,temp_C,daily,1,3.072192907333374,4.667597770690918,True,34.180427314785206
+4,temp_C,daily,5,4.3629608154296875,4.707530975341797,True,7.319551623069063
+4,temp_C,daily,10,4.6274261474609375,4.70894718170166,True,1.7311944919981797
+4,temp_C,daily,14,4.667789936065674,4.711070537567139,True,0.9186999251302983
+4,temp_C,weekly,3,2.908005475997925,2.9776451587677,True,2.338750222292968
+4,temp_C,weekly,6,3.1458892822265625,3.230708360671997,True,2.625401892598933
+4,temp_C,weekly,9,3.614116907119751,3.7000513076782227,True,2.3225191602111974
+4,temp_C,weekly,13,3.9237308502197266,4.027204513549805,True,2.5693669884887624
+4,temp_C,monthly,3,2.8991498947143555,2.8604512214660645,False,-1.3528870185892148
+4,temp_C,monthly,6,4.70853328704834,4.528201103210449,False,-3.9824243607475145
+4,precip_mm,daily,1,0.24828442931175232,0.25060752034187317,True,0.9269837660704439
+4,precip_mm,daily,5,0.2507527768611908,0.2506406307220459,False,-0.04474379864981456
+4,precip_mm,daily,10,0.2510528266429901,0.25060155987739563,False,-0.18007340649246567
+4,precip_mm,daily,14,0.25034624338150024,0.24972574412822723,False,-0.24847228123761286
+4,precip_mm,weekly,3,0.10667863488197327,0.10519914329051971,False,-1.4063722813481128
+4,precip_mm,weekly,6,0.10682600736618042,0.10542631149291992,False,-1.3276532712182545
+4,precip_mm,weekly,9,0.10662048310041428,0.10535242408514023,False,-1.2036353470607093
+4,precip_mm,weekly,13,0.10638120770454407,0.10496432334184647,False,-1.3498723352725388
+4,precip_mm,monthly,3,0.06322187185287476,0.05666923522949219,False,-11.562952273568712
+4,precip_mm,monthly,6,0.09612517803907394,0.08843648433685303,False,-8.694029121436824
+4,evap_mm,daily,1,0.11323165893554688,0.14467184245586395,True,21.73206823567534
+4,evap_mm,daily,5,0.14505314826965332,0.14645527303218842,True,0.9573740388486605
+4,evap_mm,daily,10,0.14682018756866455,0.14755891263484955,True,0.500630597633267
+4,evap_mm,daily,14,0.14760242402553558,0.14800791442394257,True,0.273965348397199
+4,evap_mm,weekly,3,0.08240608870983124,0.077967070043087,False,-5.693453228768368
+4,evap_mm,weekly,6,0.08620768040418625,0.08078456670045853,False,-6.7130566211193665
+4,evap_mm,weekly,9,0.09779331088066101,0.09162911772727966,False,-6.727330030316503
+4,evap_mm,weekly,13,0.11587555706501007,0.10869164764881134,False,-6.609440165458099
+4,evap_mm,monthly,3,0.0689452514052391,0.052196528762578964,False,-32.0878093615063
+4,evap_mm,monthly,6,0.15747159719467163,0.13510556519031525,False,-16.554486096002538
+4,snow_mm,daily,1,0.023550182580947876,0.023090632632374763,False,-1.9902007705444575
+4,snow_mm,daily,5,0.023093774914741516,0.023086005821824074,False,-0.03365282404155813
+4,snow_mm,daily,10,0.023081431165337563,0.02308095619082451,False,-0.0020578632407036632
+4,snow_mm,daily,14,0.02307755872607231,0.02307751215994358,False,-0.00020178140697332328
+4,snow_mm,weekly,3,0.009642133489251137,0.009531012736260891,False,-1.1658861032415277
+4,snow_mm,weekly,6,0.009638880379498005,0.009526828303933144,False,-1.1761739793148227
+4,snow_mm,weekly,9,0.009643505327403545,0.009524759836494923,False,-1.2467032549591366
+4,snow_mm,weekly,13,0.00965336337685585,0.009523590095341206,False,-1.3626508513646312
+4,snow_mm,monthly,3,0.005375070031732321,0.004293070640414953,False,-25.203391277362844
+4,snow_mm,monthly,6,0.0016459727194160223,0.000673326721880585,False,-144.45379426184974
+4,runoff_mm,daily,1,0.005664576776325703,0.012482045218348503,True,54.618200164834995
+4,runoff_mm,daily,5,0.004344036802649498,0.01176223624497652,True,63.06793442866978
+4,runoff_mm,daily,10,0.0062460522167384624,0.010903042741119862,True,42.71276041886887
+4,runoff_mm,daily,14,0.007328433450311422,0.010157067328691483,True,27.84892318661501
+4,runoff_mm,weekly,3,0.0073012374341487885,0.0070621850900352,False,-3.3849628842338486
+4,runoff_mm,weekly,6,0.007202621549367905,0.006353410892188549,False,-13.366216534546052
+4,runoff_mm,weekly,9,0.00688136275857687,0.006338729988783598,False,-8.56059132907479
+4,runoff_mm,weekly,13,0.006581714376807213,0.0064202770590782166,False,-2.514491450189448
+4,runoff_mm,monthly,3,0.014925064519047737,0.009071629494428635,False,-64.52462623406308
+4,runoff_mm,monthly,6,0.012454217299818993,0.006235528737306595,False,-99.72993188703559
+4,solar_Wm2,daily,1,6.629605770111084,7.281026840209961,True,8.946829676569248
+4,solar_Wm2,daily,5,7.298154830932617,7.32166862487793,True,0.321153484950359
+4,solar_Wm2,daily,10,7.310039043426514,7.334198474884033,True,0.3294079310814061
+4,solar_Wm2,daily,14,7.321281909942627,7.33941125869751,True,0.24701366520916476
+4,solar_Wm2,weekly,3,3.951737403869629,3.92327618598938,False,-0.7254451772191922
+4,solar_Wm2,weekly,6,3.8475732803344727,3.836881160736084,False,-0.27866694720191565
+4,solar_Wm2,weekly,9,3.886561393737793,3.883671522140503,False,-0.07441081411790648
+4,solar_Wm2,weekly,13,4.186663627624512,4.181654453277588,False,-0.11978929399576832
+4,solar_Wm2,monthly,3,2.1230885982513428,2.154175281524658,True,1.443089777323656
+4,solar_Wm2,monthly,6,3.886484384536743,3.977947473526001,True,2.2992533108584796
+4,wind_speed_ms,daily,1,1.7217248678207397,1.785883903503418,True,3.5925647550109874
+4,wind_speed_ms,daily,5,1.7946487665176392,1.786954641342163,False,-0.4305719349259529
+4,wind_speed_ms,daily,10,1.788209319114685,1.788098692893982,False,-0.006186807313419591
+4,wind_speed_ms,daily,14,1.7838108539581299,1.7805973291397095,False,-0.18047453884326647
+4,wind_speed_ms,weekly,3,0.8789682388305664,0.8778079748153687,False,-0.1321774292881988
+4,wind_speed_ms,weekly,6,0.9184548258781433,0.9132210612297058,False,-0.573110374983022
+4,wind_speed_ms,weekly,9,0.9371227622032166,0.9061100482940674,False,-3.4226211228466985
+4,wind_speed_ms,weekly,13,0.9119011163711548,0.9009010791778564,False,-1.2210038868348059
+4,wind_speed_ms,monthly,3,0.4821363091468811,0.49649250507354736,True,2.8915231911788117
+4,wind_speed_ms,monthly,6,0.3194677233695984,0.31000441312789917,False,-3.052637266100764

datasets/climate_anomaly_models/evap_mm_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc008b63fc5351072fb7c648599de74daa82a5069551804a2acc4af8f1f83588
+size 23175

datasets/climate_anomaly_models/precip_mm_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:611b6ce0d3ed2d4e8179140009f6ee726bf0fd673eb4b8afeea1e7424b418490
+size 23269

datasets/climate_anomaly_models/runoff_mm_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a03ac12621ef7b068a9e24b98718e16e2c91d7da810a2b3a2f6261cd0533c118
+size 23269

datasets/climate_anomaly_models/snow_mm_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3217690f0c4fbf7ea8429cf26f2fd032ff285a589dcaf4ea488819bff0c0341c
+size 23175

datasets/climate_anomaly_models/solar_Wm2_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2c7777971858c6ea7ac68a8b09772e7c454fd425b37a169206b711c136d4b4d
+size 23269

datasets/climate_anomaly_models/temp_C_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:829557c5e6061076457b9600328061a901e47200a90a1dd194add77f581399ed
+size 23160

datasets/climate_anomaly_models/training_summary.json

@@ -0,0 +1,569 @@
+{
+  "wind_u_ms": {
+    "status": "trained",
+    "tau_learned": 24.582462310791016,
+    "best_val_nll": 1.6372566223144531,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 2.7416179180145264,
+        "climatology_only_rmse": 3.0988545417785645,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 2.6203224650530736
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 3.134211540222168,
+        "climatology_only_rmse": 3.1021032333374023,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.026211513320252
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 3.131052255630493,
+        "climatology_only_rmse": 3.0903005599975586,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.1041170837402174
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 3.103543519973755,
+        "climatology_only_rmse": 3.0672664642333984,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.0810500445686095
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 3.0170693397521973,
+        "climatology_only_rmse": 3.003436803817749,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.014521421593338
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 3.0323879718780518,
+        "climatology_only_rmse": 3.0282764434814453,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 2.9841993989261626
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 3.214184284210205,
+        "climatology_only_rmse": 3.213974952697754,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.2457718856016076
+      }
+    ]
+  },
+  "wind_v_ms": {
+    "status": "trained",
+    "tau_learned": 24.902565002441406,
+    "best_val_nll": 1.6035866737365723,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 2.658956527709961,
+        "climatology_only_rmse": 3.0347506999969482,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 2.83391925886813
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 3.018909215927124,
+        "climatology_only_rmse": 3.0155293941497803,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.1571250684467573
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 3.02223539352417,
+        "climatology_only_rmse": 3.0183560848236084,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.1342100607773764
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 2.969254732131958,
+        "climatology_only_rmse": 2.9677913188934326,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.1246906890379798
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 2.8682141304016113,
+        "climatology_only_rmse": 2.8661468029022217,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.1640924899716847
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 2.9601047039031982,
+        "climatology_only_rmse": 2.9590978622436523,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.2391207611099824
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 3.1267435550689697,
+        "climatology_only_rmse": 3.126676082611084,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.475050007851379
+      }
+    ]
+  },
+  "temp_C": {
+    "status": "trained",
+    "tau_learned": 4.595165729522705,
+    "best_val_nll": 1.7487640380859375,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 2.6014695167541504,
+        "climatology_only_rmse": 3.639011859893799,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 2.5666762954564333
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 3.546872615814209,
+        "climatology_only_rmse": 3.653485059738159,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 3.5159557949841624
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 3.6382896900177,
+        "climatology_only_rmse": 3.6745989322662354,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 3.5926695153069423
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 3.6519830226898193,
+        "climatology_only_rmse": 3.652050495147705,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 3.64111023644812
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 3.6176540851593018,
+        "climatology_only_rmse": 3.6176528930664062,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.5972857786739305
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 3.6575050354003906,
+        "climatology_only_rmse": 3.6575050354003906,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.5230417642578518
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 3.563164710998535,
+        "climatology_only_rmse": 3.563164710998535,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 3.6341837602257234
+      }
+    ]
+  },
+  "precip_mm": {
+    "status": "trained",
+    "tau_learned": 29.581716537475586,
+    "best_val_nll": -0.6593826413154602,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 0.35098960995674133,
+        "climatology_only_rmse": 0.3513760566711426,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 0.28751333195416834
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 0.3570253551006317,
+        "climatology_only_rmse": 0.3566482365131378,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.29757815460421255
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 0.3567119240760803,
+        "climatology_only_rmse": 0.3566606640815735,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.3018862256634369
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 0.355029821395874,
+        "climatology_only_rmse": 0.35490572452545166,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.3133934362365005
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 0.3549681603908539,
+        "climatology_only_rmse": 0.3548552393913269,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.33146691327805783
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 0.35744813084602356,
+        "climatology_only_rmse": 0.35721877217292786,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.3439092592300972
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 0.3003491759300232,
+        "climatology_only_rmse": 0.3003755807876587,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 0.3304699597888924
+      }
+    ]
+  },
+  "evap_mm": {
+    "status": "trained",
+    "tau_learned": 4.313884258270264,
+    "best_val_nll": -1.7300641536712646,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 0.10797975957393646,
+        "climatology_only_rmse": 0.12143297493457794,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 0.12467147451686157
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 0.12331552058458328,
+        "climatology_only_rmse": 0.12392762303352356,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 0.12232115721461156
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 0.12414608150720596,
+        "climatology_only_rmse": 0.12524272501468658,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 0.11970659574882868
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 0.12795375287532806,
+        "climatology_only_rmse": 0.1279502958059311,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.11856590466168111
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 0.1313582807779312,
+        "climatology_only_rmse": 0.13135826587677002,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.12346747960827636
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 0.1322607696056366,
+        "climatology_only_rmse": 0.1322607696056366,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.12656735717024395
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 0.11246112734079361,
+        "climatology_only_rmse": 0.11246112734079361,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.11490333415993356
+      }
+    ]
+  },
+  "snow_mm": {
+    "status": "trained",
+    "tau_learned": 0.19776128232479095,
+    "best_val_nll": -3.659635066986084,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 0.05960214138031006,
+        "climatology_only_rmse": 0.059583112597465515,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.0243306499342656
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 0.05958078056573868,
+        "climatology_only_rmse": 0.05958078056573868,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.023817760038479616
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 0.059578243643045425,
+        "climatology_only_rmse": 0.059578243643045425,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.02319194764308019
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 0.05957218259572983,
+        "climatology_only_rmse": 0.05957218259572983,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.020848873251621578
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 0.059570446610450745,
+        "climatology_only_rmse": 0.059570446610450745,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.01777055174592649
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 0.05956999212503433,
+        "climatology_only_rmse": 0.05956999212503433,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.015146745833071601
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 0.05957533419132233,
+        "climatology_only_rmse": 0.05957533419132233,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.010263173068413359
+      }
+    ]
+  },
+  "runoff_mm": {
+    "status": "trained",
+    "tau_learned": 0.19629128277301788,
+    "best_val_nll": -3.575098752975464,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 0.047677528113126755,
+        "climatology_only_rmse": 0.047616854310035706,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.029372147381345286
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 0.04761168733239174,
+        "climatology_only_rmse": 0.04761168733239174,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.02901791240810015
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 0.04760130122303963,
+        "climatology_only_rmse": 0.04760130122303963,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.028581123296834085
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 0.04744777828454971,
+        "climatology_only_rmse": 0.04744777828454971,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.026898740456013728
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 0.047346390783786774,
+        "climatology_only_rmse": 0.047346390783786774,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.024558654810867622
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 0.04725157469511032,
+        "climatology_only_rmse": 0.04725157469511032,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.02224505770190648
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 0.046008218079805374,
+        "climatology_only_rmse": 0.046008218079805374,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 0.015728631979408797
+      }
+    ]
+  },
+  "solar_Wm2": {
+    "status": "trained",
+    "tau_learned": 4.263890266418457,
+    "best_val_nll": 2.344870090484619,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 6.579422950744629,
+        "climatology_only_rmse": 7.028797149658203,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 6.413401710214481
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 7.160378456115723,
+        "climatology_only_rmse": 7.1105523109436035,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 6.894254597848917
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 7.151108741760254,
+        "climatology_only_rmse": 7.158675193786621,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 6.854381368534472
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 7.210677146911621,
+        "climatology_only_rmse": 7.210821628570557,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 6.731941367571172
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 7.321290016174316,
+        "climatology_only_rmse": 7.321290493011475,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 6.793035575972334
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 7.310159206390381,
+        "climatology_only_rmse": 7.310159206390381,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 6.570173736086686
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 6.40765380859375,
+        "climatology_only_rmse": 6.40765380859375,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 7.17354336616637
+      }
+    ]
+  },
+  "wind_speed_ms": {
+    "status": "trained",
+    "tau_learned": 7.753019332885742,
+    "best_val_nll": 1.0960512161254883,
+    "skill_by_horizon": [
+      {
+        "horizon_days": 1,
+        "n_test_examples": 9820,
+        "model_rmse": 1.8207898139953613,
+        "climatology_only_rmse": 1.9430686235427856,
+        "model_beats_climatology": true,
+        "mean_predicted_std": 1.7666910379756078
+      },
+      {
+        "horizon_days": 5,
+        "n_test_examples": 9820,
+        "model_rmse": 1.9703699350357056,
+        "climatology_only_rmse": 1.9440382719039917,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 1.925924609189249
+      },
+      {
+        "horizon_days": 10,
+        "n_test_examples": 9820,
+        "model_rmse": 1.9470051527023315,
+        "climatology_only_rmse": 1.9391731023788452,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 1.927273000370353
+      },
+      {
+        "horizon_days": 30,
+        "n_test_examples": 9820,
+        "model_rmse": 1.9093137979507446,
+        "climatology_only_rmse": 1.9085825681686401,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 1.9374849872622752
+      },
+      {
+        "horizon_days": 60,
+        "n_test_examples": 9820,
+        "model_rmse": 1.8517873287200928,
+        "climatology_only_rmse": 1.85175359249115,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 1.9388930327375131
+      },
+      {
+        "horizon_days": 90,
+        "n_test_examples": 9820,
+        "model_rmse": 1.9170551300048828,
+        "climatology_only_rmse": 1.9170544147491455,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 1.9505142463060157
+      },
+      {
+        "horizon_days": 180,
+        "n_test_examples": 9820,
+        "model_rmse": 1.9633865356445312,
+        "climatology_only_rmse": 1.9633865356445312,
+        "model_beats_climatology": false,
+        "mean_predicted_std": 1.976137801670419
+      }
+    ]
+  }
+}

datasets/climate_anomaly_models/wind_speed_ms_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f183dbbf66fcc2ae226bf46d796ca3cf5b63d0f65fcd43a4c30de59ed5bd9eb
+size 23329

datasets/climate_anomaly_models/wind_u_ms_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cab4ec79a561043e388c55c56dea65266c9e67fa749834fc6c453a1d08fc51c0
+size 23269

datasets/climate_anomaly_models/wind_v_ms_anomaly_model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:61240d69c9974604a295c63098fe8f8951a2292f38b15e77511748af2ed73608
+size 23269

datasets/ecmwf_forecast.csv

@@ -0,0 +1,271 @@
+time,precipitation_sum,temperature_2m_max,temperature_2m_min,wind_speed_10m_max,shortwave_radiation_sum,et0_fao_evapotranspiration,latitude,longitude,station_code
+2026-08-31,2.6,20.9,14.9,28.5,18.07,3.26,48.748705218,0.58156195,H602021010
+2026-09-01,0.0,19.9,11.9,11.3,17.47,2.95,48.748705218,0.58156195,H602021010
+2026-09-02,0.0,22.3,11.9,14.3,17.52,3.28,48.748705218,0.58156195,H602021010
+2026-09-03,0.0,24.7,14.7,14.5,14.75,3.03,48.748705218,0.58156195,H602021010
+2026-09-04,0.0,25.8,16.4,14.5,20.83,3.98,48.748705218,0.58156195,H602021010
+2026-09-05,0.0,27.0,16.9,9.3,17.49,3.34,48.748705218,0.58156195,H602021010
+2026-09-06,0.0,31.8,18.1,11.7,19.35,4.51,48.748705218,0.58156195,H602021010
+2026-09-07,0.0,26.3,15.4,11.7,19.48,3.79,48.748705218,0.58156195,H602021010
+2026-09-08,0.2,23.8,16.9,22.8,14.53,3.36,48.748705218,0.58156195,H602021010
+2026-09-09,8.2,19.0,14.2,20.6,3.13,1.03,48.748705218,0.58156195,H602021010
+2026-08-31,0.9,22.3,15.8,27.3,19.01,3.84,49.034118991,0.808378297,H605641401
+2026-09-01,0.0,20.6,12.2,10.9,16.69,2.94,49.034118991,0.808378297,H605641401
+2026-09-02,0.0,22.4,11.3,14.8,16.92,3.17,49.034118991,0.808378297,H605641401
+2026-09-03,0.0,25.3,14.6,15.7,13.51,2.92,49.034118991,0.808378297,H605641401
+2026-09-04,0.0,26.0,17.1,13.9,19.93,3.81,49.034118991,0.808378297,H605641401
+2026-09-05,0.1,27.7,17.3,10.9,15.21,2.96,49.034118991,0.808378297,H605641401
+2026-09-06,0.0,32.6,17.1,11.0,19.27,4.27,49.034118991,0.808378297,H605641401
+2026-09-07,0.0,26.1,16.1,8.4,18.58,3.57,49.034118991,0.808378297,H605641401
+2026-09-08,0.0,25.8,16.4,20.5,15.34,3.66,49.034118991,0.808378297,H605641401
+2026-09-09,4.6,19.9,15.1,16.8,3.43,1.17,49.034118991,0.808378297,H605641401
+2026-08-31,1.2,21.7,15.6,29.5,19.23,3.63,49.183882705,0.714759598,H620021010
+2026-09-01,0.1,20.2,12.7,13.5,14.54,2.6,49.183882705,0.714759598,H620021010
+2026-09-02,0.0,21.8,11.7,15.5,15.87,2.92,49.183882705,0.714759598,H620021010
+2026-09-03,0.3,24.3,15.4,16.3,11.32,2.38,49.183882705,0.714759598,H620021010
+2026-09-04,0.0,25.1,16.6,15.7,18.54,3.36,49.183882705,0.714759598,H620021010
+2026-09-05,0.0,25.9,17.3,9.5,15.4,2.87,49.183882705,0.714759598,H620021010
+2026-09-06,0.0,30.4,17.3,11.8,19.2,4.08,49.183882705,0.714759598,H620021010
+2026-09-07,0.0,24.5,16.6,9.4,18.1,3.31,49.183882705,0.714759598,H620021010
+2026-09-08,0.4,24.8,16.3,20.8,14.88,2.95,49.183882705,0.714759598,H620021010
+2026-09-09,2.4,18.7,15.0,18.1,3.76,1.12,49.183882705,0.714759598,H620021010
+2026-08-31,1.2,21.9,15.8,29.5,19.23,3.66,49.242018124,0.694854113,H623021010
+2026-09-01,0.1,20.4,12.9,13.5,14.54,2.61,49.242018124,0.694854113,H623021010
+2026-09-02,0.0,22.0,11.9,15.5,15.87,2.93,49.242018124,0.694854113,H623021010
+2026-09-03,0.3,24.5,15.6,16.3,11.32,2.39,49.242018124,0.694854113,H623021010
+2026-09-04,0.0,25.3,16.8,15.7,18.54,3.38,49.242018124,0.694854113,H623021010
+2026-09-05,0.0,26.1,17.5,9.5,15.4,2.88,49.242018124,0.694854113,H623021010
+2026-09-06,0.0,30.6,17.5,11.8,19.2,4.1,49.242018124,0.694854113,H623021010
+2026-09-07,0.0,24.7,16.8,9.4,18.1,3.33,49.242018124,0.694854113,H623021010
+2026-09-08,0.4,25.0,16.5,20.8,14.88,2.97,49.242018124,0.694854113,H623021010
+2026-09-09,2.4,18.9,15.2,18.1,3.76,1.13,49.242018124,0.694854113,H623021010
+2026-08-31,2.6,20.7,14.7,28.5,18.07,3.24,48.718036538,0.430946285,H600021010
+2026-09-01,0.0,19.7,11.7,11.3,17.47,2.94,48.718036538,0.430946285,H600021010
+2026-09-02,0.0,22.1,11.7,14.3,17.52,3.26,48.718036538,0.430946285,H600021010
+2026-09-03,0.0,24.5,14.5,14.5,14.75,3.02,48.718036538,0.430946285,H600021010
+2026-09-04,0.0,25.6,16.2,14.5,20.83,3.96,48.718036538,0.430946285,H600021010
+2026-09-05,0.0,26.8,16.7,9.3,17.49,3.33,48.718036538,0.430946285,H600021010
+2026-09-06,0.0,31.6,17.9,11.7,19.35,4.5,48.718036538,0.430946285,H600021010
+2026-09-07,0.0,26.1,15.2,11.7,19.48,3.77,48.718036538,0.430946285,H600021010
+2026-09-08,0.2,23.6,16.7,22.8,14.53,3.34,48.718036538,0.430946285,H600021010
+2026-09-09,8.2,18.8,14.0,20.6,3.13,1.02,48.718036538,0.430946285,H600021010
+2026-08-31,1.3,21.7,15.2,27.5,18.5,3.58,48.822095344,0.710239413,H603021010
+2026-09-01,0.0,21.0,12.2,9.8,19.29,3.35,48.822095344,0.710239413,H603021010
+2026-09-02,0.0,22.5,11.3,13.8,17.41,3.28,48.822095344,0.710239413,H603021010
+2026-09-03,0.0,25.6,14.5,14.3,16.18,3.37,48.822095344,0.710239413,H603021010
+2026-09-04,0.0,26.9,16.6,13.4,20.87,4.17,48.822095344,0.710239413,H603021010
+2026-09-05,0.0,27.5,16.9,8.3,17.3,3.32,48.822095344,0.710239413,H603021010
+2026-09-06,0.0,32.5,17.5,9.5,19.41,4.37,48.822095344,0.710239413,H603021010
+2026-09-07,0.0,27.0,15.8,10.1,19.59,3.73,48.822095344,0.710239413,H603021010
+2026-09-08,0.0,24.9,16.9,21.7,15.21,3.62,48.822095344,0.710239413,H603021010
+2026-09-09,9.8,19.6,14.8,18.4,3.11,1.09,48.822095344,0.710239413,H603021010
+2026-08-31,0.9,22.1,15.6,27.3,19.01,3.82,48.952759299,0.743674684,H605021010
+2026-09-01,0.0,20.4,12.0,10.9,16.69,2.93,48.952759299,0.743674684,H605021010
+2026-09-02,0.0,22.2,11.1,14.8,16.92,3.16,48.952759299,0.743674684,H605021010
+2026-09-03,0.0,25.1,14.4,15.7,13.51,2.91,48.952759299,0.743674684,H605021010
+2026-09-04,0.0,25.8,16.9,13.9,19.93,3.79,48.952759299,0.743674684,H605021010
+2026-09-05,0.1,27.5,17.1,10.9,15.21,2.95,48.952759299,0.743674684,H605021010
+2026-09-06,0.0,32.4,16.9,11.0,19.27,4.26,48.952759299,0.743674684,H605021010
+2026-09-07,0.0,25.9,15.9,8.4,18.58,3.56,48.952759299,0.743674684,H605021010
+2026-09-08,0.0,25.6,16.2,20.5,15.34,3.64,48.952759299,0.743674684,H605021010
+2026-09-09,4.6,19.7,14.9,16.8,3.43,1.16,48.952759299,0.743674684,H605021010
+2026-08-31,0.9,22.4,15.8,27.3,19.01,3.84,49.030154418,0.814920403,H605022010
+2026-09-01,0.0,20.7,12.2,10.9,16.69,2.94,49.030154418,0.814920403,H605022010
+2026-09-02,0.0,22.5,11.4,14.8,16.92,3.18,49.030154418,0.814920403,H605022010
+2026-09-03,0.0,25.4,14.7,15.7,13.51,2.93,49.030154418,0.814920403,H605022010
+2026-09-04,0.0,26.1,17.2,13.9,19.93,3.81,49.030154418,0.814920403,H605022010
+2026-09-05,0.1,27.8,17.4,10.9,15.21,2.96,49.030154418,0.814920403,H605022010
+2026-09-06,0.0,32.6,17.1,11.0,19.27,4.27,49.030154418,0.814920403,H605022010
+2026-09-07,0.0,26.2,16.1,8.4,18.58,3.57,49.030154418,0.814920403,H605022010
+2026-09-08,0.0,25.9,16.4,20.5,15.34,3.67,49.030154418,0.814920403,H605022010
+2026-09-09,4.6,20.0,15.1,16.8,3.43,1.17,49.030154418,0.814920403,H605022010
+2026-08-31,0.9,22.2,15.7,27.3,19.01,3.83,48.984918738,0.789022568,H605641101
+2026-09-01,0.0,20.5,12.1,10.9,16.69,2.93,48.984918738,0.789022568,H605641101
+2026-09-02,0.0,22.3,11.2,14.8,16.92,3.17,48.984918738,0.789022568,H605641101
+2026-09-03,0.0,25.2,14.5,15.7,13.51,2.92,48.984918738,0.789022568,H605641101
+2026-09-04,0.0,25.9,17.0,13.9,19.93,3.8,48.984918738,0.789022568,H605641101
+2026-09-05,0.1,27.6,17.2,10.9,15.21,2.95,48.984918738,0.789022568,H605641101
+2026-09-06,0.0,32.5,17.0,11.0,19.27,4.26,48.984918738,0.789022568,H605641101
+2026-09-07,0.0,26.0,16.0,8.4,18.58,3.56,48.984918738,0.789022568,H605641101
+2026-09-08,0.0,25.7,16.3,20.5,15.34,3.65,48.984918738,0.789022568,H605641101
+2026-09-09,4.6,19.8,15.0,16.8,3.43,1.17,48.984918738,0.789022568,H605641101
+2026-08-31,0.9,22.4,15.9,27.3,19.01,3.85,49.047068778,0.799839826,H605641201
+2026-09-01,0.0,20.7,12.3,10.9,16.69,2.94,49.047068778,0.799839826,H605641201
+2026-09-02,0.0,22.5,11.4,14.8,16.92,3.18,49.047068778,0.799839826,H605641201
+2026-09-03,0.0,25.4,14.7,15.7,13.51,2.93,49.047068778,0.799839826,H605641201
+2026-09-04,0.0,26.1,17.2,13.9,19.93,3.82,49.047068778,0.799839826,H605641201
+2026-09-05,0.1,27.8,17.4,10.9,15.21,2.96,49.047068778,0.799839826,H605641201
+2026-09-06,0.0,32.7,17.2,11.0,19.27,4.28,49.047068778,0.799839826,H605641201
+2026-09-07,0.0,26.2,16.2,8.4,18.58,3.57,49.047068778,0.799839826,H605641201
+2026-09-08,0.0,25.9,16.5,20.5,15.34,3.67,49.047068778,0.799839826,H605641201
+2026-09-09,4.6,20.0,15.2,16.8,3.43,1.17,49.047068778,0.799839826,H605641201
+2026-08-31,0.9,22.5,15.9,27.3,19.01,3.86,49.091746041,0.759834916,H606000101
+2026-09-01,0.0,20.8,12.3,10.9,16.69,2.95,49.091746041,0.759834916,H606000101
+2026-09-02,0.0,22.6,11.5,14.8,16.92,3.18,49.091746041,0.759834916,H606000101
+2026-09-03,0.0,25.5,14.8,15.7,13.51,2.94,49.091746041,0.759834916,H606000101
+2026-09-04,0.0,26.2,17.3,13.9,19.93,3.82,49.091746041,0.759834916,H606000101
+2026-09-05,0.1,27.9,17.5,10.9,15.21,2.97,49.091746041,0.759834916,H606000101
+2026-09-06,0.0,32.7,17.2,11.0,19.27,4.28,49.091746041,0.759834916,H606000101
+2026-09-07,0.0,26.3,16.2,8.4,18.58,3.58,49.091746041,0.759834916,H606000101
+2026-09-08,0.0,26.0,16.5,20.5,15.34,3.68,49.091746041,0.759834916,H606000101
+2026-09-09,4.6,20.1,15.2,16.8,3.43,1.18,49.091746041,0.759834916,H606000101
+2026-08-31,1.9,21.7,15.8,31.5,19.12,3.63,49.340856041,0.58089832,H623022010
+2026-09-01,0.3,20.4,12.8,15.4,15.34,2.73,49.340856041,0.58089832,H623022010
+2026-09-02,0.0,22.4,12.1,16.3,16.46,3.05,49.340856041,0.58089832,H623022010
+2026-09-03,0.1,24.7,15.9,18.2,11.79,2.5,49.340856041,0.58089832,H623022010
+2026-09-04,0.0,24.6,16.9,17.5,17.64,3.16,49.340856041,0.58089832,H623022010
+2026-09-05,0.2,26.1,17.9,9.7,15.43,2.88,49.340856041,0.58089832,H623022010
+2026-09-06,0.0,30.1,17.8,17.1,19.42,4.16,49.340856041,0.58089832,H623022010
+2026-09-07,0.0,24.8,16.9,10.9,17.97,3.36,49.340856041,0.58089832,H623022010
+2026-09-08,0.6,24.4,17.1,23.6,14.37,2.82,49.340856041,0.58089832,H623022010
+2026-09-09,2.4,18.8,15.1,20.3,3.91,1.15,49.340856041,0.58089832,H623022010
+2026-08-31,2.4,22.3,15.5,27.2,18.91,3.8,48.44823808,1.205047075,H402021010
+2026-09-01,0.0,22.3,12.5,9.3,20.75,3.62,48.44823808,1.205047075,H402021010
+2026-09-02,0.0,23.0,12.2,11.2,19.19,3.58,48.44823808,1.205047075,H402021010
+2026-09-03,0.0,26.0,14.0,13.6,18.44,3.7,48.44823808,1.205047075,H402021010
+2026-09-04,0.0,27.3,17.0,13.9,20.95,4.26,48.44823808,1.205047075,H402021010
+2026-09-05,0.0,27.8,16.5,10.5,18.48,3.57,48.44823808,1.205047075,H402021010
+2026-09-06,0.0,33.0,18.5,10.6,19.2,4.59,48.44823808,1.205047075,H402021010
+2026-09-07,0.2,27.9,16.8,10.4,19.69,3.82,48.44823808,1.205047075,H402021010
+2026-09-08,0.0,25.5,18.0,22.0,17.0,3.75,48.44823808,1.205047075,H402021010
+2026-09-09,7.2,19.7,15.1,18.1,2.88,1.09,48.44823808,1.205047075,H402021010
+2026-08-31,2.5,22.8,15.8,26.9,18.84,3.92,48.386346937,1.477223775,H403301101
+2026-09-01,0.0,22.5,12.7,7.8,20.55,3.53,48.386346937,1.477223775,H403301101
+2026-09-02,0.0,23.4,11.7,13.0,19.33,3.68,48.386346937,1.477223775,H403301101
+2026-09-03,0.0,26.5,14.0,12.8,18.62,3.77,48.386346937,1.477223775,H403301101
+2026-09-04,0.0,28.3,17.1,13.7,20.85,4.35,48.386346937,1.477223775,H403301101
+2026-09-05,0.0,27.5,16.7,10.8,18.06,3.49,48.386346937,1.477223775,H403301101
+2026-09-06,0.0,33.3,18.5,10.8,19.18,4.61,48.386346937,1.477223775,H403301101
+2026-09-07,0.2,28.5,17.4,10.2,19.49,3.9,48.386346937,1.477223775,H403301101
+2026-09-08,0.0,26.5,17.5,21.5,17.51,3.89,48.386346937,1.477223775,H403301101
+2026-09-09,5.4,20.5,15.2,18.0,2.99,1.34,48.386346937,1.477223775,H403301101
+2026-08-31,2.5,22.8,15.8,26.9,18.84,3.92,48.441954003,1.498980903,H404021010
+2026-09-01,0.0,22.5,12.7,7.8,20.55,3.53,48.441954003,1.498980903,H404021010
+2026-09-02,0.0,23.4,11.7,13.0,19.33,3.68,48.441954003,1.498980903,H404021010
+2026-09-03,0.0,26.5,14.0,12.8,18.62,3.77,48.441954003,1.498980903,H404021010
+2026-09-04,0.0,28.3,17.1,13.7,20.85,4.35,48.441954003,1.498980903,H404021010
+2026-09-05,0.0,27.5,16.7,10.8,18.06,3.49,48.441954003,1.498980903,H404021010
+2026-09-06,0.0,33.3,18.5,10.8,19.18,4.61,48.441954003,1.498980903,H404021010
+2026-09-07,0.2,28.5,17.4,10.2,19.49,3.9,48.441954003,1.498980903,H404021010
+2026-09-08,0.0,26.5,17.5,21.5,17.51,3.89,48.441954003,1.498980903,H404021010
+2026-09-09,5.4,20.5,15.2,18.0,2.99,1.34,48.441954003,1.498980903,H404021010
+2026-08-31,1.5,23.3,16.1,26.8,19.0,4.05,48.77678274,1.379648722,H420021010
+2026-09-01,0.0,22.8,13.0,8.3,20.62,3.6,48.77678274,1.379648722,H420021010
+2026-09-02,0.0,24.1,12.0,11.8,17.32,3.31,48.77678274,1.379648722,H420021010
+2026-09-03,0.0,26.8,14.4,13.4,16.56,3.46,48.77678274,1.379648722,H420021010
+2026-09-04,0.0,28.5,17.1,12.3,20.68,4.31,48.77678274,1.379648722,H420021010
+2026-09-05,0.0,27.2,16.8,10.1,16.12,3.08,48.77678274,1.379648722,H420021010
+2026-09-06,0.0,33.2,18.0,11.3,19.16,4.47,48.77678274,1.379648722,H420021010
+2026-09-07,0.0,27.9,17.0,8.1,19.43,3.65,48.77678274,1.379648722,H420021010
+2026-09-08,0.2,27.0,16.7,21.0,17.64,3.93,48.77678274,1.379648722,H420021010
+2026-09-09,7.2,20.8,15.4,16.5,3.23,1.3,48.77678274,1.379648722,H420021010
+2026-08-31,1.5,23.3,16.1,26.8,19.0,4.05,48.776718416,1.379891235,H420021011
+2026-09-01,0.0,22.8,13.0,8.3,20.62,3.6,48.776718416,1.379891235,H420021011
+2026-09-02,0.0,24.1,12.0,11.8,17.32,3.31,48.776718416,1.379891235,H420021011
+2026-09-03,0.0,26.8,14.4,13.4,16.56,3.46,48.776718416,1.379891235,H420021011
+2026-09-04,0.0,28.5,17.1,12.3,20.68,4.31,48.776718416,1.379891235,H420021011
+2026-09-05,0.0,27.2,16.8,10.1,16.12,3.08,48.776718416,1.379891235,H420021011
+2026-09-06,0.0,33.2,18.0,11.3,19.16,4.47,48.776718416,1.379891235,H420021011
+2026-09-07,0.0,27.9,17.0,8.1,19.43,3.65,48.776718416,1.379891235,H420021011
+2026-09-08,0.2,27.0,16.7,21.0,17.64,3.93,48.776718416,1.379891235,H420021011
+2026-09-09,7.2,20.8,15.4,16.5,3.23,1.3,48.776718416,1.379891235,H420021011
+2026-08-31,2.4,22.4,15.6,27.2,18.91,3.81,48.449158817,1.291192655,H403021010
+2026-09-01,0.0,22.4,12.6,9.3,20.75,3.63,48.449158817,1.291192655,H403021010
+2026-09-02,0.0,23.1,12.3,11.2,19.19,3.58,48.449158817,1.291192655,H403021010
+2026-09-03,0.0,26.1,14.1,13.6,18.44,3.71,48.449158817,1.291192655,H403021010
+2026-09-04,0.0,27.4,17.1,13.9,20.95,4.27,48.449158817,1.291192655,H403021010
+2026-09-05,0.0,27.9,16.6,10.5,18.48,3.58,48.449158817,1.291192655,H403021010
+2026-09-06,0.0,33.1,18.6,10.6,19.2,4.6,48.449158817,1.291192655,H403021010
+2026-09-07,0.2,28.0,16.9,10.4,19.69,3.82,48.449158817,1.291192655,H403021010
+2026-09-08,0.0,25.6,18.1,22.0,17.0,3.76,48.449158817,1.291192655,H403021010
+2026-09-09,7.2,19.8,15.2,18.1,2.88,1.09,48.449158817,1.291192655,H403021010
+2026-08-31,2.5,22.9,15.8,26.9,18.84,3.93,48.46726995,1.483793322,H404021101
+2026-09-01,0.0,22.5,12.8,7.8,20.55,3.54,48.46726995,1.483793322,H404021101
+2026-09-02,0.0,23.5,11.8,13.0,19.33,3.69,48.46726995,1.483793322,H404021101
+2026-09-03,0.0,26.5,14.1,12.8,18.62,3.78,48.46726995,1.483793322,H404021101
+2026-09-04,0.0,28.3,17.1,13.7,20.85,4.36,48.46726995,1.483793322,H404021101
+2026-09-05,0.0,27.6,16.7,10.8,18.06,3.49,48.46726995,1.483793322,H404021101
+2026-09-06,0.0,33.3,18.5,10.8,19.18,4.62,48.46726995,1.483793322,H404021101
+2026-09-07,0.2,28.5,17.5,10.2,19.49,3.9,48.46726995,1.483793322,H404021101
+2026-09-08,0.0,26.5,17.6,21.5,17.51,3.9,48.46726995,1.483793322,H404021101
+2026-09-09,5.4,20.5,15.2,18.0,2.99,1.34,48.46726995,1.483793322,H404021101
+2026-08-31,1.5,23.1,15.9,26.8,19.0,4.04,48.645793808,1.542543476,H414000201
+2026-09-01,0.0,22.6,12.8,8.3,20.62,3.59,48.645793808,1.542543476,H414000201
+2026-09-02,0.0,23.9,11.8,11.8,17.32,3.3,48.645793808,1.542543476,H414000201
+2026-09-03,0.0,26.7,14.2,13.4,16.56,3.45,48.645793808,1.542543476,H414000201
+2026-09-04,0.0,28.4,16.9,12.3,20.68,4.3,48.645793808,1.542543476,H414000201
+2026-09-05,0.0,27.0,16.7,10.1,16.12,3.07,48.645793808,1.542543476,H414000201
+2026-09-06,0.0,33.0,17.9,11.3,19.16,4.45,48.645793808,1.542543476,H414000201
+2026-09-07,0.0,27.8,16.8,8.1,19.43,3.64,48.645793808,1.542543476,H414000201
+2026-09-08,0.2,26.9,16.6,21.0,17.64,3.92,48.645793808,1.542543476,H414000201
+2026-09-09,7.2,20.7,15.3,16.5,3.23,1.29,48.645793808,1.542543476,H414000201
+2026-08-31,1.5,23.2,16.0,26.8,19.0,4.05,48.700852684,1.442183941,H416021010
+2026-09-01,0.0,22.7,12.9,8.3,20.62,3.6,48.700852684,1.442183941,H416021010
+2026-09-02,0.0,24.0,11.9,11.8,17.32,3.31,48.700852684,1.442183941,H416021010
+2026-09-03,0.0,26.8,14.3,13.4,16.56,3.46,48.700852684,1.442183941,H416021010
+2026-09-04,0.0,28.5,17.0,12.3,20.68,4.3,48.700852684,1.442183941,H416021010
+2026-09-05,0.0,27.1,16.8,10.1,16.12,3.08,48.700852684,1.442183941,H416021010
+2026-09-06,0.0,33.1,18.0,11.3,19.16,4.46,48.700852684,1.442183941,H416021010
+2026-09-07,0.0,27.9,16.9,8.1,19.43,3.64,48.700852684,1.442183941,H416021010
+2026-09-08,0.2,27.0,16.7,21.0,17.64,3.92,48.700852684,1.442183941,H416021010
+2026-09-09,7.2,20.8,15.4,16.5,3.23,1.29,48.700852684,1.442183941,H416021010
+2026-08-31,1.5,23.3,16.1,26.8,19.0,4.05,48.777083556,1.379365749,H420021012
+2026-09-01,0.0,22.8,13.0,8.3,20.62,3.6,48.777083556,1.379365749,H420021012
+2026-09-02,0.0,24.1,12.0,11.8,17.32,3.31,48.777083556,1.379365749,H420021012
+2026-09-03,0.0,26.8,14.4,13.4,16.56,3.46,48.777083556,1.379365749,H420021012
+2026-09-04,0.0,28.5,17.1,12.3,20.68,4.31,48.777083556,1.379365749,H420021012
+2026-09-05,0.0,27.2,16.8,10.1,16.12,3.08,48.777083556,1.379365749,H420021012
+2026-09-06,0.0,33.2,18.0,11.3,19.16,4.47,48.777083556,1.379365749,H420021012
+2026-09-07,0.0,27.9,17.0,8.1,19.43,3.65,48.777083556,1.379365749,H420021012
+2026-09-08,0.2,27.0,16.7,21.0,17.64,3.93,48.777083556,1.379365749,H420021012
+2026-09-09,7.2,20.8,15.4,16.5,3.23,1.3,48.777083556,1.379365749,H420021012
+2026-08-31,1.5,23.3,16.1,26.8,19.0,4.05,48.792029301,1.376915068,H426021010
+2026-09-01,0.0,22.8,13.0,8.3,20.62,3.6,48.792029301,1.376915068,H426021010
+2026-09-02,0.0,24.1,12.0,11.8,17.32,3.31,48.792029301,1.376915068,H426021010
+2026-09-03,0.0,26.8,14.4,13.4,16.56,3.46,48.792029301,1.376915068,H426021010
+2026-09-04,0.0,28.5,17.1,12.3,20.68,4.31,48.792029301,1.376915068,H426021010
+2026-09-05,0.0,27.2,16.8,10.1,16.12,3.08,48.792029301,1.376915068,H426021010
+2026-09-06,0.0,33.2,18.0,11.3,19.16,4.47,48.792029301,1.376915068,H426021010
+2026-09-07,0.0,27.9,17.0,8.1,19.43,3.65,48.792029301,1.376915068,H426021010
+2026-09-08,0.2,27.0,16.7,21.0,17.64,3.93,48.792029301,1.376915068,H426021010
+2026-09-09,7.2,20.8,15.4,16.5,3.23,1.3,48.792029301,1.376915068,H426021010
+2026-08-31,1.5,23.3,16.1,26.8,19.0,4.05,48.792090349,1.37670564,H426021201
+2026-09-01,0.0,22.8,13.0,8.3,20.62,3.6,48.792090349,1.37670564,H426021201
+2026-09-02,0.0,24.1,12.0,11.8,17.32,3.31,48.792090349,1.37670564,H426021201
+2026-09-03,0.0,26.8,14.4,13.4,16.56,3.46,48.792090349,1.37670564,H426021201
+2026-09-04,0.0,28.5,17.1,12.3,20.68,4.31,48.792090349,1.37670564,H426021201
+2026-09-05,0.0,27.2,16.8,10.1,16.12,3.08,48.792090349,1.37670564,H426021201
+2026-09-06,0.0,33.2,18.0,11.3,19.16,4.47,48.792090349,1.37670564,H426021201
+2026-09-07,0.0,27.9,17.0,8.1,19.43,3.65,48.792090349,1.37670564,H426021201
+2026-09-08,0.2,27.0,16.7,21.0,17.64,3.93,48.792090349,1.37670564,H426021201
+2026-09-09,7.2,20.8,15.4,16.5,3.23,1.3,48.792090349,1.37670564,H426021201
+2026-08-31,0.9,23.3,16.2,26.4,18.92,4.08,49.01307446,1.378224742,H431021010
+2026-09-01,0.0,21.7,13.1,9.4,18.32,3.22,49.01307446,1.378224742,H431021010
+2026-09-02,0.0,23.6,12.1,12.6,16.36,3.15,49.01307446,1.378224742,H431021010
+2026-09-03,0.0,26.5,14.5,14.8,14.01,3.07,49.01307446,1.378224742,H431021010
+2026-09-04,0.0,28.0,17.1,13.1,20.41,4.15,49.01307446,1.378224742,H431021010
+2026-09-05,0.1,26.9,17.7,10.3,14.6,2.82,49.01307446,1.378224742,H431021010
+2026-09-06,0.0,32.7,17.4,10.4,19.23,4.14,49.01307446,1.378224742,H431021010
+2026-09-07,0.0,26.7,17.3,10.6,19.0,3.72,49.01307446,1.378224742,H431021010
+2026-09-08,0.0,27.6,16.8,19.8,16.53,3.91,49.01307446,1.378224742,H431021010
+2026-09-09,9.4,21.3,15.5,16.6,3.06,1.36,49.01307446,1.378224742,H431021010
+2026-08-31,0.9,23.3,16.4,27.2,19.14,4.11,49.114382066,1.208807304,H432021010
+2026-09-01,0.0,21.6,13.2,10.4,17.6,3.17,49.114382066,1.208807304,H432021010
+2026-09-02,0.0,23.7,12.5,13.0,16.43,3.22,49.114382066,1.208807304,H432021010
+2026-09-03,0.0,26.6,14.9,15.5,14.0,3.12,49.114382066,1.208807304,H432021010
+2026-09-04,0.0,27.9,17.6,14.8,20.48,4.16,49.114382066,1.208807304,H432021010
+2026-09-05,0.1,27.5,18.1,10.7,14.69,2.89,49.114382066,1.208807304,H432021010
+2026-09-06,0.0,33.2,17.6,11.4,19.2,4.32,49.114382066,1.208807304,H432021010
+2026-09-07,0.0,26.8,17.2,10.8,18.87,3.75,49.114382066,1.208807304,H432021010
+2026-09-08,0.0,27.5,17.1,18.0,16.16,3.85,49.114382066,1.208807304,H432021010
+2026-09-09,7.6,21.1,15.7,16.1,3.18,1.32,49.114382066,1.208807304,H432021010
+2026-08-31,1.2,22.3,16.4,25.8,19.06,3.76,49.224003868,1.177680812,H438021010
+2026-09-01,0.0,20.6,13.4,10.0,15.12,2.71,49.224003868,1.177680812,H438021010
+2026-09-02,0.0,22.1,12.2,13.3,15.67,2.88,49.224003868,1.177680812,H438021010
+2026-09-03,0.1,24.7,14.5,15.1,10.91,2.29,49.224003868,1.177680812,H438021010
+2026-09-04,0.0,25.6,16.8,12.2,19.56,3.62,49.224003868,1.177680812,H438021010
+2026-09-05,0.0,25.9,18.0,8.9,15.34,2.84,49.224003868,1.177680812,H438021010
+2026-09-06,0.0,30.3,17.1,8.6,18.92,3.77,49.224003868,1.177680812,H438021010
+2026-09-07,0.0,25.7,17.2,8.0,17.86,3.37,49.224003868,1.177680812,H438021010
+2026-09-08,0.2,25.9,16.4,16.1,15.72,3.29,49.224003868,1.177680812,H438021010
+2026-09-09,4.2,19.9,15.9,14.3,3.5,1.12,49.224003868,1.177680812,H438021010

datasets/eure_risle_stations_full.csv

@@ -0,0 +1,28 @@
+code_site,libelle_site,code_station,libelle_station,type_station,coordonnee_x_station,coordonnee_y_station,code_projection,longitude_station,latitude_station,influence_locale_station,commentaire_station,altitude_ref_alti_station,code_systeme_alti_site,code_commune_station,libelle_commune,code_departement,code_region,libelle_region,code_cours_eau,libelle_cours_eau,uri_cours_eau,descriptif_station,date_maj_station,date_ouverture_station,date_fermeture_station,commentaire_influence_locale_station,code_regime_station,qualification_donnees_station,code_finalite_station,type_contexte_loi_stat_station,type_loi_station,code_sandre_reseau_station,date_debut_ref_alti_station,date_activation_ref_alti_station,date_maj_ref_alti_station,libelle_departement,en_service,coordLatLon,already_in_roster
+H4260210,L'Eure à Saint-Georges-Motel,H426021010,L'Eure à Saint-Georges-Motel,STD,580769.0,6855886.0,26,1.3769150684531133,48.79202930059512,,,69.92,3.0,27543,SAINT-GEORGES-MOTEL,27,28,NORMANDIE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-10-20T00:00:00Z,1999-11-26T00:00:00Z,,,1,16,2,1.0,2.0,"POH010,POH047,POH126,POH903,RIC065,RIC315",1999-11-26T00:00:00Z,2013-11-18T00:00:00Z,2013-11-18T00:00:00Z,EURE,True,"48.79202930059512,1.3769150684531133",True
+H4260212,L'Eure à Saint-Georges-Motel [site virtuel],H426021201,L'Eure à Saint-Georges-Motel [site virtuel],FICTIVE,580753.7572121,6855893.1027166,26,1.3767056402577862,48.79209034910749,,,,,27543,SAINT-GEORGES-MOTEL,27,28,NORMANDIE,H4267701,L'Eure,http://id.eaufrance.fr/CEA/H4267701,,2024-02-13T00:00:00Z,2021-09-01T00:00:00Z,,,1,12,2 3 7,,,POH047,,,,EURE,True,"48.79209034910749,1.3767056402577862",True
+H4310210,L'Eure à Pacy-sur-Eure,H431021010,L'Eure à Pacy-sur-Eure - Station partenaire SEBV,STD,581370.0,6880459.0,26,1.37822474192181,49.01307446045349,,Syndicat mixte Eure Blaise Vesgre (SEBV),40.87,3.0,27448,PACY-SUR-EURE,27,28,NORMANDIE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,Station partenaire SEBV,2025-09-25T00:00:00Z,1999-12-22T00:00:00Z,,,1,16,2,1.0,2.0,"POH047,RIC315",1999-11-26T00:00:00Z,2013-11-18T00:00:00Z,2025-06-20T00:00:00Z,EURE,True,"49.01307446045349,1.37822474192181",True
+H4320210,L'Eure à Cailly-sur-Eure,H432021010,L'Eure à Cailly-sur-Eure,STD,569235.0,6891991.0,26,1.2088073037095033,49.11438206622232,1.0,,21.46,3.0,27124,CAILLY-SUR-EURE,27,28,NORMANDIE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-10-20T00:00:00Z,1968-01-01T00:00:00Z,,,1,20,1 2,1.0,2.0,"BSH004,BSH053,BSH104,BSH153,POH010,POH047,POH126,POH903,RIC016,RIC065,RIC265,RIC315",1968-01-01T00:00:00Z,2013-11-18T00:00:00Z,2013-11-18T00:00:00Z,EURE,True,"49.11438206622232,1.2088073037095033",True
+H4380210,L'Eure à Louviers,H438021010,L'Eure à Louviers,STD,567244.3974367,6904232.7688029,26,1.1776808123006943,49.22400386751953,1.0,,10.75,3.0,27375,LOUVIERS,27,28,NORMANDIE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-10-20T00:00:00Z,1971-03-22T00:00:00Z,,,1,20,1 2 7,1.0,2.0,"BSH004,BSH053,BSH104,BSH153,POH010,POH047,POH126,POH903,RIC016,RIC065,RIC265,RIC315",1986-03-01T00:00:00Z,2013-11-18T00:00:00Z,2020-02-04T00:00:00Z,EURE,True,"49.22400386751953,1.1776808123006943",True
+H6030210,La Risle à Rugles [observateur],H603021010,La Risle à Rugles [observateur],STD,531904.0,6860441.0,26,0.7102394129364239,48.82209534358593,2.0,,,,27502,RUGLES,27,28,NORMANDIE,H6031001,La Risle,http://id.eaufrance.fr/CEA/H6031001,,2023-10-05T00:00:00Z,2006-08-01T00:00:00Z,2013-06-01T00:00:00Z,,1,16,,2.0,3.0,"POH010,POH047,POH903",,,,EURE,False,"48.82209534358593,0.7102394129364239",True
+H6050210,La Risle à Bosc-Renoult-en-Ouche,H605021010,La Risle à Bosc-Renoult-en-Ouche,STD,534773.0,6874894.0,26,0.7436746842294717,48.95275929896315,,,138.27,3.0,27088,BOSC-RENOULT-EN-OUCHE,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2023-01-27T00:00:00Z,2001-07-04T00:00:00Z,2022-11-15T00:00:00Z,,1,16,2,1.0,2.0,"POH126,POH010,POH047,POH903",2001-07-04T00:00:00Z,2023-01-27T00:00:00Z,2023-01-27T00:00:00Z,EURE,False,"48.95275929896315,0.7436746842294717",True
+H6050220,La Risle à Grosley-sur-Risle,H605022010,La Risle à Grosley-sur-Risle,STD,540227.0,6883351.0,26,0.8149204033339469,49.03015441769096,1.0,Altitude du 0 de l'échelle non confirmée par un géomètre expert.,10400.0,3.0,27300,GROSLEY-SUR-RISLE,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2025-07-15T00:00:00Z,1970-01-01T00:00:00Z,1988-01-01T00:00:00Z,,1,20,1 2,1.0,2.0,"POH126,POH010,POH047",1970-01-01T00:00:00Z,2025-07-15T00:00:00Z,2025-07-15T00:00:00Z,EURE,False,"49.03015441769096,0.8149204033339469",True
+H6056411,La Risle à Ajou [amont bétoire],H605641101,La Risle à Ajou [amont bétoire],STD,538193.0,6878375.0,26,0.7890225683598991,48.98491873752821,2.0,,,,27007,AJOU,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2023-10-05T00:00:00Z,2012-08-22T00:00:00Z,2020-01-15T00:00:00Z,,1,16,0,,,"POH010,POH047,POH126,POH903",,,,EURE,False,"48.98491873752821,0.7890225683598991",True
+H6056412,La Risle à Grosley-sur-Risle [aval bétoire],H605641201,La Risle à Grosley-sur-Risle [aval bétoire],STD,539177.0,6885262.0,26,0.799839825688955,49.04706877832192,2.0,,,,27300,GROSLEY-SUR-RISLE,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2023-10-05T00:00:00Z,2013-07-11T00:00:00Z,2020-01-15T00:00:00Z,,1,16,0,,,"POH010,POH126,POH047",,,,EURE,False,"49.047068778321915,0.799839825688955",True
+H6056414,La Risle à Barquet,H605641401,La Risle à Barquet,STD,539761.0,6883805.0,26,0.808378297462809,49.03411899138815,0.0,,,,27040,BARQUET,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2025-10-20T00:00:00Z,2020-10-01T00:00:00Z,,,1,16,1 2 7,,,"POH010,POH047,POH126,RIC016,RIC065,RIC165,RIC315",,,,EURE,True,"49.03411899138815,0.808378297462809",True
+H6060001,La Risle à Beaumontel,H606000101,La Risle à Beaumontel,STD,484772.3,2456019.1,5,0.7598349157788882,49.09174604142581,,,,,27050,BEAUMONTEL,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2023-10-05T00:00:00Z,1972-03-01T00:00:00Z,1983-01-01T00:00:00Z,,1,16,7,,,POH047,,,,EURE,False,"49.09174604142581,0.7598349157788882",True
+H6200210,La Risle à Brionne,H620021010,La Risle à Brionne,STD,533400.0,6900649.0,26,0.714759598241397,49.1838827045184,,,5485.0,3.0,27116,BRIONNE,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2025-10-20T00:00:00Z,2002-01-29T00:00:00Z,,,1,16,2,1.0,2.0,"POH010,POH047,POH126,POH903,RIC065,RIC315",2002-01-29T00:00:00Z,2025-07-15T00:00:00Z,2025-07-15T00:00:00Z,EURE,True,"49.1838827045184,0.714759598241397",True
+H6230210,La Risle à Pont-Authou,H623021010,La Risle à Pont-Authou,STD,532138.0,6907155.0,26,0.694854112706974,49.2420181235553,1.0,Station arrêtée du 01/12/1991 au 08/04/1999.,4102.0,3.0,27468,PONT-AUTHOU,27,28,NORMANDIE,H6227501,La Risle,http://id.eaufrance.fr/CEA/H6227501,,2025-10-20T00:00:00Z,1967-05-01T00:00:00Z,,,1,20,1 2,1.0,2.0,"BSH004,BSH053,BSH104,BSH153,POH010,POH047,POH126,POH903,RIC016,RIC065,RIC265,RIC315",1967-05-01T00:00:00Z,2025-07-15T00:00:00Z,2025-07-15T00:00:00Z,EURE,True,"49.2420181235553,0.694854112706974",True
+H6230220,La Risle à Corneville-sur-Risle [Les Ruelles],H623022010,La Risle à Corneville-sur-Risle [Les Ruelles],STD,524179.0,6918393.0,26,0.5808983200584252,49.34085604126936,2.0,,14.64,3.0,27174,CORNEVILLE-SUR-RISLE,27,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2023-10-05T00:00:00Z,2001-07-04T00:00:00Z,2017-09-13T00:00:00Z,,1,16,2,1.0,2.0,"POH010,POH047,POH126,POH903",2001-07-04T00:00:00Z,2013-11-18T00:00:00Z,2023-10-05T00:00:00Z,EURE,False,"49.34085604126936,0.5808983200584252",True
+H4020210,L'Eure à Chuisnes [observateur],H402021010,L'Eure à Chuisnes [observateur] - observateur,STD,567277.0,6817949.0,26,1.205047075362921,48.44823807964378,2.0,,,,28099,CHUISNES,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,observateur,2025-07-11T00:00:00Z,2006-08-01T00:00:00Z,2013-07-01T00:00:00Z,,1,16,,2.0,3.0,"POH010,POH047,POH903",,,,EURE-ET-LOIR,False,"48.44823807964378,1.205047075362921",True
+H4030210,L'Eure à Saint-Luperce,H403021010,L'Eure à Saint-Luperce,STD,573648.0,6817910.0,26,1.2911926548135038,48.4491588174238,1.0,,153.9,3.0,28350,SAINT-LUPERCE,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-10-20T00:00:00Z,1964-12-17T00:00:00Z,,,1,20,1 2,1.0,2.0,"BSH004,BSH051,BSH053,BSH104,BSH151,POH010,POH047,POH126,POH903,RIC016,RIC065,RIC265,RIC315",2006-12-14T00:00:00Z,2013-11-18T00:00:00Z,2025-07-15T00:00:00Z,EURE-ET-LOIR,True,"48.4491588174238,1.2911926548135038",True
+H4033011,L'Eure à Ver-lès-Chartres,H403301101,L'Eure à Ver-lès-Chartres - Loché,STD,587267.0,6810648.0,26,1.4772237751209467,48.38634693666215,1.0,"Station reprise de manière formelle par la DREAL Centre (SEMA) en 2000. Elle a été exploitée par le Service de la Navigation de la Seine (ex. Circonscription Electrique Nord et Paris) jusqu'en décembre 1985. Débits vraisemblablement approximatifs, surtout peut-être en crue : à vérifier.",,,28403,VER-LES-CHARTRES,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,Loché,2020-06-26T00:00:00Z,1969-01-01T00:00:00Z,1986-01-02T00:00:00Z,,1,12,1 2,1.0,2.0,,,,,EURE-ET-LOIR,False,"48.38634693666215,1.4772237751209467",True
+H4040210,L'Eure à Chartres,H404021010,L'Eure à Chartres,DEB,588995.0,6816797.0,26,1.4989809032314128,48.44195400344526,2.0,,12459.0,3.0,28085,CHARTRES,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-07-15T00:00:00Z,2010-05-05T00:00:00Z,2017-12-31T00:00:00Z,,1,16,2,1.0,0.0,"POH126,POH010,POH047,POH903",2010-05-05T00:00:00Z,2025-07-15T00:00:00Z,2025-07-15T00:00:00Z,EURE-ET-LOIR,False,"48.44195400344526,1.4989809032314128",True
+H4040211,L'Eure à Lèves,H404021101,L'Eure à Lèves,STD,587926.0,6819632.0,26,1.4837933224994202,48.46726995012251,0.0,,,,28209,LEVES,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-10-20T00:00:00Z,2017-07-20T00:00:00Z,,,1,16,1 2 7,,,"POH010,POH047,POH126,POH903,RIC065,RIC165,RIC315",,,,EURE-ET-LOIR,True,"48.46726995012251,1.4837933224994202",True
+H4140002,Le bras de l'Eure à Lormaye [SEBV],H414000201,Le bras de l'Eure à Lormaye [SEBV],STD,592634.4397549,6839392.7393497,26,1.5425434758151677,48.64579380760316,,,92.35,3.0,28213,LORMAYE,28,24,CENTRE-VAL DE LOIRE,H41-4301,Bras de l'Eure,http://id.eaufrance.fr/CEA/H41-4301,,2026-04-07T00:00:00Z,2025-07-01T00:00:00Z,,,1,16,2 3,,,POH047,2025-07-01T00:00:00Z,2026-04-07T00:00:00Z,2026-04-07T00:00:00Z,EURE-ET-LOIR,True,"48.64579380760316,1.5425434758151675",True
+H4160210,L'Eure à Charpont,H416021010,L'Eure à Charpont,STD,585363.0,6845654.0,26,1.4421839413328292,48.70085268423738,1.0,Altitude du 0 d'échelle non confirmée par un géomètre expert.,8171.0,3.0,28082,CHARPONT,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,,2025-07-15T00:00:00Z,1984-10-09T00:00:00Z,,,1,20,1 2,1.0,2.0,"POH126,RIC016,POH010,BSH004,BSH051,BSH053,RIC065,RIC315,POH047,POH903,BSH104,BSH151,BSH153",1984-10-09T00:00:00Z,2025-07-15T00:00:00Z,2025-07-15T00:00:00Z,EURE-ET-LOIR,True,"48.70085268423738,1.4421839413328292",True
+H4200210,L'Eure à Montreuil,H420021010,L'Eure à Montreuil - observateur,STD,580935.0,6854187.0,26,1.3796487223244014,48.77678274043829,2.0,,,,28267,MONTREUIL,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,observateur,2024-04-04T00:00:00Z,2006-08-01T00:00:00Z,2011-11-01T00:00:00Z,,1,16,,2.0,3.0,"POH010,POH047,POH903",,,,EURE-ET-LOIR,False,"48.77678274043829,1.3796487223244014",True
+H4200210,L'Eure à Montreuil,H420021011,L'Eure à Montreuil - IJINUS,STD,580952.6708716,6854179.4837869,26,1.3798912354796424,48.77671841583632,,"La première mesure réelle commence en janvier 2021. Cependant, une chronique antérieure a été reconstituée (QMJ) grâce aux deux stations de Saint-Georges-Motel (Eure) et de Muzy (Avre). Cette chronique peut-être considérée fiable. Le 0 de l'échelle n'a pas été levé par un géomètre expert, la précision n'est pas connue.",71.08,3.0,28267,MONTREUIL,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,IJINUS,2025-07-18T00:00:00Z,2000-01-01T00:00:00Z,2024-06-13T00:00:00Z,,1,16,2 7,,,"POH126,POH010,POH047",2021-01-01T00:00:00Z,2025-07-18T00:00:00Z,2025-07-18T00:00:00Z,EURE-ET-LOIR,False,"48.776718415836314,1.3798912354796424",True
+H4200210,L'Eure à Montreuil,H420021012,L'Eure à Montreuil - LNS,STD,580914.8962014,6854220.866527,26,1.3793657492806526,48.77708355621209,2.0,,71.5,3.0,28267,MONTREUIL,28,24,CENTRE-VAL DE LOIRE,H4--0200,L'Eure,http://id.eaufrance.fr/CEA/H4--0200,LNS,2025-10-20T00:00:00Z,2023-08-04T00:00:00Z,,,1,20,2 7,,,"POH010,POH047,POH126,RIC016,RIC065,RIC165,RIC315",2020-01-01T00:00:00Z,2025-07-18T00:00:00Z,2025-07-18T00:00:00Z,EURE-ET-LOIR,True,"48.77708355621209,1.3793657492806526",True
+H6000210,La Risle à Échauffour et à Sainte-Gauburge-Sainte-Colombe [observateur],H600021010,La Risle à Échauffour [observateur],STD,511031.0,6849508.0,26,0.4309462853451878,48.71803653764536,2.0,,,,61150,ECHAUFFOUR,61,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2023-10-05T00:00:00Z,2006-08-01T00:00:00Z,2013-05-01T00:00:00Z,,1,16,,2.0,3.0,"POH010,POH047,POH903",,,,ORNE,False,"48.71803653764536,0.43094628534518786",True
+H6020210,La Risle à Rai,H602021010,La Risle à Rai,STD,522210.514122,6852566.3515797,26,0.5815619502671305,48.7487052183159,1.0,,205.32,3.0,61342,RAI,61,28,NORMANDIE,H6--0200,La Risle,http://id.eaufrance.fr/CEA/H6--0200,,2025-10-20T00:00:00Z,1965-09-01T00:00:00Z,,,1,20,1 2,1.0,2.0,"BSH004,BSH053,BSH104,BSH153,POH010,POH047,POH126,POH903,RIC016,RIC065,RIC115,RIC315",1965-09-01T00:00:00Z,2013-11-18T00:00:00Z,2023-12-04T00:00:00Z,ORNE,True,"48.7487052183159,0.5815619502671305",True

datasets/flood_detection_evaluation.csv

@@ -0,0 +1,11 @@
+horizon_days,tp,fp,fn,tn,precision,recall
+1,38,43,0,1029,0.4691358024691358,1.0
+2,37,53,1,1019,0.4111111111111111,0.9736842105263158
+3,37,69,1,1003,0.3490566037735849,0.9736842105263158
+4,38,70,0,1002,0.35185185185185186,1.0
+5,38,106,0,966,0.2638888888888889,1.0
+10,35,188,3,884,0.15695067264573992,0.9210526315789473
+30,12,837,26,235,0.014134275618374558,0.3157894736842105
+60,0,750,0,360,0.0,
+90,0,644,0,466,0.0,
+180,0,667,0,443,0.0,

datasets/flood_events_discharge.csv

@@ -0,0 +1,126 @@
+start_date,end_date,duration_days,peak_value,peak_date,station_code,threshold,threshold_percentile
+2018-01-02,2018-01-09,8,16351.0,2018-01-07,H404021101,11109.799999999992,95.0
+2018-01-20,2018-01-28,9,17311.0,2018-01-23,H404021101,11109.799999999992,95.0
+2018-02-12,2018-02-18,7,17907.0,2018-02-14,H404021101,11109.799999999992,95.0
+2018-03-11,2018-03-18,8,28880.0,2018-03-13,H404021101,11109.799999999992,95.0
+2018-06-12,2018-06-12,1,13497.0,2018-06-12,H404021101,11109.799999999992,95.0
+2018-06-15,2018-06-15,1,12810.0,2018-06-15,H404021101,11109.799999999992,95.0
+2019-02-12,2019-02-14,3,15910.0,2019-02-13,H404021101,11109.799999999992,95.0
+2019-12-16,2019-12-16,1,11126.0,2019-12-16,H404021101,11109.799999999992,95.0
+2019-12-22,2019-12-25,4,14173.0,2019-12-24,H404021101,11109.799999999992,95.0
+2020-02-03,2020-02-07,5,18330.0,2020-02-05,H404021101,11109.799999999992,95.0
+2020-03-01,2020-03-10,10,24552.0,2020-03-05,H404021101,11109.799999999992,95.0
+2020-12-27,2020-12-27,1,13241.0,2020-12-27,H404021101,11109.799999999992,95.0
+2020-12-29,2021-01-01,4,15910.0,2020-12-31,H404021101,11109.799999999992,95.0
+2021-01-31,2021-02-07,8,16887.0,2021-02-03,H404021101,11109.799999999992,95.0
+2022-01-07,2022-01-07,1,11849.0,2022-01-07,H404021101,11109.799999999992,95.0
+2022-01-11,2022-01-12,2,13248.0,2022-01-12,H404021101,11109.799999999992,95.0
+2022-04-11,2022-04-12,2,12363.0,2022-04-11,H404021101,11109.799999999992,95.0
+2023-12-12,2023-12-14,3,13508.0,2023-12-13,H404021101,11109.799999999992,95.0
+2024-01-05,2024-01-07,3,17816.0,2024-01-06,H404021101,11109.799999999992,95.0
+2024-01-19,2024-01-22,4,20477.0,2024-01-20,H404021101,11109.799999999992,95.0
+2024-02-13,2024-02-14,2,11946.0,2024-02-13,H404021101,11109.799999999992,95.0
+2024-02-24,2024-03-06,12,23276.0,2024-02-29,H404021101,11109.799999999992,95.0
+2024-04-07,2024-04-07,1,11427.0,2024-04-07,H404021101,11109.799999999992,95.0
+2024-05-04,2024-05-05,2,13017.0,2024-05-05,H404021101,11109.799999999992,95.0
+2024-10-09,2024-10-14,6,28066.0,2024-10-12,H404021101,11109.799999999992,95.0
+2024-11-26,2024-11-28,3,12643.0,2024-11-27,H404021101,11109.799999999992,95.0
+2024-12-11,2024-12-11,1,12249.0,2024-12-11,H404021101,11109.799999999992,95.0
+2025-01-05,2025-01-13,9,26842.0,2025-01-11,H404021101,11109.799999999992,95.0
+2025-01-26,2025-02-04,10,31386.0,2025-01-28,H404021101,11109.799999999992,95.0
+2025-02-10,2025-02-11,2,11936.0,2025-02-10,H404021101,11109.799999999992,95.0
+2025-02-27,2025-03-01,3,13515.0,2025-02-28,H404021101,11109.799999999992,95.0
+2026-02-07,2026-02-09,3,12080.0,2026-02-08,H404021101,11109.799999999992,95.0
+2026-02-13,2026-02-24,12,28903.0,2026-02-21,H404021101,11109.799999999992,95.0
+2021-02-01,2021-02-08,8,33221.0,2021-02-04,H420021011,24332.75,95.0
+2022-01-09,2022-01-12,4,28190.0,2022-01-10,H420021011,24332.75,95.0
+2022-04-09,2022-04-10,2,24535.0,2022-04-10,H420021011,24332.75,95.0
+2023-12-10,2023-12-11,2,27126.0,2023-12-11,H420021011,24332.75,95.0
+2023-12-13,2023-12-14,2,24956.0,2023-12-14,H420021011,24332.75,95.0
+2024-01-18,2024-01-22,5,36664.0,2024-01-19,H420021011,24332.75,95.0
+2024-02-23,2024-03-07,14,39148.0,2024-02-27,H420021011,24332.75,95.0
+2024-01-19,2024-01-20,2,34978.0,2024-01-19,H420021012,28601.199999999957,95.0
+2024-02-24,2024-02-24,1,29098.0,2024-02-24,H420021012,28601.199999999957,95.0
+2024-02-26,2024-03-06,10,38152.0,2024-02-27,H420021012,28601.199999999957,95.0
+2024-10-10,2024-10-15,6,45610.0,2024-10-14,H420021012,28601.199999999957,95.0
+2024-10-20,2024-10-20,1,29595.0,2024-10-20,H420021012,28601.199999999957,95.0
+2025-01-06,2025-01-13,8,44782.0,2025-01-10,H420021012,28601.199999999957,95.0
+2025-01-26,2025-02-04,8,47715.0,2025-01-31,H420021012,28601.199999999957,95.0
+2026-02-13,2026-02-24,12,46714.0,2026-02-20,H420021012,28601.199999999957,95.0
+2013-01-28,2013-01-28,1,10879.0,2013-01-28,H605641101,8347.999999999998,95.0
+2013-01-30,2013-01-31,2,9878.0,2013-01-30,H605641101,8347.999999999998,95.0
+2013-02-02,2013-02-02,1,8542.0,2013-02-02,H605641101,8347.999999999998,95.0
+2013-02-06,2013-02-07,2,9347.0,2013-02-07,H605641101,8347.999999999998,95.0
+2013-12-25,2013-12-25,1,11450.0,2013-12-25,H605641101,8347.999999999998,95.0
+2014-01-03,2014-01-03,1,10032.0,2014-01-03,H605641101,8347.999999999998,95.0
+2014-01-28,2014-02-01,5,12797.0,2014-01-30,H605641101,8347.999999999998,95.0
+2014-02-06,2014-02-10,4,12976.0,2014-02-08,H605641101,8347.999999999998,95.0
+2014-02-13,2014-02-16,3,13992.0,2014-02-15,H605641101,8347.999999999998,95.0
+2014-03-01,2014-03-02,2,11683.0,2014-03-01,H605641101,8347.999999999998,95.0
+2014-05-29,2014-05-29,1,9119.0,2014-05-29,H605641101,8347.999999999998,95.0
+2014-10-17,2014-10-18,2,10420.0,2014-10-18,H605641101,8347.999999999998,95.0
+2014-12-13,2014-12-13,1,9864.0,2014-12-13,H605641101,8347.999999999998,95.0
+2014-12-18,2014-12-20,2,12424.0,2014-12-18,H605641101,8347.999999999998,95.0
+2015-01-16,2015-01-16,1,11899.0,2015-01-16,H605641101,8347.999999999998,95.0
+2015-01-20,2015-01-20,1,8937.0,2015-01-20,H605641101,8347.999999999998,95.0
+2015-01-30,2015-02-02,4,13134.0,2015-01-31,H605641101,8347.999999999998,95.0
+2015-02-21,2015-02-22,2,10055.0,2015-02-21,H605641101,8347.999999999998,95.0
+2015-02-24,2015-02-28,5,10719.0,2015-02-27,H605641101,8347.999999999998,95.0
+2015-03-03,2015-03-03,1,8353.0,2015-03-03,H605641101,8347.999999999998,95.0
+2013-11-22,2013-11-22,1,7688.0,2013-11-22,H605641201,6639.9000000000015,95.0
+2013-12-25,2013-12-28,4,12007.0,2013-12-26,H605641201,6639.9000000000015,95.0
+2014-01-03,2014-01-03,1,9504.0,2014-01-03,H605641201,6639.9000000000015,95.0
+2014-01-28,2014-02-03,7,13308.0,2014-02-02,H605641201,6639.9000000000015,95.0
+2014-02-06,2014-02-17,12,12983.0,2014-02-15,H605641201,6639.9000000000015,95.0
+2014-02-23,2014-02-23,1,6721.0,2014-02-23,H605641201,6639.9000000000015,95.0
+2014-02-28,2014-03-04,5,10271.0,2014-03-01,H605641201,6639.9000000000015,95.0
+2014-05-29,2014-05-29,1,8390.0,2014-05-29,H605641201,6639.9000000000015,95.0
+2014-10-17,2014-10-18,2,10182.0,2014-10-18,H605641201,6639.9000000000015,95.0
+2014-12-13,2014-12-15,3,12421.0,2014-12-14,H605641201,6639.9000000000015,95.0
+2014-12-18,2014-12-21,4,11709.0,2014-12-19,H605641201,6639.9000000000015,95.0
+2015-01-16,2015-01-17,2,10407.0,2015-01-16,H605641201,6639.9000000000015,95.0
+2015-01-20,2015-01-20,1,8459.0,2015-01-20,H605641201,6639.9000000000015,95.0
+2015-01-30,2015-02-03,5,11412.0,2015-01-31,H605641201,6639.9000000000015,95.0
+2015-02-21,2015-02-22,2,9377.0,2015-02-22,H605641201,6639.9000000000015,95.0
+2015-02-24,2015-02-28,5,9140.0,2015-02-27,H605641201,6639.9000000000015,95.0
+2015-03-02,2015-03-03,2,7914.0,2015-03-03,H605641201,6639.9000000000015,95.0
+2015-03-31,2015-03-31,1,7436.0,2015-03-31,H605641201,6639.9000000000015,95.0
+2016-02-09,2016-02-11,3,10817.0,2016-02-11,H605641201,6639.9000000000015,95.0
+2016-02-14,2016-02-16,3,10704.0,2016-02-15,H605641201,6639.9000000000015,95.0
+2016-03-10,2016-03-10,1,8093.0,2016-03-10,H605641201,6639.9000000000015,95.0
+2016-03-29,2016-03-30,2,8883.0,2016-03-29,H605641201,6639.9000000000015,95.0
+2016-04-01,2016-04-01,1,7758.0,2016-04-01,H605641201,6639.9000000000015,95.0
+2016-04-12,2016-04-12,1,8064.0,2016-04-12,H605641201,6639.9000000000015,95.0
+2017-03-07,2017-03-08,2,9700.0,2017-03-07,H605641201,6639.9000000000015,95.0
+2020-12-24,2020-12-25,2,9744.0,2020-12-24,H605641401,6781.299999999995,95.0
+2020-12-28,2020-12-30,3,11347.0,2020-12-29,H605641401,6781.299999999995,95.0
+2021-01-22,2021-01-22,1,7132.0,2021-01-22,H605641401,6781.299999999995,95.0
+2021-01-29,2021-02-07,10,12671.0,2021-02-05,H605641401,6781.299999999995,95.0
+2021-07-14,2021-07-14,1,7331.0,2021-07-14,H605641401,6781.299999999995,95.0
+2021-12-05,2021-12-07,3,8873.0,2021-12-06,H605641401,6781.299999999995,95.0
+2021-12-11,2021-12-11,1,10480.0,2021-12-11,H605641401,6781.299999999995,95.0
+2021-12-29,2021-12-30,2,9686.0,2021-12-30,H605641401,6781.299999999995,95.0
+2022-01-04,2022-01-06,3,11876.0,2022-01-05,H605641401,6781.299999999995,95.0
+2022-01-09,2022-01-10,2,10720.0,2022-01-10,H605641401,6781.299999999995,95.0
+2022-04-09,2022-04-10,2,12211.0,2022-04-10,H605641401,6781.299999999995,95.0
+2023-01-16,2023-01-18,3,10377.0,2023-01-17,H605641401,6781.299999999995,95.0
+2023-11-13,2023-11-14,2,9373.0,2023-11-13,H605641401,6781.299999999995,95.0
+2023-11-17,2023-11-17,1,8777.0,2023-11-17,H605641401,6781.299999999995,95.0
+2023-11-19,2023-11-20,2,9851.0,2023-11-19,H605641401,6781.299999999995,95.0
+2023-12-10,2023-12-12,3,11931.0,2023-12-11,H605641401,6781.299999999995,95.0
+2024-01-03,2024-01-06,4,13520.0,2024-01-04,H605641401,6781.299999999995,95.0
+2024-01-18,2024-01-20,3,16904.0,2024-01-19,H605641401,6781.299999999995,95.0
+2024-01-23,2024-01-24,2,10793.0,2024-01-23,H605641401,6781.299999999995,95.0
+2024-02-10,2024-02-12,3,8111.0,2024-02-11,H605641401,6781.299999999995,95.0
+2024-02-23,2024-02-24,2,11420.0,2024-02-23,H605641401,6781.299999999995,95.0
+2024-02-26,2024-02-27,2,11611.0,2024-02-27,H605641401,6781.299999999995,95.0
+2024-03-02,2024-03-03,2,9772.0,2024-03-02,H605641401,6781.299999999995,95.0
+2024-04-05,2024-04-06,2,10961.0,2024-04-05,H605641401,6781.299999999995,95.0
+2024-05-02,2024-05-03,2,11787.0,2024-05-03,H605641401,6781.299999999995,95.0
+2024-05-22,2024-05-22,1,10274.0,2024-05-22,H605641401,6781.299999999995,95.0
+2024-11-24,2024-11-27,4,8873.0,2024-11-26,H605641401,6781.299999999995,95.0
+2024-12-08,2024-12-10,3,10847.0,2024-12-09,H605641401,6781.299999999995,95.0
+2024-12-23,2024-12-23,1,8832.0,2024-12-23,H605641401,6781.299999999995,95.0
+2025-01-03,2025-01-11,9,16184.0,2025-01-10,H605641401,6781.299999999995,95.0
+2025-01-24,2025-02-02,8,17946.0,2025-01-28,H605641401,6781.299999999995,95.0
+2026-02-12,2026-02-22,11,15240.0,2026-02-20,H605641401,6781.299999999995,95.0

datasets/flood_events_waterlevel.csv

@@ -0,0 +1,343 @@
+start_date,end_date,duration_days,peak_value,peak_date,station_code,threshold,threshold_percentile
+2015-05-04,2015-05-04,1,660.0,2015-05-04,H404021010,646.0999999999999,95.0
+2015-08-13,2015-08-13,1,650.0,2015-08-13,H404021010,646.0999999999999,95.0
+2015-08-27,2015-08-27,1,660.0,2015-08-27,H404021010,646.0999999999999,95.0
+2015-09-12,2015-09-13,2,900.0,2015-09-13,H404021010,646.0999999999999,95.0
+2015-09-15,2015-09-16,2,670.0,2015-09-15,H404021010,646.0999999999999,95.0
+2015-10-05,2015-10-06,2,880.0,2015-10-06,H404021010,646.0999999999999,95.0
+2015-10-28,2015-10-28,1,680.0,2015-10-28,H404021010,646.0999999999999,95.0
+2015-11-20,2015-11-23,4,740.0,2015-11-20,H404021010,646.0999999999999,95.0
+2016-01-09,2016-01-11,3,680.0,2016-01-11,H404021010,646.0999999999999,95.0
+2016-01-30,2016-02-01,3,710.0,2016-02-01,H404021010,646.0999999999999,95.0
+2016-03-02,2016-03-03,2,660.0,2016-03-02,H404021010,646.0999999999999,95.0
+2016-03-06,2016-03-06,1,675.0,2016-03-06,H404021010,646.0999999999999,95.0
+2016-03-30,2016-03-31,2,658.0,2016-03-31,H404021010,646.0999999999999,95.0
+2016-04-13,2016-04-13,1,768.0,2016-04-13,H404021010,646.0999999999999,95.0
+2016-05-23,2016-05-24,2,709.0,2016-05-23,H404021010,646.0999999999999,95.0
+2016-05-29,2016-05-31,3,851.0,2016-05-29,H404021010,646.0999999999999,95.0
+2016-06-12,2016-06-12,1,675.0,2016-06-12,H404021010,646.0999999999999,95.0
+2016-06-15,2016-06-18,4,732.0,2016-06-18,H404021010,646.0999999999999,95.0
+2016-07-12,2016-07-12,1,953.0,2016-07-12,H404021010,646.0999999999999,95.0
+2017-03-03,2017-03-03,1,664.0,2017-03-03,H404021010,646.0999999999999,95.0
+2017-05-18,2017-05-18,1,735.0,2017-05-18,H404021010,646.0999999999999,95.0
+2017-06-02,2017-06-02,1,819.0,2017-06-02,H404021010,646.0999999999999,95.0
+2017-06-09,2017-06-09,1,741.0,2017-06-09,H404021010,646.0999999999999,95.0
+2017-07-10,2017-07-10,1,757.0,2017-07-10,H404021010,646.0999999999999,95.0
+2017-09-08,2017-09-08,1,776.0,2017-09-08,H404021010,646.0999999999999,95.0
+2017-09-13,2017-09-14,2,721.0,2017-09-13,H404021010,646.0999999999999,95.0
+2017-12-11,2017-12-11,1,699.0,2017-12-11,H404021010,646.0999999999999,95.0
+2018-01-01,2018-01-01,1,977.0,2018-01-01,H404021101,864.0,95.0
+2018-01-03,2018-01-09,7,1080.0,2018-01-07,H404021101,864.0,95.0
+2018-01-18,2018-01-18,1,908.0,2018-01-18,H404021101,864.0,95.0
+2018-01-20,2018-01-26,7,1106.0,2018-01-23,H404021101,864.0,95.0
+2018-02-12,2018-02-15,4,1159.0,2018-02-14,H404021101,864.0,95.0
+2018-02-17,2018-02-18,2,874.0,2018-02-17,H404021101,864.0,95.0
+2018-03-11,2018-03-18,8,1632.0,2018-03-13,H404021101,864.0,95.0
+2018-06-11,2018-06-12,2,1808.0,2018-06-11,H404021101,864.0,95.0
+2018-06-15,2018-06-16,2,920.0,2018-06-15,H404021101,864.0,95.0
+2019-02-12,2019-02-14,3,1058.0,2019-02-13,H404021101,864.0,95.0
+2019-05-19,2019-05-19,1,1043.0,2019-05-19,H404021101,864.0,95.0
+2019-12-22,2019-12-25,4,977.0,2019-12-24,H404021101,864.0,95.0
+2020-02-03,2020-02-07,5,1157.0,2020-02-05,H404021101,864.0,95.0
+2020-03-01,2020-03-10,10,1486.0,2020-03-05,H404021101,864.0,95.0
+2020-12-23,2020-12-23,1,885.0,2020-12-23,H404021101,864.0,95.0
+2020-12-27,2020-12-28,2,1127.0,2020-12-27,H404021101,864.0,95.0
+2020-12-30,2021-01-01,3,1042.0,2020-12-31,H404021101,864.0,95.0
+2021-01-31,2021-02-07,8,1200.0,2021-02-03,H404021101,864.0,95.0
+2021-06-19,2021-06-19,1,1212.0,2021-06-19,H404021101,864.0,95.0
+2022-01-07,2022-01-07,1,867.0,2022-01-07,H404021101,864.0,95.0
+2022-01-11,2022-01-12,2,929.0,2022-01-12,H404021101,864.0,95.0
+2022-04-08,2022-04-08,1,936.0,2022-04-08,H404021101,864.0,95.0
+2022-04-11,2022-04-12,2,922.0,2022-04-12,H404021101,864.0,95.0
+2022-06-04,2022-06-04,1,919.0,2022-06-04,H404021101,864.0,95.0
+2022-06-22,2022-06-22,1,949.0,2022-06-22,H404021101,864.0,95.0
+2023-06-18,2023-06-18,1,879.0,2023-06-18,H404021101,864.0,95.0
+2023-09-17,2023-09-17,1,901.0,2023-09-17,H404021101,864.0,95.0
+2023-12-09,2023-12-09,1,1013.0,2023-12-09,H404021101,864.0,95.0
+2023-12-12,2023-12-14,3,930.0,2023-12-13,H404021101,864.0,95.0
+2024-01-05,2024-01-07,3,1134.0,2024-01-06,H404021101,864.0,95.0
+2024-01-18,2024-01-22,5,1340.0,2024-01-20,H404021101,864.0,95.0
+2024-02-13,2024-02-14,2,873.0,2024-02-13,H404021101,864.0,95.0
+2024-02-24,2024-03-06,12,1377.0,2024-02-29,H404021101,864.0,95.0
+2024-04-07,2024-04-07,1,904.0,2024-04-07,H404021101,864.0,95.0
+2024-05-05,2024-05-06,2,943.0,2024-05-05,H404021101,864.0,95.0
+2024-05-12,2024-05-12,1,864.0,2024-05-12,H404021101,864.0,95.0
+2024-06-21,2024-06-21,1,989.0,2024-06-21,H404021101,864.0,95.0
+2024-09-26,2024-09-26,1,901.0,2024-09-26,H404021101,864.0,95.0
+2024-10-09,2024-10-14,6,1621.0,2024-10-12,H404021101,864.0,95.0
+2024-11-25,2024-11-25,1,879.0,2024-11-25,H404021101,864.0,95.0
+2024-11-27,2024-11-28,2,947.0,2024-11-27,H404021101,864.0,95.0
+2024-12-11,2024-12-11,1,882.0,2024-12-11,H404021101,864.0,95.0
+2025-01-04,2025-01-13,10,1553.0,2025-01-09,H404021101,864.0,95.0
+2025-01-26,2025-02-04,10,1751.0,2025-01-28,H404021101,864.0,95.0
+2025-02-10,2025-02-10,1,924.0,2025-02-10,H404021101,864.0,95.0
+2025-02-25,2025-03-01,5,920.0,2025-02-28,H404021101,864.0,95.0
+2025-06-13,2025-06-13,1,1033.0,2025-06-13,H404021101,864.0,95.0
+2025-08-20,2025-08-20,1,1513.0,2025-08-20,H404021101,864.0,95.0
+2026-02-05,2026-02-05,1,870.0,2026-02-05,H404021101,864.0,95.0
+2026-02-08,2026-02-08,1,867.0,2026-02-08,H404021101,864.0,95.0
+2026-02-11,2026-02-11,1,865.0,2026-02-11,H404021101,864.0,95.0
+2026-02-13,2026-02-24,12,1631.0,2026-02-21,H404021101,864.0,95.0
+2021-01-22,2021-01-25,4,509.0,2021-01-22,H420021011,449.8999999999992,95.0
+2021-01-30,2021-02-09,11,809.0,2021-02-04,H420021011,449.8999999999992,95.0
+2022-01-04,2022-01-05,2,560.0,2022-01-04,H420021011,449.8999999999992,95.0
+2022-01-08,2022-01-13,6,685.0,2022-01-09,H420021011,449.8999999999992,95.0
+2022-04-09,2022-04-12,4,581.0,2022-04-09,H420021011,449.8999999999992,95.0
+2023-12-10,2023-12-15,6,724.0,2023-12-10,H420021011,449.8999999999992,95.0
+2024-01-04,2024-01-08,5,523.0,2024-01-07,H420021011,449.8999999999992,95.0
+2024-01-18,2024-01-23,6,905.0,2024-01-19,H420021011,449.8999999999992,95.0
+2024-02-23,2024-03-07,14,956.0,2024-02-27,H420021011,449.8999999999992,95.0
+2024-04-04,2024-04-06,3,500.0,2024-04-05,H420021011,449.8999999999992,95.0
+2024-04-08,2024-04-08,1,471.0,2024-04-08,H420021011,449.8999999999992,95.0
+2024-01-18,2024-01-20,3,1277.0,2024-01-19,H420021012,1103.2499999999995,95.0
+2024-02-26,2024-03-06,10,1341.0,2024-02-27,H420021012,1103.2499999999995,95.0
+2024-10-09,2024-10-16,8,1573.0,2024-10-14,H420021012,1103.2499999999995,95.0
+2024-10-20,2024-10-21,2,1146.0,2024-10-20,H420021012,1103.2499999999995,95.0
+2025-01-05,2025-01-14,10,1505.0,2025-01-09,H420021012,1103.2499999999995,95.0
+2025-01-26,2025-02-04,10,1583.0,2025-02-01,H420021012,1103.2499999999995,95.0
+2026-02-13,2026-02-24,12,1528.0,2026-02-20,H420021012,1103.2499999999995,95.0
+2026-06-26,2026-06-26,1,2875.0,2026-06-26,H420021012,1103.2499999999995,95.0
+2013-01-01,2013-01-01,1,882.0,2013-01-01,H605021010,867.8499999999999,95.0
+2013-01-28,2013-01-28,1,943.0,2013-01-28,H605021010,867.8499999999999,95.0
+2013-02-11,2013-02-12,2,959.0,2013-02-12,H605021010,867.8499999999999,95.0
+2013-03-19,2013-03-21,3,950.0,2013-03-19,H605021010,867.8499999999999,95.0
+2013-04-10,2013-04-11,2,906.0,2013-04-10,H605021010,867.8499999999999,95.0
+2013-12-25,2013-12-26,2,1001.0,2013-12-26,H605021010,867.8499999999999,95.0
+2013-12-28,2013-12-28,1,873.0,2013-12-28,H605021010,867.8499999999999,95.0
+2014-01-30,2014-02-03,5,957.0,2014-02-02,H605021010,867.8499999999999,95.0
+2014-02-07,2014-02-10,4,954.0,2014-02-09,H605021010,867.8499999999999,95.0
+2014-02-13,2014-02-15,3,952.0,2014-02-14,H605021010,867.8499999999999,95.0
+2014-03-01,2014-03-02,2,941.0,2014-03-01,H605021010,867.8499999999999,95.0
+2014-05-29,2014-05-29,1,905.0,2014-05-29,H605021010,867.8499999999999,95.0
+2014-08-08,2014-08-08,1,869.0,2014-08-08,H605021010,867.8499999999999,95.0
+2014-10-17,2014-10-18,2,1053.0,2014-10-18,H605021010,867.8499999999999,95.0
+2014-12-13,2014-12-15,3,1123.0,2014-12-14,H605021010,867.8499999999999,95.0
+2014-12-18,2014-12-20,3,1097.0,2014-12-19,H605021010,867.8499999999999,95.0
+2015-01-16,2015-01-16,1,1052.0,2015-01-16,H605021010,867.8499999999999,95.0
+2015-01-30,2015-02-02,4,1104.0,2015-01-31,H605021010,867.8499999999999,95.0
+2015-02-21,2015-02-22,2,1059.0,2015-02-21,H605021010,867.8499999999999,95.0
+2015-02-27,2015-02-28,2,974.0,2015-02-27,H605021010,867.8499999999999,95.0
+2015-03-30,2015-03-31,2,926.0,2015-03-30,H605021010,867.8499999999999,95.0
+2016-02-09,2016-02-11,3,1085.0,2016-02-11,H605021010,867.8499999999999,95.0
+2016-02-13,2016-02-15,3,1057.0,2016-02-14,H605021010,867.8499999999999,95.0
+2016-03-10,2016-03-10,1,887.0,2016-03-10,H605021010,867.8499999999999,95.0
+2016-03-29,2016-03-29,1,935.0,2016-03-29,H605021010,867.8499999999999,95.0
+2016-04-11,2016-04-11,1,876.0,2016-04-11,H605021010,867.8499999999999,95.0
+2016-06-01,2016-06-01,1,869.0,2016-06-01,H605021010,867.8499999999999,95.0
+2017-03-07,2017-03-08,2,989.0,2017-03-07,H605021010,867.8499999999999,95.0
+2017-12-16,2017-12-16,1,871.0,2017-12-16,H605021010,867.8499999999999,95.0
+2017-12-27,2017-12-28,2,993.0,2017-12-28,H605021010,867.8499999999999,95.0
+2017-12-30,2018-01-06,8,1079.0,2018-01-05,H605021010,867.8499999999999,95.0
+2018-01-16,2018-01-25,10,1096.0,2018-01-21,H605021010,867.8499999999999,95.0
+2018-02-01,2018-02-01,1,1028.0,2018-02-01,H605021010,867.8499999999999,95.0
+2018-02-11,2018-02-12,2,1009.0,2018-02-12,H605021010,867.8499999999999,95.0
+2018-02-15,2018-02-16,2,992.0,2018-02-15,H605021010,867.8499999999999,95.0
+2018-03-09,2018-03-09,1,940.0,2018-03-09,H605021010,867.8499999999999,95.0
+2018-03-11,2018-03-12,2,1063.0,2018-03-11,H605021010,867.8499999999999,95.0
+2018-03-31,2018-03-31,1,1035.0,2018-03-31,H605021010,867.8499999999999,95.0
+2018-06-04,2018-06-06,3,1085.0,2018-06-05,H605021010,867.8499999999999,95.0
+2018-06-11,2018-06-14,4,1315.0,2018-06-12,H605021010,867.8499999999999,95.0
+2019-02-09,2019-02-12,4,942.0,2019-02-10,H605021010,867.8499999999999,95.0
+2019-11-28,2019-11-28,1,1015.0,2019-11-28,H605021010,867.8499999999999,95.0
+2019-12-13,2019-12-14,2,919.0,2019-12-13,H605021010,867.8499999999999,95.0
+2019-12-20,2019-12-23,4,1026.0,2019-12-21,H605021010,867.8499999999999,95.0
+2020-01-28,2020-01-29,2,954.0,2020-01-28,H605021010,867.8499999999999,95.0
+2020-02-01,2020-02-05,5,1071.0,2020-02-02,H605021010,867.8499999999999,95.0
+2020-02-10,2020-02-14,5,1042.0,2020-02-11,H605021010,867.8499999999999,95.0
+2020-02-17,2020-02-20,4,987.0,2020-02-17,H605021010,867.8499999999999,95.0
+2020-02-25,2020-02-28,4,995.0,2020-02-28,H605021010,867.8499999999999,95.0
+2020-03-01,2020-03-11,11,1118.0,2020-03-07,H605021010,867.8499999999999,95.0
+2020-03-13,2020-03-13,1,909.0,2020-03-13,H605021010,867.8499999999999,95.0
+2020-12-23,2020-12-25,3,1075.0,2020-12-24,H605021010,867.8499999999999,95.0
+2020-12-27,2020-12-30,4,1105.0,2020-12-28,H605021010,867.8499999999999,95.0
+2021-01-22,2021-01-22,1,892.0,2021-01-22,H605021010,867.8499999999999,95.0
+2021-01-29,2021-02-06,9,1116.0,2021-02-04,H605021010,867.8499999999999,95.0
+2021-06-22,2021-06-24,3,1021.0,2021-06-23,H605021010,867.8499999999999,95.0
+2021-07-13,2021-07-14,2,947.0,2021-07-14,H605021010,867.8499999999999,95.0
+2021-10-04,2021-10-04,1,868.0,2021-10-04,H605021010,867.8499999999999,95.0
+2021-12-05,2021-12-06,2,1014.0,2021-12-06,H605021010,867.8499999999999,95.0
+2021-12-10,2021-12-11,2,1071.0,2021-12-11,H605021010,867.8499999999999,95.0
+2021-12-28,2021-12-31,4,1058.0,2021-12-30,H605021010,867.8499999999999,95.0
+2022-01-04,2022-01-06,3,1103.0,2022-01-05,H605021010,867.8499999999999,95.0
+2022-01-08,2022-01-11,4,1105.0,2022-01-10,H605021010,867.8499999999999,95.0
+2022-04-08,2022-04-11,4,1172.0,2022-04-10,H605021010,867.8499999999999,95.0
+2013-01-01,2013-01-02,2,782.0,2013-01-01,H605641101,711.5999999999999,95.0
+2013-01-12,2013-01-12,1,712.0,2013-01-12,H605641101,711.5999999999999,95.0
+2013-01-27,2013-01-31,5,923.0,2013-01-28,H605641101,711.5999999999999,95.0
+2013-02-02,2013-02-02,1,778.0,2013-02-02,H605641101,711.5999999999999,95.0
+2013-02-06,2013-02-07,2,801.0,2013-02-07,H605641101,711.5999999999999,95.0
+2013-02-10,2013-02-13,4,1070.0,2013-02-12,H605641101,711.5999999999999,95.0
+2013-03-16,2013-03-22,7,1130.0,2013-03-20,H605641101,711.5999999999999,95.0
+2013-04-10,2013-04-12,3,790.0,2013-04-10,H605641101,711.5999999999999,95.0
+2013-12-25,2013-12-30,6,1330.0,2013-12-29,H605641101,711.5999999999999,95.0
+2014-01-02,2014-01-03,2,777.0,2014-01-03,H605641101,711.5999999999999,95.0
+2014-01-28,2014-02-03,7,1147.0,2014-02-02,H605641101,711.5999999999999,95.0
+2014-02-07,2014-02-10,4,990.0,2014-02-09,H605641101,711.5999999999999,95.0
+2014-02-13,2014-02-16,4,1009.0,2014-02-14,H605641101,711.5999999999999,95.0
+2014-03-01,2014-03-02,2,844.0,2014-03-01,H605641101,711.5999999999999,95.0
+2014-05-26,2014-05-26,1,738.0,2014-05-26,H605641101,711.5999999999999,95.0
+2014-05-29,2014-05-29,1,764.0,2014-05-29,H605641101,711.5999999999999,95.0
+2014-10-17,2014-10-18,2,845.0,2014-10-18,H605641101,711.5999999999999,95.0
+2014-12-13,2014-12-15,3,1034.0,2014-12-14,H605641101,711.5999999999999,95.0
+2014-12-18,2014-12-20,3,955.0,2014-12-19,H605641101,711.5999999999999,95.0
+2015-01-16,2015-01-16,1,847.0,2015-01-16,H605641101,711.5999999999999,95.0
+2015-01-30,2015-02-02,4,904.0,2015-01-31,H605641101,711.5999999999999,95.0
+2015-02-21,2015-02-22,2,881.0,2015-02-21,H605641101,711.5999999999999,95.0
+2015-02-24,2015-02-24,1,732.0,2015-02-24,H605641101,711.5999999999999,95.0
+2015-02-27,2015-02-28,2,825.0,2015-02-27,H605641101,711.5999999999999,95.0
+2015-03-30,2015-03-31,2,790.0,2015-03-30,H605641101,711.5999999999999,95.0
+2016-02-09,2016-02-11,3,1022.0,2016-02-11,H605641101,711.5999999999999,95.0
+2016-02-13,2016-02-15,3,916.0,2016-02-15,H605641101,711.5999999999999,95.0
+2016-03-09,2016-03-10,2,744.0,2016-03-10,H605641101,711.5999999999999,95.0
+2016-04-11,2016-04-12,2,746.0,2016-04-11,H605641101,711.5999999999999,95.0
+2017-03-07,2017-03-08,2,761.0,2017-03-07,H605641101,711.5999999999999,95.0
+2013-12-25,2013-12-30,6,2995.0,2013-12-29,H605641201,2587.6,95.0
+2014-01-02,2014-01-03,2,2684.0,2014-01-03,H605641201,2587.6,95.0
+2014-01-27,2014-02-03,8,2886.0,2014-02-03,H605641201,2587.6,95.0
+2014-02-06,2014-02-10,5,2825.0,2014-02-09,H605641201,2587.6,95.0
+2014-02-13,2014-02-16,4,2838.0,2014-02-15,H605641201,2587.6,95.0
+2014-02-28,2014-03-02,3,2726.0,2014-03-02,H605641201,2587.6,95.0
+2014-05-02,2014-05-02,1,2591.0,2014-05-02,H605641201,2587.6,95.0
+2014-05-26,2014-05-26,1,2603.0,2014-05-26,H605641201,2587.6,95.0
+2014-05-29,2014-05-29,1,2645.0,2014-05-29,H605641201,2587.6,95.0
+2014-10-17,2014-10-18,2,2713.0,2014-10-18,H605641201,2587.6,95.0
+2014-12-13,2014-12-15,3,2809.0,2014-12-15,H605641201,2587.6,95.0
+2014-12-18,2014-12-21,4,2785.0,2014-12-19,H605641201,2587.6,95.0
+2015-01-16,2015-01-17,2,2722.0,2015-01-16,H605641201,2587.6,95.0
+2015-01-19,2015-01-20,2,2612.0,2015-01-20,H605641201,2587.6,95.0
+2015-01-30,2015-02-03,5,2755.0,2015-01-31,H605641201,2587.6,95.0
+2015-02-21,2015-02-24,4,2729.0,2015-02-22,H605641201,2587.6,95.0
+2015-02-27,2015-02-28,2,2709.0,2015-02-28,H605641201,2587.6,95.0
+2015-03-02,2015-03-03,2,2602.0,2015-03-03,H605641201,2587.6,95.0
+2015-03-30,2015-03-31,2,2666.0,2015-03-31,H605641201,2587.6,95.0
+2016-02-09,2016-02-11,3,2757.0,2016-02-11,H605641201,2587.6,95.0
+2016-02-14,2016-02-15,2,2736.0,2016-02-15,H605641201,2587.6,95.0
+2016-03-10,2016-03-11,2,2619.0,2016-03-10,H605641201,2587.6,95.0
+2016-03-29,2016-03-30,2,2660.0,2016-03-29,H605641201,2587.6,95.0
+2016-04-11,2016-04-12,2,2609.0,2016-04-12,H605641201,2587.6,95.0
+2017-03-07,2017-03-08,2,2669.0,2017-03-07,H605641201,2587.6,95.0
+2020-12-23,2020-12-25,3,1079.0,2020-12-24,H605641401,943.6999999999998,95.0
+2020-12-28,2020-12-30,3,1157.0,2020-12-29,H605641401,943.6999999999998,95.0
+2021-01-29,2021-02-06,9,1244.0,2021-02-05,H605641401,943.6999999999998,95.0
+2021-07-14,2021-07-14,1,944.0,2021-07-14,H605641401,943.6999999999998,95.0
+2021-12-06,2021-12-07,2,1015.0,2021-12-06,H605641401,943.6999999999998,95.0
+2021-12-10,2021-12-11,2,1114.0,2021-12-11,H605641401,943.6999999999998,95.0
+2021-12-29,2021-12-31,3,1087.0,2021-12-30,H605641401,943.6999999999998,95.0
+2022-01-04,2022-01-06,3,1177.0,2022-01-05,H605641401,943.6999999999998,95.0
+2022-01-09,2022-01-10,2,1189.0,2022-01-10,H605641401,943.6999999999998,95.0
+2022-04-09,2022-04-10,2,1260.0,2022-04-10,H605641401,943.6999999999998,95.0
+2023-01-16,2023-01-18,3,1121.0,2023-01-17,H605641401,943.6999999999998,95.0
+2023-11-13,2023-11-14,2,1039.0,2023-11-13,H605641401,943.6999999999998,95.0
+2023-11-17,2023-11-17,1,1030.0,2023-11-17,H605641401,943.6999999999998,95.0
+2023-11-19,2023-11-20,2,1083.0,2023-11-20,H605641401,943.6999999999998,95.0
+2023-12-10,2023-12-12,3,1180.0,2023-12-11,H605641401,943.6999999999998,95.0
+2024-01-02,2024-01-06,5,1267.0,2024-01-04,H605641401,943.6999999999998,95.0
+2024-01-17,2024-01-20,4,1523.0,2024-01-19,H605641401,943.6999999999998,95.0
+2024-01-22,2024-01-24,3,1148.0,2024-01-23,H605641401,943.6999999999998,95.0
+2024-02-11,2024-02-11,1,983.0,2024-02-11,H605641401,943.6999999999998,95.0
+2024-02-22,2024-02-24,3,1193.0,2024-02-24,H605641401,943.6999999999998,95.0
+2024-02-26,2024-02-27,2,1205.0,2024-02-27,H605641401,943.6999999999998,95.0
+2024-03-02,2024-03-03,2,1124.0,2024-03-02,H605641401,943.6999999999998,95.0
+2024-04-04,2024-04-06,3,1149.0,2024-04-05,H605641401,943.6999999999998,95.0
+2024-05-02,2024-05-04,3,1194.0,2024-05-03,H605641401,943.6999999999998,95.0
+2024-05-21,2024-05-23,3,1094.0,2024-05-22,H605641401,943.6999999999998,95.0
+2024-11-24,2024-11-24,1,965.0,2024-11-24,H605641401,943.6999999999998,95.0
+2024-11-26,2024-11-27,2,990.0,2024-11-26,H605641401,943.6999999999998,95.0
+2024-12-08,2024-12-10,3,1153.0,2024-12-10,H605641401,943.6999999999998,95.0
+2024-12-22,2024-12-23,2,990.0,2024-12-23,H605641401,943.6999999999998,95.0
+2025-01-02,2025-01-11,10,1495.0,2025-01-10,H605641401,943.6999999999998,95.0
+2025-01-25,2025-02-02,9,1831.0,2025-01-26,H605641401,943.6999999999998,95.0
+2026-02-12,2026-02-15,4,1172.0,2026-02-14,H605641401,943.6999999999998,95.0
+2026-02-18,2026-02-21,4,1483.0,2026-02-20,H605641401,943.6999999999998,95.0
+2013-01-28,2013-01-31,4,390.0,2013-01-28,H620021010,304.0500000000002,95.0
+2013-02-06,2013-02-08,3,369.0,2013-02-07,H620021010,304.0500000000002,95.0
+2013-02-11,2013-02-13,3,578.0,2013-02-12,H620021010,304.0500000000002,95.0
+2013-03-17,2013-03-22,6,539.0,2013-03-21,H620021010,304.0500000000002,95.0
+2013-12-26,2013-12-30,5,593.0,2013-12-30,H620021010,304.0500000000002,95.0
+2014-01-29,2014-02-03,6,581.0,2014-02-03,H620021010,304.0500000000002,95.0
+2014-02-08,2014-02-10,3,411.0,2014-02-10,H620021010,304.0500000000002,95.0
+2014-02-14,2014-02-16,3,453.0,2014-02-15,H620021010,304.0500000000002,95.0
+2014-03-01,2014-03-02,2,389.0,2014-03-02,H620021010,304.0500000000002,95.0
+2014-10-18,2014-10-18,1,361.0,2014-10-18,H620021010,304.0500000000002,95.0
+2014-12-13,2014-12-15,3,604.0,2014-12-15,H620021010,304.0500000000002,95.0
+2014-12-18,2014-12-21,4,528.0,2014-12-19,H620021010,304.0500000000002,95.0
+2015-01-16,2015-01-17,2,350.0,2015-01-16,H620021010,304.0500000000002,95.0
+2015-01-30,2015-02-03,5,527.0,2015-02-01,H620021010,304.0500000000002,95.0
+2015-02-21,2015-02-22,2,408.0,2015-02-22,H620021010,304.0500000000002,95.0
+2015-02-24,2015-02-25,2,367.0,2015-02-24,H620021010,304.0500000000002,95.0
+2015-02-27,2015-02-28,2,402.0,2015-02-28,H620021010,304.0500000000002,95.0
+2015-03-31,2015-03-31,1,362.0,2015-03-31,H620021010,304.0500000000002,95.0
+2016-02-10,2016-02-12,3,572.0,2016-02-11,H620021010,304.0500000000002,95.0
+2016-02-14,2016-02-16,3,506.0,2016-02-15,H620021010,304.0500000000002,95.0
+2016-03-29,2016-03-30,2,389.0,2016-03-30,H620021010,304.0500000000002,95.0
+2016-04-12,2016-04-12,1,350.0,2016-04-12,H620021010,304.0500000000002,95.0
+2017-03-07,2017-03-08,2,394.0,2017-03-08,H620021010,304.0500000000002,95.0
+2017-12-12,2017-12-13,2,361.0,2017-12-13,H620021010,304.0500000000002,95.0
+2017-12-15,2017-12-17,3,361.0,2017-12-17,H620021010,304.0500000000002,95.0
+2017-12-28,2018-01-07,11,737.0,2018-01-06,H620021010,304.0500000000002,95.0
+2018-01-17,2018-01-26,10,780.0,2018-01-22,H620021010,304.0500000000002,95.0
+2018-02-01,2018-02-02,2,386.0,2018-02-02,H620021010,304.0500000000002,95.0
+2018-02-12,2018-02-13,2,434.0,2018-02-12,H620021010,304.0500000000002,95.0
+2018-02-15,2018-02-17,3,393.0,2018-02-16,H620021010,304.0500000000002,95.0
+2018-03-09,2018-03-13,5,376.0,2018-03-12,H620021010,304.0500000000002,95.0
+2018-03-31,2018-04-01,2,426.0,2018-03-31,H620021010,304.0500000000002,95.0
+2018-06-05,2018-06-07,3,637.0,2018-06-05,H620021010,304.0500000000002,95.0
+2018-06-11,2018-06-16,6,1178.0,2018-06-14,H620021010,304.0500000000002,95.0
+2019-02-10,2019-02-12,3,566.0,2019-02-12,H620021010,304.0500000000002,95.0
+2020-02-02,2020-02-06,5,509.0,2020-02-04,H620021010,304.0500000000002,95.0
+2020-02-10,2020-02-13,4,439.0,2020-02-12,H620021010,304.0500000000002,95.0
+2020-02-26,2020-02-29,4,397.0,2020-02-27,H620021010,304.0500000000002,95.0
+2020-03-02,2020-03-11,10,895.0,2020-03-07,H620021010,304.0500000000002,95.0
+2020-12-24,2020-12-25,2,499.0,2020-12-25,H620021010,304.0500000000002,95.0
+2020-12-28,2020-12-31,4,543.0,2020-12-29,H620021010,304.0500000000002,95.0
+2021-01-29,2021-02-06,9,642.0,2021-02-05,H620021010,304.0500000000002,95.0
+2021-06-23,2021-06-24,2,732.0,2021-06-24,H620021010,304.0500000000002,95.0
+2021-12-06,2021-12-07,2,340.0,2021-12-07,H620021010,304.0500000000002,95.0
+2021-12-11,2021-12-12,2,432.0,2021-12-11,H620021010,304.0500000000002,95.0
+2021-12-30,2021-12-31,2,347.0,2021-12-30,H620021010,304.0500000000002,95.0
+2022-01-04,2022-01-06,3,534.0,2022-01-06,H620021010,304.0500000000002,95.0
+2022-01-08,2022-01-11,4,516.0,2022-01-10,H620021010,304.0500000000002,95.0
+2022-04-09,2022-04-11,3,501.0,2022-04-10,H620021010,304.0500000000002,95.0
+2023-01-17,2023-01-18,2,367.0,2023-01-18,H620021010,304.0500000000002,95.0
+2023-11-19,2023-11-20,2,359.0,2023-11-20,H620021010,304.0500000000002,95.0
+2023-12-10,2023-12-12,3,530.0,2023-12-11,H620021010,304.0500000000002,95.0
+2024-01-03,2024-01-06,4,593.0,2024-01-04,H620021010,304.0500000000002,95.0
+2024-01-18,2024-01-20,3,748.0,2024-01-19,H620021010,304.0500000000002,95.0
+2024-01-23,2024-01-24,2,440.0,2024-01-23,H620021010,304.0500000000002,95.0
+2024-02-11,2024-02-12,2,349.0,2024-02-12,H620021010,304.0500000000002,95.0
+2024-02-23,2024-02-28,6,504.0,2024-02-27,H620021010,304.0500000000002,95.0
+2024-03-02,2024-03-03,2,411.0,2024-03-03,H620021010,304.0500000000002,95.0
+2024-04-05,2024-04-06,2,463.0,2024-04-06,H620021010,304.0500000000002,95.0
+2024-05-02,2024-05-04,3,687.0,2024-05-04,H620021010,304.0500000000002,95.0
+2024-05-12,2024-05-12,1,384.0,2024-05-12,H620021010,304.0500000000002,95.0
+2024-05-22,2024-05-23,2,413.0,2024-05-23,H620021010,304.0500000000002,95.0
+2024-12-08,2024-12-11,4,582.0,2024-12-10,H620021010,304.0500000000002,95.0
+2024-12-23,2024-12-23,1,323.0,2024-12-23,H620021010,304.0500000000002,95.0
+2025-01-03,2025-01-12,10,696.0,2025-01-10,H620021010,304.0500000000002,95.0
+2025-01-25,2025-02-02,9,871.0,2025-01-27,H620021010,304.0500000000002,95.0
+2026-02-12,2026-02-16,5,419.0,2026-02-14,H620021010,304.0500000000002,95.0
+2026-02-18,2026-02-22,5,594.0,2026-02-21,H620021010,304.0500000000002,95.0
+2013-01-29,2013-01-30,2,991.0,2013-01-29,H623022010,927.0,95.0
+2013-02-06,2013-02-08,3,969.0,2013-02-06,H623022010,927.0,95.0
+2013-02-10,2013-02-13,4,1102.0,2013-02-12,H623022010,927.0,95.0
+2013-03-17,2013-03-22,6,1089.0,2013-03-21,H623022010,927.0,95.0
+2013-12-26,2013-12-31,6,1105.0,2013-12-30,H623022010,927.0,95.0
+2014-01-29,2014-02-04,6,1193.0,2014-02-03,H623022010,927.0,95.0
+2014-02-06,2014-02-11,6,1026.0,2014-02-10,H623022010,927.0,95.0
+2014-02-14,2014-02-17,4,1069.0,2014-02-15,H623022010,927.0,95.0
+2014-03-01,2014-03-03,3,1014.0,2014-03-02,H623022010,927.0,95.0
+2014-10-18,2014-10-19,2,1078.0,2014-10-18,H623022010,927.0,95.0
+2014-12-14,2014-12-16,3,1158.0,2014-12-15,H623022010,927.0,95.0
+2014-12-18,2014-12-21,4,1168.0,2014-12-19,H623022010,927.0,95.0
+2015-01-16,2015-01-17,2,1018.0,2015-01-17,H623022010,927.0,95.0
+2015-01-30,2015-02-03,5,1203.0,2015-02-01,H623022010,927.0,95.0
+2015-02-21,2015-03-01,9,1020.0,2015-02-22,H623022010,927.0,95.0
+2015-03-03,2015-03-03,1,927.0,2015-03-03,H623022010,927.0,95.0
+2015-03-31,2015-03-31,1,964.0,2015-03-31,H623022010,927.0,95.0
+2016-02-01,2016-02-01,1,932.0,2016-02-01,H623022010,927.0,95.0
+2016-02-09,2016-02-16,8,1114.0,2016-02-11,H623022010,927.0,95.0
+2016-03-05,2016-03-05,1,993.0,2016-03-05,H623022010,927.0,95.0
+2016-03-10,2016-03-11,2,979.0,2016-03-11,H623022010,927.0,95.0
+2016-03-28,2016-03-30,3,1037.0,2016-03-30,H623022010,927.0,95.0
+2016-04-12,2016-04-13,2,973.0,2016-04-12,H623022010,927.0,95.0
+2017-03-06,2017-03-08,3,1031.0,2017-03-08,H623022010,927.0,95.0

datasets/fp_details_horizon_5.csv

@@ -0,0 +1,315 @@
+station_code,target_date,predicted_value,threshold,margin,days_to_nearest_real_event
+H605641201,2026-01-05,8528.576171875,6639.9000000000015,1888.6761718749985,3225
+H605641101,2026-01-06,8860.708984375,8347.999999999998,512.7089843750018,3962
+H605641201,2026-01-06,9122.71875,6639.9000000000015,2482.8187499999985,3226
+H605641101,2026-01-07,8719.134765625,8347.999999999998,371.1347656250018,3963
+H605641201,2026-01-07,8959.505859375,6639.9000000000015,2319.6058593749985,3227
+H605641101,2026-01-08,8719.12109375,8347.999999999998,371.1210937500018,3964
+H605641201,2026-01-08,8959.462890625,6639.9000000000015,2319.5628906249985,3228
+H605641201,2026-01-09,8455.52734375,6639.9000000000015,1815.6273437499985,3229
+H605641101,2026-01-10,8633.087890625,8347.999999999998,285.0878906250018,3966
+H605641201,2026-01-10,8629.1689453125,6639.9000000000015,1989.2689453124985,3230
+H605641101,2026-01-11,8416.341796875,8347.999999999998,68.34179687500182,3967
+H605641201,2026-01-11,8679.5224609375,6639.9000000000015,2039.6224609374985,3231
+H605641101,2026-01-12,8549.6171875,8347.999999999998,201.61718750000182,3968
+H605641201,2026-01-12,8790.5654296875,6639.9000000000015,2150.6654296874985,3232
+H605641201,2026-01-13,8382.22265625,6639.9000000000015,1742.3226562499985,3233
+H420021012,2026-01-14,28771.474609375,28601.19999999996,170.2746093750393,30
+H605641101,2026-01-14,8803.759765625,8347.999999999998,455.7597656250018,3970
+H605641201,2026-01-14,9052.775390625,6639.9000000000015,2412.8753906249985,3234
+H605641201,2026-01-15,8497.1171875,6639.9000000000015,1857.2171874999985,3235
+H605641201,2026-01-16,8325.703125,6639.9000000000015,1685.8031249999985,3236
+H605641201,2026-01-17,8501.5830078125,6639.9000000000015,1861.6830078124985,3237
+H605641201,2026-01-18,8325.0576171875,6639.9000000000015,1685.1576171874985,3238
+H605641201,2026-01-19,8438.431640625,6639.9000000000015,1798.5316406249985,3239
+H605641201,2026-01-20,8280.759765625,6639.9000000000015,1640.8597656249985,3240
+H605641201,2026-01-21,8293.8681640625,6639.9000000000015,1653.9681640624985,3241
+H404021101,2026-01-22,11117.796875,11109.799999999992,7.9968750000080036,16
+H605641201,2026-01-22,8389.580078125,6639.9000000000015,1749.6800781249985,3242
+H605641201,2026-01-23,8366.2490234375,6639.9000000000015,1726.3490234374985,3243
+H404021101,2026-01-24,11369.4921875,11109.799999999992,259.692187500008,14
+H605641201,2026-01-24,8405.3427734375,6639.9000000000015,1765.4427734374985,3244
+H404021101,2026-01-25,13755.953125,11109.799999999992,2646.153125000008,13
+H605641201,2026-01-25,8414.08203125,6639.9000000000015,1774.1820312499985,3245
+H404021101,2026-01-26,15700.1083984375,11109.799999999992,4590.308398437508,12
+H605641101,2026-01-26,8899.6640625,8347.999999999998,551.6640625000018,3982
+H605641201,2026-01-26,8960.4404296875,6639.9000000000015,2320.5404296874985,3246
+H605641201,2026-01-27,8413.9638671875,6639.9000000000015,1774.0638671874985,3247
+H605641201,2026-01-28,8589.099609375,6639.9000000000015,1949.1996093749985,3248
+H605641101,2026-01-29,8454.115234375,8347.999999999998,106.11523437500182,3985
+H605641201,2026-01-29,8712.046875,6639.9000000000015,2072.1468749999985,3249
+H605641201,2026-01-30,8468.86328125,6639.9000000000015,1828.9632812499985,3250
+H605641101,2026-01-31,8549.5361328125,8347.999999999998,201.53613281250182,3987
+H605641201,2026-01-31,8781.8349609375,6639.9000000000015,2141.9349609374985,3251
+H605641201,2026-02-01,8369.1591796875,6639.9000000000015,1729.2591796874985,3252
+H420021012,2026-02-02,29669.896484375,28601.19999999996,1068.6964843750393,11
+H605641201,2026-02-02,8316.609375,6639.9000000000015,1676.7093749999985,3253
+H605641201,2026-02-03,8329.962890625,6639.9000000000015,1690.0628906249985,3254
+H605641101,2026-02-04,8892.11328125,8347.999999999998,544.1132812500018,3991
+H605641201,2026-02-04,9150.9140625,6639.9000000000015,2511.0140624999985,3255
+H605641101,2026-02-05,8931.060546875,8347.999999999998,583.0605468750018,3992
+H605641201,2026-02-05,9188.083984375,6639.9000000000015,2548.1839843749985,3256
+H605641201,2026-02-06,8293.5205078125,6639.9000000000015,1653.6205078124985,3257
+H605641201,2026-02-07,8259.6484375,6639.9000000000015,1619.7484374999985,3258
+H605641201,2026-02-08,8304.7177734375,6639.9000000000015,1664.8177734374985,3259
+H420021012,2026-02-09,32454.205078125,28601.19999999996,3853.0050781250393,4
+H605641401,2026-02-09,9319.359375,6781.299999999995,2538.0593750000053,3
+H605641101,2026-02-09,9057.78515625,8347.999999999998,709.7851562500018,3996
+H605641201,2026-02-09,9461.982421875,6639.9000000000015,2822.0824218749985,3260
+H404021101,2026-02-10,14642.9375,11109.799999999992,3533.137500000008,1
+H605641401,2026-02-10,9730.62890625,6781.299999999995,2949.3289062500053,2
+H605641101,2026-02-10,8355.31640625,8347.999999999998,7.316406250001819,3997
+H605641201,2026-02-10,8620.76171875,6639.9000000000015,1980.8617187499985,3261
+H404021101,2026-02-11,16081.4140625,11109.799999999992,4971.614062500008,2
+H420021012,2026-02-11,29221.00390625,28601.19999999996,619.8039062500393,2
+H605641401,2026-02-11,10703.0634765625,6781.299999999995,3921.7634765625053,1
+H605641101,2026-02-11,8455.123046875,8347.999999999998,107.12304687500182,3998
+H605641201,2026-02-11,8594.908203125,6639.9000000000015,1955.0082031249985,3262
+H404021101,2026-02-12,16625.81640625,11109.799999999992,5516.016406250008,1
+H420021012,2026-02-12,29411.03125,28601.19999999996,809.8312500000393,1
+H605641101,2026-02-12,8833.2666015625,8347.999999999998,485.2666015625018,3999
+H605641201,2026-02-12,8795.8857421875,6639.9000000000015,2155.9857421874985,3263
+H605641201,2026-02-13,8033.7119140625,6639.9000000000015,1393.8119140624985,3264
+H605641201,2026-02-14,8186.1220703125,6639.9000000000015,1546.2220703124985,3265
+H605641101,2026-02-15,8719.833984375,8347.999999999998,371.8339843750018,4002
+H605641201,2026-02-15,8706.8125,6639.9000000000015,2066.9124999999985,3266
+H605641201,2026-02-16,8206.763671875,6639.9000000000015,1566.8636718749985,3267
+H605641101,2026-02-17,8987.53125,8347.999999999998,639.5312500000018,4004
+H605641201,2026-02-17,9041.111328125,6639.9000000000015,2401.2113281249985,3268
+H605641101,2026-02-18,8380.693359375,8347.999999999998,32.69335937500182,4005
+H605641201,2026-02-18,8084.51708984375,6639.9000000000015,1444.6170898437485,3269
+H605641201,2026-02-19,7456.0517578125,6639.9000000000015,816.1517578124985,3270
+H605641201,2026-02-20,7214.9140625,6639.9000000000015,575.0140624999985,3271
+H605641101,2026-02-21,8721.03125,8347.999999999998,373.0312500000018,4008
+H605641201,2026-02-21,7712.6279296875,6639.9000000000015,1072.7279296874985,3272
+H605641201,2026-02-22,7205.34228515625,6639.9000000000015,565.4422851562485,3273
+H605641401,2026-02-23,9585.287109375,6781.299999999995,2803.9871093750053,1
+H605641201,2026-02-23,7320.9091796875,6639.9000000000015,681.0091796874985,3274
+H605641401,2026-02-24,13155.92578125,6781.299999999995,6374.625781250005,2
+H605641201,2026-02-24,7313.6669921875,6639.9000000000015,673.7669921874985,3275
+H404021101,2026-02-25,22475.908203125,11109.799999999992,11366.108203125008,1
+H420021012,2026-02-25,45311.0703125,28601.19999999996,16709.87031250004,1
+H605641401,2026-02-25,13831.779296875,6781.299999999995,7050.479296875005,3
+H605641201,2026-02-25,7184.7978515625,6639.9000000000015,544.8978515624985,3276
+H404021101,2026-02-26,21915.78515625,11109.799999999992,10805.985156250008,2
+H420021012,2026-02-26,44744.11328125,28601.19999999996,16142.91328125004,2
+H605641401,2026-02-26,13133.146484375,6781.299999999995,6351.846484375005,4
+H605641201,2026-02-26,7063.220703125,6639.9000000000015,423.32070312499854,3277
+H404021101,2026-02-27,14260.5283203125,11109.799999999992,3150.728320312508,3
+H420021012,2026-02-27,43650.8359375,28601.19999999996,15049.63593750004,3
+H605641401,2026-02-27,9941.578125,6781.299999999995,3160.2781250000053,5
+H605641201,2026-02-27,6950.7587890625,6639.9000000000015,310.85878906249854,3278
+H420021012,2026-02-28,43123.42578125,28601.19999999996,14522.22578125004,4
+H605641401,2026-02-28,8213.2763671875,6781.299999999995,1431.9763671875053,6
+H605641201,2026-02-28,7041.146484375,6639.9000000000015,401.24648437499854,3279
+H420021012,2026-03-01,39676.58203125,28601.19999999996,11075.38203125004,5
+H605641401,2026-03-01,7570.470703125,6781.299999999995,789.1707031250053,7
+H605641201,2026-03-01,7223.05859375,6639.9000000000015,583.1585937499985,3280
+H605641401,2026-03-02,7382.1640625,6781.299999999995,600.8640625000053,8
+H605641201,2026-03-02,7615.73046875,6639.9000000000015,975.8304687499985,3281
+H605641401,2026-03-03,7429.73095703125,6781.299999999995,648.4309570312553,9
+H605641201,2026-03-03,7975.48974609375,6639.9000000000015,1335.5897460937485,3282
+H404021101,2026-03-04,12209.87109375,11109.799999999992,1100.071093750008,8
+H605641401,2026-03-04,7762.84521484375,6781.299999999995,981.5452148437553,10
+H605641101,2026-03-04,8457.6435546875,8347.999999999998,109.64355468750182,4019
+H605641201,2026-03-04,8602.3076171875,6639.9000000000015,1962.4076171874985,3283
+H404021101,2026-03-05,12661.724609375,11109.799999999992,1551.924609375008,9
+H605641401,2026-03-05,8480.791015625,6781.299999999995,1699.4910156250053,11
+H605641101,2026-03-05,8807.2021484375,8347.999999999998,459.2021484375018,4020
+H605641201,2026-03-05,8991.29296875,6639.9000000000015,2351.3929687499985,3284
+H404021101,2026-03-06,12183.1015625,11109.799999999992,1073.301562500008,10
+H605641401,2026-03-06,8764.45703125,6781.299999999995,1983.1570312500053,12
+H605641201,2026-03-06,8272.888671875,6639.9000000000015,1632.9886718749985,3285
+H404021101,2026-03-07,12106.666015625,11109.799999999992,996.866015625008,11
+H605641401,2026-03-07,9332.494140625,6781.299999999995,2551.1941406250053,13
+H605641201,2026-03-07,8269.7890625,6639.9000000000015,1629.8890624999985,3286
+H404021101,2026-03-08,13027.74609375,11109.799999999992,1917.946093750008,12
+H605641401,2026-03-08,10550.05078125,6781.299999999995,3768.7507812500053,14
+H605641101,2026-03-08,9234.32421875,8347.999999999998,886.3242187500018,4023
+H605641201,2026-03-08,9524.865234375,6639.9000000000015,2884.9652343749985,3287
+H404021101,2026-03-09,14122.197265625,11109.799999999992,3012.397265625008,13
+H605641401,2026-03-09,9996.009765625,6781.299999999995,3214.7097656250053,15
+H605641201,2026-03-09,8285.1923828125,6639.9000000000015,1645.2923828124985,3288
+H404021101,2026-03-10,15181.0537109375,11109.799999999992,4071.253710937508,14
+H605641401,2026-03-10,10359.3828125,6781.299999999995,3578.0828125000053,16
+H605641201,2026-03-10,8516.802734375,6639.9000000000015,1876.9027343749985,3289
+H404021101,2026-03-11,15828.0625,11109.799999999992,4718.262500000008,15
+H605641401,2026-03-11,10301.2646484375,6781.299999999995,3519.9646484375053,17
+H605641201,2026-03-11,8301.4873046875,6639.9000000000015,1661.5873046874985,3290
+H404021101,2026-03-12,15285.15625,11109.799999999992,4175.356250000008,16
+H605641401,2026-03-12,10400.265625,6781.299999999995,3618.9656250000053,18
+H605641101,2026-03-12,8597.3916015625,8347.999999999998,249.39160156250182,4027
+H605641201,2026-03-12,8822.44140625,6639.9000000000015,2182.5414062499985,3291
+H404021101,2026-03-13,15947.4951171875,11109.799999999992,4837.695117187508,17
+H605641401,2026-03-13,9431.3193359375,6781.299999999995,2650.0193359375053,19
+H605641201,2026-03-13,8292.8779296875,6639.9000000000015,1652.9779296874985,3292
+H404021101,2026-03-14,16404.125,11109.799999999992,5294.325000000008,18
+H605641401,2026-03-14,8768.744140625,6781.299999999995,1987.4441406250053,20
+H605641201,2026-03-14,8343.994140625,6639.9000000000015,1704.0941406249985,3293
+H404021101,2026-03-15,15703.7958984375,11109.799999999992,4593.995898437508,19
+H605641401,2026-03-15,8097.43896484375,6781.299999999995,1316.1389648437553,21
+H605641201,2026-03-15,8436.87109375,6639.9000000000015,1796.9710937499985,3294
+H404021101,2026-03-16,15224.1806640625,11109.799999999992,4114.380664062508,20
+H605641401,2026-03-16,7309.216796875,6781.299999999995,527.9167968750053,22
+H605641201,2026-03-16,8310.849609375,6639.9000000000015,1670.9496093749985,3295
+H605641401,2026-03-17,6933.4716796875,6781.299999999995,152.17167968750528,23
+H605641201,2026-03-17,8299.462890625,6639.9000000000015,1659.5628906249985,3296
+H404021101,2026-03-18,13918.859375,11109.799999999992,2809.059375000008,22
+H605641201,2026-03-18,8301.634765625,6639.9000000000015,1661.7347656249985,3297
+H404021101,2026-03-19,11165.2568359375,11109.799999999992,55.456835937508004,23
+H420021012,2026-03-19,34861.0859375,28601.19999999996,6259.885937500039,23
+H605641401,2026-03-19,9401.1494140625,6781.299999999995,2619.8494140625053,25
+H605641201,2026-03-19,8321.45703125,6639.9000000000015,1681.5570312499985,3298
+H605641401,2026-03-20,10358.408203125,6781.299999999995,3577.1082031250053,26
+H605641201,2026-03-20,8295.30078125,6639.9000000000015,1655.4007812499985,3299
+H605641401,2026-03-21,10334.1455078125,6781.299999999995,3552.8455078125053,27
+H605641201,2026-03-21,8239.3408203125,6639.9000000000015,1599.4408203124985,3300
+H605641401,2026-03-22,8479.373046875,6781.299999999995,1698.0730468750053,28
+H605641201,2026-03-22,8187.833984375,6639.9000000000015,1547.9339843749985,3301
+H605641401,2026-03-23,7655.9130859375,6781.299999999995,874.6130859375053,29
+H605641201,2026-03-23,8169.01025390625,6639.9000000000015,1529.1102539062485,3302
+H404021101,2026-03-24,13126.755859375,11109.799999999992,2016.955859375008,28
+H605641401,2026-03-24,7941.4052734375,6781.299999999995,1160.1052734375053,30
+H605641101,2026-03-24,8359.546875,8347.999999999998,11.546875000001819,4039
+H605641201,2026-03-24,8527.783203125,6639.9000000000015,1887.8832031249985,3303
+H404021101,2026-03-25,12331.326171875,11109.799999999992,1221.526171875008,29
+H605641401,2026-03-25,8461.705078125,6781.299999999995,1680.4050781250053,31
+H605641101,2026-03-25,8446.291015625,8347.999999999998,98.29101562500182,4040
+H605641201,2026-03-25,8666.98046875,6639.9000000000015,2027.0804687499985,3304
+H404021101,2026-03-26,11503.01171875,11109.799999999992,393.211718750008,30
+H605641401,2026-03-26,8209.802734375,6781.299999999995,1428.5027343750053,32
+H605641201,2026-03-26,8310.443359375,6639.9000000000015,1670.5433593749985,3305
+H404021101,2026-03-27,12259.498046875,11109.799999999992,1149.698046875008,31
+H605641401,2026-03-27,8202.5810546875,6781.299999999995,1421.2810546875053,33
+H605641101,2026-03-27,8597.2607421875,8347.999999999998,249.26074218750182,4042
+H605641201,2026-03-27,8842.20703125,6639.9000000000015,2202.3070312499985,3306
+H404021101,2026-03-28,12722.4677734375,11109.799999999992,1612.667773437508,32
+H605641201,2026-03-28,8324.3212890625,6639.9000000000015,1684.4212890624985,3307
+H404021101,2026-03-29,12594.9755859375,11109.799999999992,1485.175585937508,33
+H605641201,2026-03-29,8392.99609375,6639.9000000000015,1753.0960937499985,3308
+H404021101,2026-03-30,14730.501953125,11109.799999999992,3620.701953125008,34
+H605641201,2026-03-30,8306.3154296875,6639.9000000000015,1666.4154296874985,3309
+H404021101,2026-03-31,13680.328125,11109.799999999992,2570.528125000008,35
+H605641201,2026-03-31,8320.80859375,6639.9000000000015,1680.9085937499985,3310
+H605641101,2026-04-01,8408.6494140625,8347.999999999998,60.64941406250182,4047
+H605641201,2026-04-01,8655.03515625,6639.9000000000015,2015.1351562499985,3311
+H605641201,2026-04-02,8357.7919921875,6639.9000000000015,1717.8919921874985,3312
+H605641201,2026-04-03,8316.8984375,6639.9000000000015,1676.9984374999985,3313
+H605641201,2026-04-04,8297.298828125,6639.9000000000015,1657.3988281249985,3314
+H605641201,2026-04-05,8296.650390625,6639.9000000000015,1656.7503906249985,3315
+H605641201,2026-04-06,8296.740234375,6639.9000000000015,1656.8402343749985,3316
+H605641201,2026-04-07,8298.0,6639.9000000000015,1658.0999999999985,3317
+H605641201,2026-04-08,8300.5,6639.9000000000015,1660.5999999999985,3318
+H605641201,2026-04-09,8303.359375,6639.9000000000015,1663.4593749999985,3319
+H605641201,2026-04-10,8434.40234375,6639.9000000000015,1794.5023437499985,3320
+H605641201,2026-04-11,8405.888671875,6639.9000000000015,1765.9886718749985,3321
+H605641101,2026-04-12,8445.005859375,8347.999999999998,97.00585937500182,4058
+H605641201,2026-04-12,8691.05859375,6639.9000000000015,2051.1585937499985,3322
+H605641201,2026-04-13,8306.40234375,6639.9000000000015,1666.5023437499985,3323
+H404021101,2026-04-14,11118.26953125,11109.799999999992,8.469531250008004,49
+H605641101,2026-04-14,8568.041015625,8347.999999999998,220.04101562500182,4060
+H605641201,2026-04-14,8788.3291015625,6639.9000000000015,2148.4291015624985,3324
+H605641201,2026-04-15,8376.8359375,6639.9000000000015,1736.9359374999985,3325
+H605641201,2026-04-16,8308.453125,6639.9000000000015,1668.5531249999985,3326
+H605641201,2026-04-17,8506.70703125,6639.9000000000015,1866.8070312499985,3327
+H605641201,2026-04-18,8309.7626953125,6639.9000000000015,1669.8626953124985,3328
+H605641201,2026-04-19,8310.537109375,6639.9000000000015,1670.6371093749985,3329
+H605641201,2026-04-20,8311.6220703125,6639.9000000000015,1671.7220703124985,3330
+H605641201,2026-04-21,8310.9453125,6639.9000000000015,1671.0453124999985,3331
+H605641201,2026-04-22,8307.9560546875,6639.9000000000015,1668.0560546874985,3332
+H605641201,2026-04-23,8305.7451171875,6639.9000000000015,1665.8451171874985,3333
+H605641201,2026-04-24,8306.5234375,6639.9000000000015,1666.6234374999985,3334
+H605641201,2026-04-25,8308.66015625,6639.9000000000015,1668.7601562499985,3335
+H605641201,2026-04-26,8310.0732421875,6639.9000000000015,1670.1732421874985,3336
+H605641201,2026-04-27,8374.396484375,6639.9000000000015,1734.4964843749985,3337
+H605641201,2026-04-28,8309.712890625,6639.9000000000015,1669.8128906249985,3338
+H605641201,2026-04-29,8308.95703125,6639.9000000000015,1669.0570312499985,3339
+H605641101,2026-04-30,8651.5908203125,8347.999999999998,303.5908203125018,4076
+H605641201,2026-04-30,8873.84765625,6639.9000000000015,2233.9476562499985,3340
+H605641101,2026-05-01,8941.5751953125,8347.999999999998,593.5751953125018,4077
+H605641201,2026-05-01,9005.169921875,6639.9000000000015,2365.2699218749985,3341
+H605641101,2026-05-02,9042.626953125,8347.999999999998,694.6269531250018,4078
+H605641201,2026-05-02,9166.8837890625,6639.9000000000015,2526.9837890624985,3342
+H605641201,2026-05-03,8399.07421875,6639.9000000000015,1759.1742187499985,3343
+H605641201,2026-05-04,8309.544921875,6639.9000000000015,1669.6449218749985,3344
+H605641201,2026-05-05,8309.5185546875,6639.9000000000015,1669.6185546874985,3345
+H605641201,2026-05-06,8309.6923828125,6639.9000000000015,1669.7923828124985,3346
+H605641201,2026-05-07,8311.2080078125,6639.9000000000015,1671.3080078124985,3347
+H605641201,2026-05-08,8417.123046875,6639.9000000000015,1777.2230468749985,3348
+H605641101,2026-05-09,8934.958984375,8347.999999999998,586.9589843750018,4085
+H605641201,2026-05-09,9222.7470703125,6639.9000000000015,2582.8470703124985,3349
+H420021012,2026-05-10,29426.814453125,28601.19999999996,825.6144531250393,75
+H605641201,2026-05-10,8426.37109375,6639.9000000000015,1786.4710937499985,3350
+H420021012,2026-05-11,36290.68359375,28601.19999999996,7689.483593750039,76
+H605641201,2026-05-11,8358.1640625,6639.9000000000015,1718.2640624999985,3351
+H605641401,2026-05-12,8103.181640625,6781.299999999995,1321.8816406250053,79
+H605641201,2026-05-12,8315.142578125,6639.9000000000015,1675.2425781249985,3352
+H605641401,2026-05-13,7525.759765625,6781.299999999995,744.4597656250053,80
+H605641101,2026-05-13,8567.5634765625,8347.999999999998,219.56347656250182,4089
+H605641201,2026-05-13,8748.927734375,6639.9000000000015,2109.0277343749985,3353
+H605641101,2026-05-14,8382.240234375,8347.999999999998,34.24023437500182,4090
+H605641201,2026-05-14,8530.599609375,6639.9000000000015,1890.6996093749985,3354
+H605641201,2026-05-15,8207.67578125,6639.9000000000015,1567.7757812499985,3355
+H605641101,2026-05-16,8764.1708984375,8347.999999999998,416.1708984375018,4092
+H605641201,2026-05-16,8953.7578125,6639.9000000000015,2313.8578124999985,3356
+H605641201,2026-05-17,8313.70703125,6639.9000000000015,1673.8070312499985,3357
+H605641201,2026-05-18,8578.498046875,6639.9000000000015,1938.5980468749985,3358
+H605641201,2026-05-19,8449.2392578125,6639.9000000000015,1809.3392578124985,3359
+H605641201,2026-05-20,8339.810546875,6639.9000000000015,1699.9105468749985,3360
+H605641201,2026-05-21,8321.671875,6639.9000000000015,1681.7718749999985,3361
+H605641101,2026-05-22,8637.259765625,8347.999999999998,289.2597656250018,4098
+H605641201,2026-05-22,8858.580078125,6639.9000000000015,2218.6800781249985,3362
+H420021012,2026-05-23,29508.943359375,28601.19999999996,907.7433593750393,88
+H605641201,2026-05-23,8305.7109375,6639.9000000000015,1665.8109374999985,3363
+H605641201,2026-05-24,8306.287109375,6639.9000000000015,1666.3871093749985,3364
+H605641201,2026-05-25,8308.8330078125,6639.9000000000015,1668.9330078124985,3365
+H605641201,2026-05-26,8310.2109375,6639.9000000000015,1670.3109374999985,3366
+H605641201,2026-05-27,8309.908203125,6639.9000000000015,1670.0082031249985,3367
+H605641201,2026-05-28,8307.90234375,6639.9000000000015,1668.0023437499985,3368
+H605641201,2026-05-29,8303.9375,6639.9000000000015,1664.0374999999985,3369
+H605641201,2026-05-30,8302.287109375,6639.9000000000015,1662.3871093749985,3370
+H605641101,2026-05-31,9092.265625,8347.999999999998,744.2656250000018,4107
+H605641201,2026-05-31,9370.048828125,6639.9000000000015,2730.1488281249985,3371
+H605641101,2026-06-01,8863.017578125,8347.999999999998,515.0175781250018,4108
+H605641201,2026-06-01,9115.865234375,6639.9000000000015,2475.9652343749985,3372
+H605641201,2026-06-02,8366.14453125,6639.9000000000015,1726.2445312499985,3373
+H605641201,2026-06-03,8364.578125,6639.9000000000015,1724.6781249999985,3374
+H605641201,2026-06-04,8305.5673828125,6639.9000000000015,1665.6673828124985,3375
+H605641101,2026-06-05,8552.248046875,8347.999999999998,204.24804687500182,4112
+H605641201,2026-06-05,8765.24609375,6639.9000000000015,2125.3460937499985,3376
+H605641201,2026-06-06,8482.9560546875,6639.9000000000015,1843.0560546874985,3377
+H605641201,2026-06-07,8485.0205078125,6639.9000000000015,1845.1205078124985,3378
+H605641201,2026-06-08,8562.81640625,6639.9000000000015,1922.9164062499985,3379
+H605641201,2026-06-09,8494.7890625,6639.9000000000015,1854.8890624999985,3380
+H605641101,2026-06-10,8594.818359375,8347.999999999998,246.81835937500182,4117
+H605641201,2026-06-10,8814.7265625,6639.9000000000015,2174.8265624999985,3381
+H605641201,2026-06-11,8311.599609375,6639.9000000000015,1671.6996093749985,3382
+H605641201,2026-06-12,8308.6337890625,6639.9000000000015,1668.7337890624985,3383
+H605641201,2026-06-13,8309.2421875,6639.9000000000015,1669.3421874999985,3384
+H605641201,2026-06-14,8310.923828125,6639.9000000000015,1671.0238281249985,3385
+H605641201,2026-06-15,8327.5322265625,6639.9000000000015,1687.6322265624985,3386
+H605641201,2026-06-16,8314.373046875,6639.9000000000015,1674.4730468749985,3387
+H605641201,2026-06-17,8314.5224609375,6639.9000000000015,1674.6224609374985,3388
+H605641201,2026-06-18,8416.5048828125,6639.9000000000015,1776.6048828124985,3389
+H605641201,2026-06-19,8343.890625,6639.9000000000015,1703.9906249999985,3390
+H605641201,2026-06-20,8359.11328125,6639.9000000000015,1719.2132812499985,3391
+H605641201,2026-06-21,8475.76171875,6639.9000000000015,1835.8617187499985,3392
+H605641201,2026-06-22,8315.84375,6639.9000000000015,1675.9437499999985,3393
+H605641201,2026-06-23,8317.3671875,6639.9000000000015,1677.4671874999985,3394
+H605641201,2026-06-24,8495.111328125,6639.9000000000015,1855.2113281249985,3395
+H605641201,2026-06-25,8408.4306640625,6639.9000000000015,1768.5306640624985,3396
+H605641101,2026-06-26,8381.41796875,8347.999999999998,33.41796875000182,4133
+H605641201,2026-06-26,8631.35546875,6639.9000000000015,1991.4554687499985,3397
+H605641201,2026-06-27,8321.4765625,6639.9000000000015,1681.5765624999985,3398
+H605641201,2026-06-28,8322.6640625,6639.9000000000015,1682.7640624999985,3399
+H605641101,2026-06-29,9004.697265625,8347.999999999998,656.6972656250018,4136
+H605641201,2026-06-29,9304.4755859375,6639.9000000000015,2664.5755859374985,3400
+H605641201,2026-06-30,8325.9609375,6639.9000000000015,1686.0609374999985,3401
+H605641201,2026-07-01,8327.5537109375,6639.9000000000015,1687.6537109374985,3402
+H605641201,2026-07-02,8327.8896484375,6639.9000000000015,1687.9896484374985,3403
+H605641201,2026-07-03,8329.3046875,6639.9000000000015,1689.4046874999985,3404
+H605641201,2026-07-04,8331.1982421875,6639.9000000000015,1691.2982421874985,3405
+H605641201,2026-07-05,8332.52734375,6639.9000000000015,1692.6273437499985,3406
+H605641201,2026-07-06,8333.033203125,6639.9000000000015,1693.1332031249985,3407
+H605641201,2026-07-07,8333.318359375,6639.9000000000015,1693.4183593749985,3408
+H605641201,2026-07-08,8333.6728515625,6639.9000000000015,1693.7728515624985,3409

datasets/meteofrance_precipitation/daily_precipitation.csv

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:707c2f5a7531bebb66e6c2c75711f47a24f5a567aa5897e86d5dc416f8386ffa
+size 11316886

datasets/spatiotemporal_gnn/model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:debcf20e5f2fa6309d3fccacb0b4631bbe6a319625698a6a03c06d195b1d3022
+size 223022