Spaces:
Running
Running
File size: 8,318 Bytes
b81a86b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 | """
Data Loader — CSV → Engine Objects
====================================
Transform sample CSV ke SupplyNode, DemandNode, Kabupaten, dst.
Tim bisa pakai loader ini sebagai template saat integrate dengan data real
(PIHPS scraper, Bapanas API, dll).
Usage:
from sample_data.loader import load_all_sample_data
data = load_all_sample_data()
surplus_nodes = data["surplus"]
deficit_nodes = data["deficit"]
weather = data["weather"]
historical = data["historical_prices"]
"""
from __future__ import annotations
import csv
import os
from datetime import datetime
from typing import Dict, List, Tuple
from matching_engine.models import (
Commodity, DemandNode, Kabupaten, SupplyNode, Tier, WeatherForecast,
)
SAMPLE_DIR = os.path.dirname(os.path.abspath(__file__))
def load_kabupaten() -> Dict[str, Kabupaten]:
"""Load 38 kabupaten Jatim dari kabupaten_jatim.csv."""
path = os.path.join(SAMPLE_DIR, "kabupaten_jatim.csv")
out: Dict[str, Kabupaten] = {}
with open(path, encoding="utf-8") as f:
reader = csv.DictReader(f)
for row in reader:
tier = Tier.HIGH if row["tier"] == "TIER_1_HIGH" else Tier.MEDIUM
out[row["kab_id"]] = Kabupaten(
id=row["kab_id"],
nama=row["nama"],
latitude=float(row["latitude"]),
longitude=float(row["longitude"]),
ipm=float(row["ipm_2024"]),
tier=tier,
population=int(row["population_2024"]),
)
return out
def load_komoditas() -> Dict[str, Commodity]:
"""Load 19 komoditas dari komoditas_constraints.csv."""
path = os.path.join(SAMPLE_DIR, "komoditas_constraints.csv")
out: Dict[str, Commodity] = {}
with open(path, encoding="utf-8") as f:
reader = csv.DictReader(f)
for row in reader:
out[row["code"]] = Commodity(
code=row["code"],
nama=row["nama"],
max_distance_km=float(row["max_distance_km"]),
min_viable_tons=float(row["min_viable_tons"]),
max_fresh_age_days=int(row["max_fresh_age_days"]),
)
return out
def load_surplus_deficit(
kabupaten: Dict[str, Kabupaten],
komoditas: Dict[str, Commodity],
*,
csv_filename: str = "surplus_deficit.csv",
) -> Tuple[List[SupplyNode], List[DemandNode]]:
"""Load surplus & deficit dari surplus_deficit.csv (or a named override).
Args:
kabupaten: dict of Kabupaten objects keyed by kab_id.
komoditas: dict of Commodity objects keyed by code.
csv_filename: filename within SAMPLE_DIR to load.
Defaults to the canonical "surplus_deficit.csv".
Pass "surplus_deficit_constrained.csv" for the
La Nina supply-shock scenario.
"""
path = os.path.join(SAMPLE_DIR, csv_filename)
surplus: List[SupplyNode] = []
deficit: List[DemandNode] = []
now = datetime.now()
with open(path, encoding="utf-8") as f:
reader = csv.DictReader(f)
for row in reader:
kab = kabupaten[row["kab_id"]]
komo = komoditas[row["commodity_code"]]
if row["role"] == "SURPLUS":
surplus.append(SupplyNode(
kabupaten=kab, commodity=komo,
volume_tons=float(row["volume_tons"]),
price_per_kg=float(row["price_idr_per_kg"]),
harvest_age_days=int(row["harvest_age_days"]),
timestamp=now,
data_source="SAMPLE_CSV",
))
elif row["role"] == "DEFICIT":
deficit.append(DemandNode(
kabupaten=kab, commodity=komo,
volume_tons=float(row["volume_tons"]),
price_per_kg=float(row["price_idr_per_kg"]),
timestamp=now,
data_source="SAMPLE_CSV",
))
else:
raise ValueError(f"Unknown role: {row['role']}")
return surplus, deficit
def load_weather() -> Dict[str, WeatherForecast]:
"""
Load weather forecast dari weather_forecast.csv.
Returned dict keyed by '{origin_kab_id}_{dest_kab_id}'
untuk kompatibilitas dengan engine.run_matching.
"""
path = os.path.join(SAMPLE_DIR, "weather_forecast.csv")
out: Dict[str, WeatherForecast] = {}
with open(path, encoding="utf-8") as f:
reader = csv.DictReader(f)
for row in reader:
key = f"{row['origin_kab_id']}_{row['dest_kab_id']}"
out[key] = WeatherForecast(
origin_kab_id=row["origin_kab_id"],
dest_kab_id=row["dest_kab_id"],
max_rain_mm=float(row["max_rain_mm"]),
transit_window_days=int(row["transit_window_days"]),
source=row["source"],
)
return out
def load_historical_prices() -> Dict[str, Tuple[float, float]]:
"""
Load historical price stats dari historical_price_stats.csv.
Return dict commodity_code → (median, std).
"""
path = os.path.join(SAMPLE_DIR, "historical_price_stats.csv")
out: Dict[str, Tuple[float, float]] = {}
with open(path, encoding="utf-8") as f:
reader = csv.DictReader(f)
for row in reader:
out[row["commodity_code"]] = (
float(row["median_idr_per_kg"]),
float(row["std_idr_per_kg"]),
)
return out
def load_all_sample_data(*, surplus_deficit_csv: str = "surplus_deficit.csv"):
"""One-shot loader untuk semua sample data.
Args:
surplus_deficit_csv: filename within SAMPLE_DIR for supply/deficit rows.
Defaults to canonical "surplus_deficit.csv".
Pass "surplus_deficit_constrained.csv" for the
La Nina supply-shock scenario fixture.
Pass "surplus_deficit_real.csv" for BPS real data.
"""
kab = load_kabupaten()
komo = load_komoditas()
surplus, deficit = load_surplus_deficit(kab, komo, csv_filename=surplus_deficit_csv)
weather = load_weather()
historical = load_historical_prices()
return {
"kabupaten": kab,
"komoditas": komo,
"surplus": surplus,
"deficit": deficit,
"weather": weather,
"historical_prices": historical,
}
def load_real_data():
"""Load BPS Jawa Timur 2022 real data (6 komoditas: beras_premium, beras_medium,
cabai_merah, cabai_rawit, bawang_merah, bawang_putih).
Convenience wrapper around load_all_sample_data(surplus_deficit_csv=...).
Muat HANYA komoditas yang ada di surplus_deficit_real.csv — tidak ada
data sintetis di-merge.
The komoditas dict is filtered to match. Without that filter this function
returned all 19 rows of komoditas_constraints.csv, so /api/v1/commodities
advertised 13 commodities that have no real nodes behind them — a dropdown
where two thirds of the options render an empty map, each one backed by a
row marked SYNTHETIC in historical_price_stats.csv. Filtering here keeps
the promise the docstring above already made.
Returns same dict shape as load_all_sample_data().
"""
data = load_all_sample_data(surplus_deficit_csv="surplus_deficit_real.csv")
with_nodes = (
{n.commodity.code for n in data["surplus"]}
| {n.commodity.code for n in data["deficit"]}
)
data["komoditas"] = {
code: c for code, c in data["komoditas"].items() if code in with_nodes
}
return data
if __name__ == "__main__":
print("Testing sample data loader...")
data = load_all_sample_data()
print(f" Kabupaten: {len(data['kabupaten'])}")
print(f" Komoditas: {len(data['komoditas'])}")
print(f" Surplus: {len(data['surplus'])}")
print(f" Deficit: {len(data['deficit'])}")
print(f" Weather: {len(data['weather'])}")
print(f" Historical: {len(data['historical_prices'])}")
print("\nSample kabupaten (3 first):")
for kab_id in list(data["kabupaten"])[:3]:
k = data["kabupaten"][kab_id]
print(f" {kab_id} {k.nama:20s} IPM={k.ipm} tier={k.tier.value}")
|