speed up by removing dask

components.py

@@ -1,4 +1,3 @@
-import dask.dataframe as dd
 import holoviews as hv
 import numpy as np
 import pandas as pd
@@ -66,9 +65,9 @@ class MapInput(pn.viewable.Viewer):
     @param.depends("data", watch=True, on_init=True)
     def _handle_data_dask_change(self):
         with self.param.update(updating=True):
-            data_dask = dd.from_pandas(self.data).persist()
+            data = self.data[["centre_easting", "centre_northing"]].copy()
             points = hv.Points(
-                data_dask, kdims=["centre_easting", "centre_northing"], vdims=[]
+                data, kdims=["centre_easting", "centre_northing"], vdims=[]
             )
 
             rangexy = hv.streams.RangeXY(source=points)
@@ -149,6 +148,7 @@ class ImageInput(pn.viewable.Viewer):
                     self.param._timestamp,
                     button_style="outline",
                     align="end",
+                    disabled=self.param.updating,
                 ),
                 pn.widgets.Select.from_param(
                     self.param.column_name, disabled=self.param.updating
@@ -157,7 +157,6 @@ class ImageInput(pn.viewable.Viewer):
             pn.Tabs(
                 pn.pane.HoloViews(
                     self.param.plot,
-                    loading=self.param.updating,
                     height=800,
                     width=800,
                     name="Interactive Image",
@@ -165,17 +164,16 @@ class ImageInput(pn.viewable.Viewer):
                 pn.pane.Image(
                     self.param.image,
                     name="Static Image",
-                    loading=self.param.updating,
                     width=800,
                 ),
                 pn.widgets.Tabulator(
                     self.param.meta_data,
                     name="Meta Data",
-                    loading=self.param.updating,
                     disabled=True,
                 ),
                 pn.pane.Markdown(self.code, name="Code"),
                 dynamic=True,
+                loading=self.param.updating,
             ),
         )
 
@@ -291,7 +289,10 @@ class App(param.Parameterized):
     def _create_sidebar(self):
         return pn.Column(
             pn.pane.Image(
-                MAJOR_TOM_LOGO, link_url=MAJOR_TOM_REF_URL, sizing_mode="stretch_width"
+                MAJOR_TOM_LOGO,
+                link_url=MAJOR_TOM_REF_URL,
+                height=60,
+                sizing_mode="stretch_width",
             ),
             pn.pane.Image(
                 MAJOR_TOM_PICTURE,

requirements.txt

@@ -1,5 +1,4 @@
 aiohttp
-dask[dataframe]
 datashader
 fsspec
 holoviews

utils.py

@@ -9,7 +9,7 @@ from fsspec.parquet import open_parquet_file
 from holoviews import opts
 from PIL import Image
 
-MAJOR_TOM_LOGO = "https://cdn-uploads.huggingface.co/production/uploads/6304c06eeb6d777a838eab63/BJKsLwX0GG4W3-gdf40TJ.png"
+MAJOR_TOM_LOGO = "assets/major-tom-esa-logo.png"
 MAJOR_TOM_PICTURE = (
     "https://upload.wikimedia.org/wikipedia/en/6/6d/Major_tom_space_oddity_video.JPG"
 )
@@ -135,6 +135,7 @@ def _meta_data_path(dataset="Core-S2L2A", repository=REPOSITORY):
 
 
 def get_meta_data(dataset="Core-S2L2A", repository=REPOSITORY):
+    print(f"Loading {dataset}")
     path = _meta_data_path(dataset=dataset)
     if not path.exists():
         data = pd.read_parquet(_meta_data_url(dataset=dataset))
@@ -146,6 +147,10 @@ def get_meta_data(dataset="Core-S2L2A", repository=REPOSITORY):
             data["centre_lon"], data["centre_lat"]
         )
     )
+    # Optimize Performance
+    data["timestamp"] = pd.to_datetime(data["timestamp"])
+    numeric_cols = ["cloud_cover", "nodata", "centre_lat", "centre_lon"]
+    data[numeric_cols] = data[numeric_cols].astype("float32")
 
     return data
 
@@ -167,17 +172,17 @@ def euclidean_distance(x, y, target_x, target_y):
     return np.sqrt((x - target_x) ** 2 + (y - target_y) ** 2)
 
 
-def get_closest_row(df, target_easting, target_northing):
+def get_closest_row(data, target_easting, target_northing):
     distance = euclidean_distance(
-        df["centre_easting"], df["centre_northing"], target_easting, target_northing
+        data["centre_easting"], data["centre_northing"], target_easting, target_northing
     )
-    closest_row = df.loc[distance.idxmin()]
+    closest_row = data.loc[distance.idxmin()]
     return closest_row
 
 
-def get_closest_rows(df, target_easting, target_northing):
+def get_closest_rows(data, target_easting, target_northing):
     distance = euclidean_distance(
-        df["centre_easting"], df["centre_northing"], target_easting, target_northing
+        data["centre_easting"], data["centre_northing"], target_easting, target_northing
     )
-    closest_rows = df[distance == distance.min()]
+    closest_rows = data[distance == distance.min()]
     return closest_rows