fix: fixed a bug where sizes weren't scaled on first render

src/data_hub/plotting.py

@@ -4,7 +4,7 @@
 from beartype.typing import Optional
 from bokeh.io import show
 from bokeh.layouts import column, row
-from bokeh.models import ColumnDataSource, CustomJS, HoverTool, Select, TapTool
+from bokeh.models import ColumnDataSource, CustomJS, HoverTool, Select
 from bokeh.plotting import figure
 from deeporigin.data_hub.dataframe import DataFrame
 from deeporigin.exceptions import DeepOriginException
@@ -17,13 +17,21 @@ def scatter(
     y: Optional[str] = None,
     size: Optional[str] = None,
 ):
-    """function to make a scatter plot from a Deep Origin dataframe, with support for interactivity"""
+    """function to make a scatter plot from a Deep Origin dataframe, with support for interactivity
+
+    Args:
+        df (DataFrame): Deep Origin dataframe
+        x (Optional[str], optional): name of column to use for x axis. Defaults to None.
+        y (Optional[str], optional): name of column to use for y axis. Defaults to None.
+        size (Optional[str], optional): name of column to use for size. Defaults to None.
+
+    `df` should be a Deep Origin dataframe with at least two numeric columns.
+
+    """
 
     figure_width = 500
     select_width = int(figure_width * 0.3)
-
-    if df.shape[1] < 2:
-        raise ValueError("DataFrame must contain at least two columns.")
+    js_code = _read_js_code()
 
     cols = df.attrs["metadata"]["cols"]
     cols = [col["name"] for col in cols if col["type"] in ["float", "integer"]]
@@ -40,16 +48,18 @@ def scatter(
     if size is None:
         size = cols[0]
 
-    # Set up initial data for scatter plot
-    initial_x = x
-    initial_y = y
-    initial_size = size
+    # normalize sizes. this should match what's in
+    # axes_callback.js
+    sizes = list(df[size])
+    min_size = min(sizes)
+    max_size = max(sizes)
+    sizes = [2 + 15 * (value - min_size) / (max_size - min_size) for value in sizes]
 
     # CDS for scatter data
     data = dict(
-        x=list(df[initial_x]),
-        y=list(df[initial_y]),
-        size=list(df[initial_size]),
+        x=list(df[x]),
+        y=list(df[y]),
+        size=sizes,
     )
 
     scatter_source = ColumnDataSource(data)
@@ -86,34 +96,31 @@ def scatter(
     )
 
     # Set initial axis labels
-    p.xaxis.axis_label = initial_x
-    p.yaxis.axis_label = initial_y
+    p.xaxis.axis_label = x
+    p.yaxis.axis_label = y
 
     # Create dropdown selectors for X and Y axes
     x_select = Select(
         title="X-Axis",
-        value=initial_x,
+        value=x,
         options=cols,
         width=select_width,
     )
     y_select = Select(
         title="Y-Axis",
-        value=initial_y,
+        value=y,
         options=cols,
         width=select_width,
     )
     size_select = Select(
         title="Size",
-        value=initial_size,
+        value=size,
         options=cols,
         width=select_width,
     )
 
     # JavaScript callback to update data, axis labels, and point sizes on select change
 
-    with importlib.resources.open_text("deeporigin.data_hub", "axes_callback.js") as f:
-        axes_callback_js = f.read()
-
     axes_callback = CustomJS(
         args=dict(
             source=scatter_source,
@@ -124,18 +131,15 @@ def scatter(
             x_axis=p.xaxis[0],
             y_axis=p.yaxis[0],
         ),
-        code=axes_callback_js,
+        code=js_code["axes_callback"],
     )
 
     # JS code, will run in browser
     # this updates the value of the slider to the currently
     # hovered point
-    code = """
-    
-        """
 
     callback = CustomJS(
-        code=code,
+        code=js_code["hover_callback"],
         args=dict(
             marker_source=marker_source,
             scatter_source=scatter_source,