improve map plot

apps/kpi_analysis/gsm_capacity.py

@@ -5,6 +5,7 @@ import streamlit as st
 from process_kpi.process_gsm_capacity import analyze_gsm_data
 from utils.convert_to_excel import (  # Import convert_dfs from the appropriate module
     convert_gsm_dfs,
+    save_dataframe,
 )
 from utils.kpi_analysis_utils import GsmCapacity
 
@@ -242,25 +243,78 @@ if (
             )
             st.plotly_chart(fig, use_container_width=True)
 
-        # create a map plotly with gsm_analysis_df and max_tch_call_blocking_bh
+        # create a map plot with scatter_map with gsm_analysis_df and max_tch_call_blocking_bh
         st.markdown("***")
         st.markdown(":blue[**Max TCH Call Blocking BH distribution**]")
-        fig = px.scatter_mapbox(
-            gsm_analysis_df.dropna(
-                subset=["max_tch_call_blocking_bh", "Latitude", "Longitude"]
-            ),
+
+        # Select and clean the necessary columns
+        map_df = gsm_analysis_df[
+            ["code", "max_tch_call_blocking_bh", "Latitude", "Longitude"]
+        ].dropna(subset=["code", "max_tch_call_blocking_bh", "Latitude", "Longitude"])
+
+        # Group by code and max max_tch_call_blocking_bh, keep first occurrence of other columns
+        map_df = (
+            map_df.groupby("code")
+            .agg(
+                {
+                    "max_tch_call_blocking_bh": "max",
+                    "Latitude": "first",
+                    "Longitude": "first",
+                }
+            )
+            .reset_index()
+        )
+        # save_dataframe(map_df, "max_tch_call_blocking_bh_map")
+        # Create a color column based on the threshold
+        map_df["color"] = map_df["max_tch_call_blocking_bh"].apply(
+            lambda x: (
+                "Above Threshold" if x > tch_blocking_threshold else "Below Threshold"
+            )
+        )
+
+        # Apply minimum size to make small values more visible
+        min_bubble_size = 5  # Minimum size for visibility
+        max_bubble_size = 30  # Maximum size for scaling
+
+        # Scale the size to make small values more visible while maintaining relative sizes
+        size_scale = max_bubble_size / map_df["max_tch_call_blocking_bh"].max()
+        map_df["scaled_size"] = map_df["max_tch_call_blocking_bh"].apply(
+            lambda x: max(x * size_scale, min_bubble_size)
+        )
+
+        fig = px.scatter_map(
+            map_df,
             lat="Latitude",
             lon="Longitude",
-            color=[
-                "red" if val > tch_blocking_threshold else "green"
-                for val in gsm_analysis_df[
-                    "max_tch_call_blocking_bh"
-                ].dropna()  # .values
-            ],
-            size="max_tch_call_blocking_bh",
+            color="color",
+            color_discrete_map={"Above Threshold": "red", "Below Threshold": "green"},
+            size="scaled_size",
+            size_max=max_bubble_size,
+            hover_data={
+                "code": True,  # Show code in hover data
+                "max_tch_call_blocking_bh": ":.2f",
+                "scaled_size": False,
+            },
+            hover_name="code",  # This will show as the title of the hover box
             zoom=10,
             height=600,
             title="Max TCH Call Blocking BH distribution",
         )
-        fig.update_layout(mapbox_style="open-street-map")
+
+        # Update traces to show code on bubbles and customize hover
+        fig.update_traces(
+            text=map_df["code"],  # Show code on the bubble
+            textposition="middle center",
+            textfont=dict(size=18, color="black"),
+            # hovertemplate="<b>%{hovertext}</b><br>"
+            # + "Blocking: %{customdata[1]:.2f}%<extra></extra>",
+        )
+
+        # Adjust layout for better text visibility
+        fig.update_layout(
+            mapbox_style="open-street-map",
+            showlegend=True,
+            margin=dict(l=10, r=10, t=40, b=10),
+        )
+
         st.plotly_chart(fig, use_container_width=True)