Update app.py

app.py

@@ -27,35 +27,21 @@ def plot_scatter(cat, x, y, col):
     grouped_cat = data.groupby(["model", "tag"]).size().reset_index(name="count").sort_values(by="count", ascending=False)
     grouped_cat["count"] = grouped_cat.groupby(["model"])["count"].transform(lambda x: x / x.sum())
     
-    # minimal example
-    pivot_df = grouped_cat.pivot(index='model', columns='tag', values='count').fillna(0).reset_index()
-
+    # Pivot the data for stacking
+    pivot_df = grouped_cat.pivot(index='model', columns='tag', values='count').fillna(0)
+    # pivot_df = pivot_df.sort_values(by="A", ascending=False)
+    # add color vis
     if col == "Size":
-        pivot_df[col] = pivot_df["model"].map(size_map)
+        pivot_df[col] = pivot_df.index.map(size_map)
+        grouped_cat = grouped_cat.dropna(inplace=True)
     else:
-        pivot_df[col] = pivot_df["model"].str.split("/").str[0]
-    
-    fig = px.scatter_3d(pivot_df, x=x, y=y, z=z,
-        hover_name="model",
-        title=f'{x} vs {y} vs {z}',
-        color=col,
-        color_continuous_scale="agsunset")
-
-    # # Pivot the data for stacking
-    # pivot_df = grouped_cat.pivot(index='model', columns='tag', values='count').fillna(0)
-    # # pivot_df = pivot_df.sort_values(by="A", ascending=False)
-    # # add color vis
-    # if col == "Size":
-    #     pivot_df[col] = pivot_df.index.map(size_map)
-    #     grouped_cat = grouped_cat.dropna(inplace=True)
-    # else:
-    #     pivot_df[col] = pivot_df.index.str.split("/").str[0]
+        pivot_df[col] = pivot_df.index.str.split("/").str[0]
 
-    # # Create an interactive scatter plot
-    # fig = px.scatter(pivot_df, x=x, y=y, hover_name=pivot_df.index, title=f'{x} vs {y}', color=col, color_continuous_scale="agsunset")
+    # Create an interactive scatter plot
+    fig = px.scatter(pivot_df, x=x, y=y, hover_name=pivot_df.index, title=f'{x} vs {y}', color=col, color_continuous_scale="agsunset")
 
-    # # Show the plot
-    # return fig
+    # Show the plot
+    return fig
 
 # Tab 3
 def plot_scatter_tab3(subcat, col):
@@ -131,13 +117,6 @@ def plot_scatter_tab5(cat, x, y, z, col):
     grouped_cat = data.groupby(["model", "tag"]).size().reset_index(name="count").sort_values(by="count", ascending=False)
     grouped_cat["count"] = grouped_cat.groupby(["model"])["count"].transform(lambda x: x / x.sum())
 
-    # Pivot the data for stacking
-    # pivot_df = grouped_cat.pivot(index='model', columns='tag', values='count').fillna(0)
-    # if col == "Size":
-    #     pivot_df[col] = pivot_df.index.map(size_map)
-    # else:
-    #     pivot_df[col] = pivot_df.index.str.split("/").str[0]
-
     pivot_df = grouped_cat.pivot(index='model', columns='tag', values='count').fillna(0).reset_index()
 
     if col == "Size":
@@ -151,18 +130,10 @@ def plot_scatter_tab5(cat, x, y, z, col):
         color=col,
         color_continuous_scale="agsunset")
 
-
-    # Create an interactive scatter plot
-    # fig = px.scatter(pivot_df, x=x, y=y, hover_name=pivot_df.index, title=f'{x} vs {y}', color=col, color_continuous_scale="agsunset")
-    # fig = plt.figure() 
-    
-    # plot =  px.scatter_3d(pivot_df[x], pivot_df[y], pivot_df[z]) #c=pivot_df[col], cmap='viridis')
     print(pivot_df)
     
-    # fig = px.scatter_3d(pivot_df, x=x, y=y,z=z, hover_name=pivot_df.index, title=f'{x} vs {y} vs {z}', color=col, color_continuous_scale="agsunset")
     return fig
 
-
 # Tab 6
 data_with_text = pd.read_csv("./tagged_data_with_text.csv")
 def random_sample(r: gr.Request):