Lighter App

app.py

@@ -487,8 +487,6 @@ def compute_global_regression(df_combined, embedding_cols, tsne_params, df_f1, r
                          perplexity=tsne_params["perplexity"], 
                          learning_rate=tsne_params["learning_rate"])
     
-    nan_rows = df_combined[df_combined[embedding_cols].isnull().any(axis=1)]
-    print(nan_rows)
     reduced = reducer.fit_transform(df_combined[embedding_cols].values)
     # Guardamos el embedding completo (por ejemplo, 4 dimensiones en PCA)
     df_combined['embedding'] = list(reduced)
@@ -1165,83 +1163,83 @@ def run_model(model_name):
 
 
 
-        real_labels_new = sorted(df_all["real"]['label'].unique().tolist())
-        df_distances_new = compute_cluster_distances_synthetic_individual(
-            df_all["synthetic"],
-            df_all["real"],
-            real_labels_new,
-            metric="wasserstein",  # O la métrica que prefieras
-            bins=20
-        )
+        # real_labels_new = sorted(df_all["real"]['label'].unique().tolist())
+        # df_distances_new = compute_cluster_distances_synthetic_individual(
+        #     df_all["synthetic"],
+        #     df_all["real"],
+        #     real_labels_new,
+        #     metric="wasserstein",  # O la métrica que prefieras
+        #     bins=20
+        # )
         
-        global_distances_new = {}
-        for idx in df_distances_new.index:
-            if idx.startswith("Global"):
-                source_name = idx.split("(")[1].rstrip(")")
-                global_distances_new[source_name] = df_distances_new.loc[idx].values
+        # global_distances_new = {}
+        # for idx in df_distances_new.index:
+        #     if idx.startswith("Global"):
+        #         source_name = idx.split("(")[1].rstrip(")")
+        #         global_distances_new[source_name] = df_distances_new.loc[idx].values
         
-        all_x_new = []
-        all_y_new = []
-        for source in df_f1.columns:
-            if source in global_distances_new:
-                x_vals = global_distances_new[source]
-                y_vals = df_f1[source].values
-                all_x_new.extend(x_vals)
-                all_y_new.extend(y_vals)
-        all_x_arr_new = np.array(all_x_new).reshape(-1, 1)
-        all_y_arr_new = np.array(all_y_new)
+        # all_x_new = []
+        # all_y_new = []
+        # for source in df_f1.columns:
+        #     if source in global_distances_new:
+        #         x_vals = global_distances_new[source]
+        #         y_vals = df_f1[source].values
+        #         all_x_new.extend(x_vals)
+        #         all_y_new.extend(y_vals)
+        # all_x_arr_new = np.array(all_x_new).reshape(-1, 1)
+        # all_y_arr_new = np.array(all_y_new)
         
-        model_global_new = LinearRegression().fit(all_x_arr_new, all_y_arr_new)
-        r2_new = model_global_new.score(all_x_arr_new, all_y_arr_new)
-        slope_new = model_global_new.coef_[0]
-        intercept_new = model_global_new.intercept_
+        # model_global_new = LinearRegression().fit(all_x_arr_new, all_y_arr_new)
+        # r2_new = model_global_new.score(all_x_arr_new, all_y_arr_new)
+        # slope_new = model_global_new.coef_[0]
+        # intercept_new = model_global_new.intercept_
         
-        scatter_fig_new = figure(
-            width=600,
-            height=600,
-            tools="pan,wheel_zoom,reset,save,hover",
-            active_scroll="wheel_zoom",
-            title="Scatter Plot: Distance vs F1 (Nueva PCA)",
-            background_fill_color="white",
-            y_range=(0, 1)
-        )
-        scatter_fig_new.xgrid.grid_line_color = None
-        scatter_fig_new.ygrid.grid_line_color = "gray"
-        scatter_fig_new.match_aspect = True
+        # scatter_fig_new = figure(
+        #     width=600,
+        #     height=600,
+        #     tools="pan,wheel_zoom,reset,save,hover",
+        #     active_scroll="wheel_zoom",
+        #     title="Scatter Plot: Distance vs F1 (Nueva PCA)",
+        #     background_fill_color="white",
+        #     y_range=(0, 1)
+        # )
+        # scatter_fig_new.xgrid.grid_line_color = None
+        # scatter_fig_new.ygrid.grid_line_color = "gray"
+        # scatter_fig_new.match_aspect = True
         
-        source_colors = {
-            "es-digital-paragraph-degradation-seq": "blue",
-            "es-digital-line-degradation-seq": "green",
-            "es-digital-seq": "red",
-            "es-digital-zoom-degradation-seq": "orange",
-            "es-digital-rotation-degradation-seq": "purple",
-            "es-digital-rotation-zoom-degradation-seq": "brown",
-            "es-render-seq": "cyan"
-        }
+        # source_colors = {
+        #     "es-digital-paragraph-degradation-seq": "blue",
+        #     "es-digital-line-degradation-seq": "green",
+        #     "es-digital-seq": "red",
+        #     "es-digital-zoom-degradation-seq": "orange",
+        #     "es-digital-rotation-degradation-seq": "purple",
+        #     "es-digital-rotation-zoom-degradation-seq": "brown",
+        #     "es-render-seq": "cyan"
+        # }
         
-        for source in df_f1.columns:
-            if source in global_distances_new:
-                x_vals = global_distances_new[source]
-                y_vals = df_f1[source].values
-                data = {"x": x_vals, "y": y_vals, "Fuente": [source]*len(x_vals)}
-                cds = ColumnDataSource(data=data)
-                scatter_fig_new.circle(
-                    'x', 'y', size=8, alpha=0.7, source=cds,
-                    fill_color=source_colors.get(source, "gray"),
-                    line_color=source_colors.get(source, "gray"),
-                    legend_label=source
-                )
+        # for source in df_f1.columns:
+        #     if source in global_distances_new:
+        #         x_vals = global_distances_new[source]
+        #         y_vals = df_f1[source].values
+        #         data = {"x": x_vals, "y": y_vals, "Fuente": [source]*len(x_vals)}
+        #         cds = ColumnDataSource(data=data)
+        #         scatter_fig_new.circle(
+        #             'x', 'y', size=8, alpha=0.7, source=cds,
+        #             fill_color=source_colors.get(source, "gray"),
+        #             line_color=source_colors.get(source, "gray"),
+        #             legend_label=source
+        #         )
         
-        scatter_fig_new.xaxis.axis_label = "Distance (Global, por Colegio) - Nueva PCA"
-        scatter_fig_new.yaxis.axis_label = "F1 Score"
-        scatter_fig_new.legend.location = "top_right"
-        hover_tool_new = scatter_fig_new.select_one(HoverTool)
-        hover_tool_new.tooltips = [("Distance", "@x"), ("F1", "@y"), ("Subset", "@Fuente")]
-        x_line_new = np.linspace(all_x_arr_new.min(), all_x_arr_new.max(), 100)
-        y_line_new = model_global_new.predict(x_line_new.reshape(-1,1))
-        scatter_fig_new.line(x_line_new, y_line_new, line_width=2, line_color="black", legend_label="Global Regression")
-        st.bokeh_chart(scatter_fig_new)
-        st.write(f"Regresión global (Nueva PCA): R² = {r2_new:.4f}, Slope = {slope_new:.4f}, Intercept = {intercept_new:.4f}")
+        # scatter_fig_new.xaxis.axis_label = "Distance (Global, por Colegio) - Nueva PCA"
+        # scatter_fig_new.yaxis.axis_label = "F1 Score"
+        # scatter_fig_new.legend.location = "top_right"
+        # hover_tool_new = scatter_fig_new.select_one(HoverTool)
+        # hover_tool_new.tooltips = [("Distance", "@x"), ("F1", "@y"), ("Subset", "@Fuente")]
+        # x_line_new = np.linspace(all_x_arr_new.min(), all_x_arr_new.max(), 100)
+        # y_line_new = model_global_new.predict(x_line_new.reshape(-1,1))
+        # scatter_fig_new.line(x_line_new, y_line_new, line_width=2, line_color="black", legend_label="Global Regression")
+        # st.bokeh_chart(scatter_fig_new)
+        # st.write(f"Regresión global (Nueva PCA): R² = {r2_new:.4f}, Slope = {slope_new:.4f}, Intercept = {intercept_new:.4f}")
         
         # -------------------------------------------------------------------------
         # 5. BLOQUE: Heatmap de Características