add min max

app.py

@@ -274,6 +274,12 @@ def plot_bars(
 
     return fig
 
+def get_desc(data):
+    res = {name: list(dt.to_dict().keys()) for name, dt in data.items()}
+    return "\n".join([
+        f"**{name}**: min={min(dt)}, max={max(dt)}" for name, dt in res.items()
+    ])
+
 
 def update_graph(
         base_folder,
@@ -307,7 +313,7 @@ def update_graph(
 
     data = {path: result for path, result in zip(datasets, data)}
     return plot_data(data, metric_name, normalization, rounding, grouping, top_k, direction, regex, log_scale_x,
-                     log_scale_y, progress), data, export_data(data, metric_name)
+                     log_scale_y, progress), data, export_data(data, metric_name), get_desc(data)
 
 
 def plot_data(data, metric_name, normalization, rounding, grouping, top_k, direction, regex, log_scale_x, log_scale_y,
@@ -413,6 +419,8 @@ The data might not be 100% representative, due to the sampling and optimistic me
             with gr.Row():
                 # export_data_button = gr.Button("Export data", visible=True, link=export_data_json)
                 export_data_json = gr.File(visible=False)
+        with gr.Column(visible=False) as min_max_hist:
+            min_max_hist_data = gr.Markdown()
         with gr.Column(scale=4):
             with gr.Row(visible=False) as group_choices:
                 with gr.Column(scale=2):
@@ -484,7 +492,7 @@ The data might not be 100% representative, due to the sampling and optimistic me
             direction_checkbox,
             group_regex,
         ],
-        outputs=[graph_output, exported_data, export_data_json],
+        outputs=[graph_output, exported_data, export_data_json, min_max_hist_data],
     )
 
     for inp in [normalization_checkbox, rounding, group_regex, direction_checkbox, top_select, log_scale_x_checkbox,
@@ -565,18 +573,20 @@ The data might not be 100% representative, due to the sampling and optimistic me
             return {
                 normalization_checkbox: gr.Column(visible=True),
                 group_choices: gr.Column(visible=False),
+                min_max_hist: gr.Column(visible=True),
             }
         else:
             return {
                 normalization_checkbox: gr.Column(visible=False),
                 group_choices: gr.Column(visible=True),
+                min_max_hist: gr.Column(visible=False),
             }
 
 
     grouping_dropdown.select(
         fn=update_grouping_options,
         inputs=[grouping_dropdown],
-        outputs=[normalization_checkbox, group_choices],
+        outputs=[normalization_checkbox, group_choices, min_max_hist],
     )
 
 # Launch the application