more fixes

app.py

@@ -12,7 +12,6 @@ source_finders = []
 questions_dict = {}
 source_finders_dict = {}
 question_options = []
-baseline_rankers = []
 baseline_rankers_dict = {}
 baseline_ranker_options = []
 run_ids = []
@@ -26,43 +25,44 @@ run_id_dropdown = None
 
 # Initialize data in a single async function
 async def initialize_data():
-    global questions, source_finders, questions_dict, source_finders_dict, question_options, finder_options, finder_labels, baseline_rankers, baseline_ranker_options
+    global questions, source_finders, questions_dict, source_finders_dict, question_options, finder_options, finder_labels, baseline_rankers_dict, source_finders_dict, baseline_ranker_options
 
     questions = await get_questions()
     source_finders = await get_source_finders()
 
     baseline_rankers = await get_baseline_rankers()
-    baseline_rankers_dict = {f["id"]: f["name"] for f in baseline_rankers}
+    baseline_rankers_dict = {f["name"]: f["id"] for f in baseline_rankers}
 
     # Convert to dictionaries for easier lookup
-    questions_dict = {q["id"]: q["text"] for q in questions}
-    source_finders_dict = {f["id"]: f["name"] for f in source_finders}
+    questions_dict = {q["text"]: q["id"] for q in questions}
+    baseline_rankers_dict = {f["name"]: f["id"] for f in baseline_rankers}
+    source_finders_dict = {f["name"]: f["id"] for f in source_finders}
 
     # Create formatted options for dropdowns
-    question_options = [f"{q['id']}: {q['text']}" for q in questions]
-    finder_options = [str(f["id"]) for f in source_finders]
+    question_options = [q['text'] for q in questions]
+    finder_options = [s["name"] for s in source_finders]
     finder_labels = {str(f["id"]): f["name"] for f in source_finders}
-    baseline_ranker_options = [f["id"] for f in baseline_rankers]
-    baseline_ranker_labels = {str(f["id"]): f["name"] for f in source_finders}
+    baseline_ranker_options = [b["name"] for b in baseline_rankers]
 
 
-def update_sources_list(question_option, source_finder_id, baseline_ranker_id: str, run_id: str, evt: gr.EventData = None):
+
+def update_sources_list(question_option, source_finder_id, run_id: str, baseline_ranker_id: str, evt: gr.EventData = None):
     global previous_run_id
-    if evt and evt.target and evt.target.elem_id == "run_id_dropdown":
-        if run_id == previous_run_id:
+    if evt.target.elem_id == "run_id_dropdown":
+        if type(run_id) == list or run_id == previous_run_id:
             return gr.skip(), gr.skip(), gr.skip(), gr.skip()
-        # Store the current run_id for future comparison
+
     previous_run_id = run_id
-    return asyncio.run(update_sources_list_async(question_option, source_finder_id, baseline_ranker_id, run_id))
+    return asyncio.run(update_sources_list_async(question_option, source_finder_id, run_id, baseline_ranker_id))
 
 
 # Main function to handle UI interactions
-async def update_sources_list_async(question_option, source_finder_id, baseline_ranker_id: str, run_id: str):
+async def update_sources_list_async(question_option, source_finder_name, run_id:str, baseline_ranker_name: str):
     if not question_option:
         return gr.skip(), gr.skip(), gr.skip(), "No question selected"
 
     # Extract question ID from selection
-    question_id = int(question_option.split(":")[0])
+    question_id = questions_dict.get(question_option)
 
     available_run_ids = await get_run_ids(question_id)
     run_id_options = [str(r_id) for r_id in available_run_ids]
@@ -70,10 +70,15 @@ async def update_sources_list_async(question_option, source_finder_id, baseline_
         run_id = run_id_options[0]
 
     run_id_int = int(run_id)
-    finder_id_int = None if len(source_finder_id) == 0 else int(source_finder_id)
-    if type(baseline_ranker_id) == list:
-        baseline_ranker_id = baseline_ranker_id[0]
-    baseline_ranker_id_int = 1 if len(baseline_ranker_id) == 0 else int(baseline_ranker_id)
+    if len(source_finder_name):
+        finder_id_int = source_finders_dict.get(source_finder_name)
+    else:
+        finder_id_int = None
+
+    if type(baseline_ranker_name) == list:
+        baseline_ranker_name = baseline_ranker_name[0]
+
+    baseline_ranker_id_int = 1 if len(baseline_ranker_name) == 0 else baseline_rankers_dict.get(baseline_ranker_name)
 
     source_runs = None
     stats = None
@@ -144,31 +149,6 @@ async def main():
                             interactive=True,
                             elem_id="run_id_dropdown"
                         )
-
-                result_text = gr.Markdown("Select a question to view source runs")
-                gr.Markdown("# Source Run Statistics")
-                statistics_table = gr.DataFrame(
-                    headers=["num_high_ranked_baseline_sources",
-                             "num_high_ranked_found_sources",
-                             "overlap_count",
-                             "overlap_percentage",
-                             "high_ranked_overlap_count",
-                             "high_ranked_overlap_percentage"
-                             ],
-                    interactive=False,
-                )
-                gr.Markdown("# Sources Found")
-                results_table = gr.DataFrame(
-                    headers=['Source Finder', 'Run ID', 'Sugya ID', 'Tractate', 'Folio', 'Rank', 'Reason'],
-                    interactive=False
-                )
-
-                # download_button = gr.DownloadButton(
-                #     label="Download Results as CSV",
-                #     interactive=True,
-                #     visible=True
-                # )
-
             with gr.Column(scale=1):
                 # Sidebar area
                 gr.Markdown("### About")
@@ -179,12 +159,37 @@ async def main():
                 gr.Markdown(f"Total Questions: {len(questions)}")
                 gr.Markdown(f"Source Finders: {len(source_finders)}")
 
-                gr.Markdown("### Source Finders")
-                for f in source_finders:
-                    gr.Markdown(f"**{f['id']}**: {f['name']}")
-                gr.Markdown("### Baseline Source Rankers")
-                for f in baseline_rankers:
-                    gr.Markdown(f"**{f['id']}**: {f['name']}")
+
+        with gr.Row():
+            result_text = gr.Markdown("Select a question to view source runs")
+        with gr.Row():
+            gr.Markdown("# Source Run Statistics")
+        with gr.Row():
+            statistics_table = gr.DataFrame(
+                headers=["num_high_ranked_baseline_sources",
+                         "num_high_ranked_found_sources",
+                         "overlap_count",
+                         "overlap_percentage",
+                         "high_ranked_overlap_count",
+                         "high_ranked_overlap_percentage"
+                         ],
+                interactive=False,
+            )
+        with gr.Row():
+            gr.Markdown("# Sources Found")
+        with gr.Row():
+            results_table = gr.DataFrame(
+                headers=['Source Finder', 'Run ID', 'Sugya ID', 'Tractate', 'Folio', 'Rank', 'Reason'],
+                interactive=False
+            )
+
+            # download_button = gr.DownloadButton(
+            #     label="Download Results as CSV",
+            #     interactive=True,
+            #     visible=True
+            # )
+
+
 
         # Set up event handlers
         question_dropdown.change(