Spaces:
Restarting
Restarting
jasonshaoshun
commited on
Commit
·
0c85aa2
1
Parent(s):
5b3b90b
debug
Browse files
app.py
CHANGED
|
@@ -318,63 +318,131 @@ from src.about import TasksMib_Subgraph
|
|
| 318 |
# )
|
| 319 |
|
| 320 |
|
| 321 |
-
|
| 322 |
-
|
| 323 |
-
|
| 324 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 325 |
|
| 326 |
-
|
| 327 |
-
|
| 328 |
-
|
| 329 |
-
|
| 330 |
-
|
| 331 |
-
|
| 332 |
-
|
| 333 |
-
|
| 334 |
-
col_name = f"{task.value.benchmark}_{model}"
|
| 335 |
-
if col_name in dataframe.columns:
|
| 336 |
-
task_cols.append(col_name)
|
| 337 |
|
| 338 |
-
|
| 339 |
-
|
| 340 |
-
|
| 341 |
-
|
| 342 |
-
|
| 343 |
-
|
| 344 |
-
|
| 345 |
-
|
| 346 |
-
|
| 347 |
-
|
| 348 |
-
|
| 349 |
-
|
| 350 |
-
|
| 351 |
-
|
| 352 |
-
|
| 353 |
|
| 354 |
-
|
| 355 |
-
|
| 356 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 357 |
|
| 358 |
-
|
| 359 |
-
|
| 360 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 361 |
|
| 362 |
-
|
| 363 |
-
|
| 364 |
-
|
| 365 |
-
|
| 366 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 367 |
|
| 368 |
return Leaderboard(
|
| 369 |
value=dataframe,
|
| 370 |
datatype=[c.type for c in fields(AutoEvalColumn_mib_subgraph)],
|
| 371 |
select_columns=SelectColumns(
|
| 372 |
-
default_selection=
|
| 373 |
-
label="
|
| 374 |
),
|
| 375 |
search_columns=["Method"],
|
| 376 |
-
hide_columns=[
|
| 377 |
-
bool_checkboxgroup_label="Hide models",
|
| 378 |
interactive=False,
|
| 379 |
)
|
| 380 |
|
|
@@ -382,11 +450,6 @@ def init_leaderboard_mib_subgraph(dataframe, track):
|
|
| 382 |
|
| 383 |
|
| 384 |
|
| 385 |
-
|
| 386 |
-
|
| 387 |
-
|
| 388 |
-
|
| 389 |
-
|
| 390 |
def init_leaderboard_mib_causalgraph(dataframe, track):
|
| 391 |
# print("Debugging column issues:")
|
| 392 |
# print("\nActual DataFrame columns:")
|
|
|
|
| 318 |
# )
|
| 319 |
|
| 320 |
|
| 321 |
+
|
| 322 |
+
|
| 323 |
+
|
| 324 |
+
|
| 325 |
+
|
| 326 |
+
|
| 327 |
+
# def init_leaderboard_mib_subgraph(dataframe, track):
|
| 328 |
+
# """Initialize the subgraph leaderboard with verified task/model column selection"""
|
| 329 |
+
# if dataframe is None or dataframe.empty:
|
| 330 |
+
# raise ValueError("Leaderboard DataFrame is empty or None.")
|
| 331 |
+
|
| 332 |
+
# # First, let's identify which columns actually exist in our dataframe
|
| 333 |
+
# print("Available columns in dataframe:", dataframe.columns.tolist())
|
| 334 |
|
| 335 |
+
# # Create task selections based on TasksMib_Subgraph definition
|
| 336 |
+
# task_selections = []
|
| 337 |
+
# for task in TasksMib_Subgraph:
|
| 338 |
+
# task_cols = []
|
| 339 |
+
# for model in task.value.models:
|
| 340 |
+
# col_name = f"{task.value.benchmark}_{model}"
|
| 341 |
+
# if col_name in dataframe.columns:
|
| 342 |
+
# task_cols.append(col_name)
|
|
|
|
|
|
|
|
|
|
| 343 |
|
| 344 |
+
# if task_cols: # Only add tasks that have data
|
| 345 |
+
# print(f"Task {task.value.benchmark} has columns:", task_cols)
|
| 346 |
+
# task_selections.append(f"Task: {task.value.benchmark}")
|
| 347 |
+
|
| 348 |
+
# # Create model selections by checking which models appear in columns
|
| 349 |
+
# model_selections = []
|
| 350 |
+
# all_models = list(set(model for task in TasksMib_Subgraph for model in task.value.models))
|
| 351 |
+
|
| 352 |
+
# for model in all_models:
|
| 353 |
+
# model_cols = []
|
| 354 |
+
# for task in TasksMib_Subgraph:
|
| 355 |
+
# if model in task.value.models:
|
| 356 |
+
# col_name = f"{task.value.benchmark}_{model}"
|
| 357 |
+
# if col_name in dataframe.columns:
|
| 358 |
+
# model_cols.append(col_name)
|
| 359 |
|
| 360 |
+
# if model_cols: # Only add models that have data
|
| 361 |
+
# print(f"Model {model} has columns:", model_cols)
|
| 362 |
+
# model_selections.append(f"Model: {model}")
|
| 363 |
+
|
| 364 |
+
# # Combine all selections
|
| 365 |
+
# selections = task_selections + model_selections
|
| 366 |
+
# print("Final selection options:", selections)
|
| 367 |
+
|
| 368 |
+
# # Print DataFrame information
|
| 369 |
+
# print("\nDebugging DataFrame:")
|
| 370 |
+
# print("DataFrame columns:", dataframe.columns.tolist())
|
| 371 |
+
# print("DataFrame shape:", dataframe.shape)
|
| 372 |
+
# print("DataFrame head:\n", dataframe.head())
|
| 373 |
|
| 374 |
+
# return Leaderboard(
|
| 375 |
+
# value=dataframe,
|
| 376 |
+
# datatype=[c.type for c in fields(AutoEvalColumn_mib_subgraph)],
|
| 377 |
+
# select_columns=SelectColumns(
|
| 378 |
+
# default_selection=selections,
|
| 379 |
+
# label="Select Tasks or Models:"
|
| 380 |
+
# ),
|
| 381 |
+
# search_columns=["Method"],
|
| 382 |
+
# hide_columns=[c.name for c in fields(AutoEvalColumn_mib_subgraph) if c.hidden],
|
| 383 |
+
# bool_checkboxgroup_label="Hide models",
|
| 384 |
+
# interactive=False,
|
| 385 |
+
# )
|
| 386 |
|
| 387 |
+
|
| 388 |
+
|
| 389 |
+
def init_leaderboard_mib_subgraph(dataframe, track):
|
| 390 |
+
"""Initialize the subgraph leaderboard with benchmark and model filtering capabilities."""
|
| 391 |
+
if dataframe is None or dataframe.empty:
|
| 392 |
+
raise ValueError("Leaderboard DataFrame is empty or None.")
|
| 393 |
+
|
| 394 |
+
# Print DataFrame information for debugging
|
| 395 |
+
print("\nDebugging DataFrame columns:", dataframe.columns.tolist())
|
| 396 |
+
|
| 397 |
+
# Get result columns (excluding Method and Average)
|
| 398 |
+
result_columns = [col for col in dataframe.columns
|
| 399 |
+
if col not in ['Method', 'Average'] and '_' in col]
|
| 400 |
+
|
| 401 |
+
# Create benchmark and model selections
|
| 402 |
+
benchmarks = set()
|
| 403 |
+
models = set()
|
| 404 |
+
|
| 405 |
+
# Extract unique benchmarks and models from column names
|
| 406 |
+
for col in result_columns:
|
| 407 |
+
benchmark, model = col.split('_')
|
| 408 |
+
benchmarks.add(benchmark)
|
| 409 |
+
models.add(model)
|
| 410 |
+
|
| 411 |
+
# Create selection groups
|
| 412 |
+
benchmark_selections = {
|
| 413 |
+
# For each benchmark, store which columns should be shown
|
| 414 |
+
benchmark: [col for col in result_columns if col.startswith(f"{benchmark}_")]
|
| 415 |
+
for benchmark in benchmarks
|
| 416 |
+
}
|
| 417 |
+
|
| 418 |
+
model_selections = {
|
| 419 |
+
# For each model, store which columns should be shown
|
| 420 |
+
model: [col for col in result_columns if col.endswith(f"_{model}")]
|
| 421 |
+
for model in models
|
| 422 |
+
}
|
| 423 |
+
|
| 424 |
+
# Combine the selection mappings
|
| 425 |
+
selection_groups = {
|
| 426 |
+
**benchmark_selections,
|
| 427 |
+
**model_selections
|
| 428 |
+
}
|
| 429 |
+
|
| 430 |
+
print("\nDebugging Selection Groups:")
|
| 431 |
+
print("Benchmarks:", benchmark_selections.keys())
|
| 432 |
+
print("Models:", model_selections.keys())
|
| 433 |
+
|
| 434 |
+
# Convert keys to list for selection options
|
| 435 |
+
selection_options = list(selection_groups.keys())
|
| 436 |
|
| 437 |
return Leaderboard(
|
| 438 |
value=dataframe,
|
| 439 |
datatype=[c.type for c in fields(AutoEvalColumn_mib_subgraph)],
|
| 440 |
select_columns=SelectColumns(
|
| 441 |
+
default_selection=selection_options, # Show all options by default
|
| 442 |
+
label="Filter by Benchmark or Model:"
|
| 443 |
),
|
| 444 |
search_columns=["Method"],
|
| 445 |
+
hide_columns=[],
|
|
|
|
| 446 |
interactive=False,
|
| 447 |
)
|
| 448 |
|
|
|
|
| 450 |
|
| 451 |
|
| 452 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 453 |
def init_leaderboard_mib_causalgraph(dataframe, track):
|
| 454 |
# print("Debugging column issues:")
|
| 455 |
# print("\nActual DataFrame columns:")
|