Spaces:
Running
Running
File size: 18,633 Bytes
cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 ce626d3 cfbcff1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 |
# src/plotting.py
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import matplotlib.colors as mcolors
from colorsys import rgb_to_hls, hls_to_rgb
import plotly.graph_objects as go
import plotly.express as px
from plotly.subplots import make_subplots
import pandas as pd
import numpy as np
from collections import defaultdict
from typing import Dict, List, Optional, Union
from config import LANGUAGE_NAMES, ALL_UG40_LANGUAGES, GOOGLE_SUPPORTED_LANGUAGES, METRICS_CONFIG
plt.style.use('default')
plt.rcParams['figure.facecolor'] = 'white'
plt.rcParams['axes.facecolor'] = 'white'
def create_leaderboard_ranking_plot(df: pd.DataFrame, metric: str = 'quality_score', top_n: int = 15) -> go.Figure:
"""Create interactive leaderboard ranking plot using Plotly."""
if df.empty:
fig = go.Figure()
fig.add_annotation(
text="No data available",
xref="paper", yref="paper",
x=0.5, y=0.5, showarrow=False,
font=dict(size=16)
)
fig.update_layout(title="No Data Available")
return fig
# Get top N models
top_models = df.head(top_n)
# Create horizontal bar chart
fig = go.Figure(data=[
go.Bar(
y=top_models['model_name'],
x=top_models[metric],
orientation='h',
marker=dict(
color=top_models[metric],
colorscale='Viridis',
showscale=True,
colorbar=dict(title=metric.replace('_', ' ').title())
),
text=[f"{score:.3f}" for score in top_models[metric]],
textposition='auto',
hovertemplate=(
"<b>%{y}</b><br>" +
f"{metric.replace('_', ' ').title()}: %{{x:.4f}}<br>" +
"Author: %{customdata[0]}<br>" +
"Coverage: %{customdata[1]:.1%}<br>" +
"<extra></extra>"
),
customdata=list(zip(top_models['author'], top_models['coverage_rate']))
)
])
fig.update_layout(
title=f"π SALT Translation Leaderboard - {metric.replace('_', ' ').title()}",
xaxis_title=f"{metric.replace('_', ' ').title()} Score",
yaxis_title="Models",
height=max(400, len(top_models) * 30 + 100),
margin=dict(l=20, r=20, t=60, b=20),
plot_bgcolor='white',
paper_bgcolor='white'
)
# Reverse y-axis to show best model at top
fig.update_yaxes(autorange="reversed")
return fig
def create_metrics_comparison_plot(df: pd.DataFrame, models: List[str] = None, max_models: int = 8) -> go.Figure:
"""Create radar chart comparing multiple metrics across models."""
if df.empty:
fig = go.Figure()
fig.add_annotation(text="No data available", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title="No Data Available")
return fig
# Select models to compare
if models is None:
selected_models = df.head(max_models)
else:
selected_models = df[df['model_name'].isin(models)].head(max_models)
if len(selected_models) == 0:
fig = go.Figure()
fig.add_annotation(text="No models found", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title="No Models Found")
return fig
# Metrics to include in radar chart
metrics = ['quality_score', 'bleu', 'chrf', 'rouge1', 'rougeL']
metric_labels = ['Quality Score', 'BLEU (/100)', 'ChrF', 'ROUGE-1', 'ROUGE-L']
fig = go.Figure()
colors = px.colors.qualitative.Set1[:len(selected_models)]
for i, (_, model) in enumerate(selected_models.iterrows()):
# Normalize BLEU to 0-1 scale for radar chart
values = []
for metric in metrics:
value = model[metric]
if metric == 'bleu':
value = value / 100.0 # Normalize BLEU
values.append(value)
# Close the radar chart
values += values[:1]
metric_labels_closed = metric_labels + [metric_labels[0]]
fig.add_trace(go.Scatterpolar(
r=values,
theta=metric_labels_closed,
fill='toself',
name=model['model_name'],
line_color=colors[i % len(colors)],
fillcolor=colors[i % len(colors)],
opacity=0.6
))
fig.update_layout(
polar=dict(
radialaxis=dict(
visible=True,
range=[0, 1]
)
),
showlegend=True,
title="π Multi-Metric Model Comparison",
height=600
)
return fig
def create_language_pair_heatmap(results_dict: Dict, metric: str = 'quality_score') -> go.Figure:
"""Create heatmap showing performance across language pairs."""
if not results_dict or 'pair_metrics' not in results_dict:
fig = go.Figure()
fig.add_annotation(text="No language pair data available", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title="No Language Pair Data Available")
return fig
pair_metrics = results_dict['pair_metrics']
# Create matrix for heatmap
languages = ALL_UG40_LANGUAGES
matrix = np.zeros((len(languages), len(languages)))
for i, src_lang in enumerate(languages):
for j, tgt_lang in enumerate(languages):
if src_lang != tgt_lang:
pair_key = f"{src_lang}_to_{tgt_lang}"
if pair_key in pair_metrics and metric in pair_metrics[pair_key]:
matrix[i, j] = pair_metrics[pair_key][metric]
else:
matrix[i, j] = np.nan
else:
matrix[i, j] = np.nan
# Create language labels
lang_labels = [LANGUAGE_NAMES.get(lang, lang) for lang in languages]
fig = go.Figure(data=go.Heatmap(
z=matrix,
x=lang_labels,
y=lang_labels,
colorscale='Viridis',
showscale=True,
colorbar=dict(title=metric.replace('_', ' ').title()),
hovertemplate=(
"Source: %{y}<br>" +
"Target: %{x}<br>" +
f"{metric.replace('_', ' ').title()}: %{{z:.3f}}<br>" +
"<extra></extra>"
)
))
fig.update_layout(
title=f"πΊοΈ Language Pair Performance - {metric.replace('_', ' ').title()}",
xaxis_title="Target Language",
yaxis_title="Source Language",
height=600,
width=700
)
return fig
def create_coverage_analysis_plot(df: pd.DataFrame) -> go.Figure:
"""Create plot analyzing test set coverage across submissions."""
if df.empty:
fig = go.Figure()
fig.add_annotation(text="No data available", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title="No Data Available")
return fig
fig = make_subplots(
rows=2, cols=2,
subplot_titles=(
"Coverage Distribution",
"Language Pairs Covered",
"Sample Count vs Quality",
"Google Comparable Coverage"
),
specs=[[{"type": "bar"}, {"type": "scatter"}],
[{"type": "scatter"}, {"type": "bar"}]]
)
# Coverage distribution
coverage_bins = pd.cut(df['coverage_rate'],
bins=[0, 0.5, 0.8, 0.9, 0.95, 1.0],
labels=['<50%', '50-80%', '80-90%', '90-95%', '95-100%'])
coverage_counts = coverage_bins.value_counts()
fig.add_trace(
go.Bar(x=coverage_counts.index, y=coverage_counts.values, name="Coverage"),
row=1, col=1
)
# Language pairs covered vs quality
fig.add_trace(
go.Scatter(
x=df['language_pairs_covered'],
y=df['quality_score'],
mode='markers',
text=df['model_name'],
name="Quality vs Coverage"
),
row=1, col=2
)
# Sample count vs quality
fig.add_trace(
go.Scatter(
x=df['total_samples'],
y=df['quality_score'],
mode='markers',
text=df['model_name'],
name="Quality vs Samples"
),
row=2, col=1
)
# Google comparable coverage
google_coverage = df['google_pairs_covered'].value_counts().sort_index()
fig.add_trace(
go.Bar(x=google_coverage.index, y=google_coverage.values, name="Google Coverage"),
row=2, col=2
)
fig.update_layout(
title="π Test Set Coverage Analysis",
height=800,
showlegend=False
)
return fig
def create_model_performance_timeline(df: pd.DataFrame) -> go.Figure:
"""Create timeline showing model performance over time."""
if df.empty:
fig = go.Figure()
fig.add_annotation(text="No data available", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title="No Data Available")
return fig
# Convert submission_date to datetime
df_copy = df.copy()
df_copy['submission_date'] = pd.to_datetime(df_copy['submission_date'])
df_copy = df_copy.sort_values('submission_date')
fig = go.Figure()
# Add scatter plot for each submission
fig.add_trace(go.Scatter(
x=df_copy['submission_date'],
y=df_copy['quality_score'],
mode='markers+lines',
marker=dict(
size=10,
color=df_copy['quality_score'],
colorscale='Viridis',
showscale=True,
colorbar=dict(title="Quality Score")
),
text=df_copy['model_name'],
hovertemplate=(
"<b>%{text}</b><br>" +
"Date: %{x}<br>" +
"Quality Score: %{y:.4f}<br>" +
"<extra></extra>"
),
name="Models"
))
# Add trend line
if len(df_copy) > 1:
z = np.polyfit(range(len(df_copy)), df_copy['quality_score'], 1)
trend_line = np.poly1d(z)(range(len(df_copy)))
fig.add_trace(go.Scatter(
x=df_copy['submission_date'],
y=trend_line,
mode='lines',
line=dict(dash='dash', color='red'),
name="Trend",
hoverinfo='skip'
))
fig.update_layout(
title="π
Model Performance Timeline",
xaxis_title="Submission Date",
yaxis_title="Quality Score",
height=500
)
return fig
def create_google_comparison_plot(df: pd.DataFrame) -> go.Figure:
"""Create plot comparing models on Google Translate-comparable language pairs."""
# Filter models that have Google comparable results
google_models = df[df['google_pairs_covered'] > 0].copy()
if google_models.empty:
fig = go.Figure()
fig.add_annotation(
text="No models with Google Translate comparable results",
x=0.5, y=0.5, showarrow=False
)
fig.update_layout(title="No Google Comparable Models")
return fig
fig = go.Figure()
# Create scatter plot
fig.add_trace(go.Scatter(
x=google_models['google_bleu'],
y=google_models['google_quality_score'],
mode='markers+text',
marker=dict(
size=12,
color=google_models['google_chrf'],
colorscale='Plasma',
showscale=True,
colorbar=dict(title="ChrF Score")
),
text=google_models['model_name'],
textposition="top center",
hovertemplate=(
"<b>%{text}</b><br>" +
"BLEU: %{x:.2f}<br>" +
"Quality: %{y:.4f}<br>" +
"ChrF: %{marker.color:.4f}<br>" +
"<extra></extra>"
),
name="Models"
))
fig.update_layout(
title="π€ Google Translate Comparable Performance",
xaxis_title="BLEU Score",
yaxis_title="Quality Score",
height=500
)
return fig
def create_detailed_model_analysis(model_results: Dict, model_name: str) -> go.Figure:
"""Create detailed analysis plot for a specific model - FIXED version."""
if not model_results or 'pair_metrics' not in model_results:
fig = go.Figure()
fig.add_annotation(text="No detailed results available", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title=f"No Data for {model_name}")
return fig
pair_metrics = model_results['pair_metrics']
# Extract language pair data
pairs = []
bleu_scores = []
quality_scores = []
sample_counts = []
google_comparable = []
for pair_key, metrics in pair_metrics.items():
if 'sample_count' in metrics and metrics['sample_count'] > 0:
src, tgt = pair_key.split('_to_')
pair_label = f"{LANGUAGE_NAMES.get(src, src)} β {LANGUAGE_NAMES.get(tgt, tgt)}"
pairs.append(pair_label)
bleu_scores.append(metrics.get('bleu', 0))
quality_scores.append(metrics.get('quality_score', 0))
sample_counts.append(metrics.get('sample_count', 0))
is_google = (src in GOOGLE_SUPPORTED_LANGUAGES and tgt in GOOGLE_SUPPORTED_LANGUAGES)
google_comparable.append(is_google)
if not pairs:
fig = go.Figure()
fig.add_annotation(text="No language pair data found", x=0.5, y=0.5, showarrow=False)
fig.update_layout(title=f"No Language Pair Data for {model_name}")
return fig
# Create subplot with proper spacing and titles
fig = make_subplots(
rows=2, cols=1,
subplot_titles=(
f"BLEU Scores by Language Pair",
f"Quality Scores by Language Pair"
),
vertical_spacing=0.15,
row_heights=[0.45, 0.45]
)
# Color code by Google comparable
colors = ['#1f77b4' if gc else '#ff7f0e' for gc in google_comparable]
# BLEU scores (top subplot)
fig.add_trace(
go.Bar(
x=pairs,
y=bleu_scores,
marker_color=colors,
name="BLEU",
text=[f"{score:.1f}" for score in bleu_scores],
textposition='outside',
textfont=dict(size=10),
showlegend=True
),
row=1, col=1
)
# Quality scores (bottom subplot)
fig.add_trace(
go.Bar(
x=pairs,
y=quality_scores,
marker_color=colors,
name="Quality",
text=[f"{score:.3f}" for score in quality_scores],
textposition='outside',
textfont=dict(size=10),
showlegend=False
),
row=2, col=1
)
# Update layout
fig.update_layout(
height=900,
title=dict(
text=f"π Detailed Analysis: {model_name}",
x=0.5,
xanchor='center'
),
showlegend=True,
margin=dict(l=50, r=50, t=100, b=150)
)
# Update x-axes to rotate labels properly
fig.update_xaxes(
tickangle=45,
tickfont=dict(size=10),
row=1, col=1
)
fig.update_xaxes(
tickangle=45,
tickfont=dict(size=10),
row=2, col=1
)
# Update y-axes
fig.update_yaxes(title_text="BLEU Score", row=1, col=1)
fig.update_yaxes(title_text="Quality Score", row=2, col=1)
# Add legend manually for Google vs UG40 only
fig.add_trace(
go.Scatter(
x=[None], y=[None],
mode='markers',
marker=dict(size=15, color='#1f77b4', symbol='square'),
name="Google Comparable",
showlegend=True
)
)
fig.add_trace(
go.Scatter(
x=[None], y=[None],
mode='markers',
marker=dict(size=15, color='#ff7f0e', symbol='square'),
name="UG40 Only",
showlegend=True
)
)
return fig
def create_submission_summary_plot(validation_info: Dict, evaluation_results: Dict) -> go.Figure:
"""Create summary plot for a new submission."""
fig = make_subplots(
rows=2, cols=2,
subplot_titles=(
"Sample Distribution",
"Primary Metrics",
"Error Analysis",
"Coverage Summary"
),
specs=[[{"type": "pie"}, {"type": "bar"}],
[{"type": "bar"}, {"type": "bar"}]]
)
# Sample distribution (pie chart)
coverage = validation_info.get('coverage', 0.8)
fig.add_trace(
go.Pie(
labels=["Evaluated", "Missing"],
values=[coverage * 100, (1 - coverage) * 100],
name="Samples"
),
row=1, col=1
)
# Primary metrics
if 'summary' in evaluation_results:
metrics_data = evaluation_results['summary']['primary_metrics']
metric_names = list(metrics_data.keys())
metric_values = list(metrics_data.values())
fig.add_trace(
go.Bar(
x=metric_names,
y=metric_values,
name="Metrics",
text=[f"{val:.3f}" for val in metric_values],
textposition='auto'
),
row=1, col=2
)
# Error analysis (CER, WER)
if 'averages' in evaluation_results:
error_metrics = ['cer', 'wer']
error_values = [evaluation_results['averages'].get(m, 0) for m in error_metrics]
fig.add_trace(
go.Bar(
x=error_metrics,
y=error_values,
name="Errors",
text=[f"{val:.3f}" for val in error_values],
textposition='auto'
),
row=2, col=1
)
# Coverage summary
if 'summary' in evaluation_results:
summary = evaluation_results['summary']
coverage_labels = ["Total Samples", "Lang Pairs", "Google Pairs"]
coverage_values = [
summary.get('total_samples', 0),
summary.get('language_pairs_covered', 0),
summary.get('google_comparable_pairs', 0)
]
fig.add_trace(
go.Bar(
x=coverage_labels,
y=coverage_values,
name="Coverage",
text=[f"{val}" for val in coverage_values],
textposition='auto'
),
row=2, col=2
)
fig.update_layout(
title="π Submission Summary",
height=700,
showlegend=False
)
return fig |