Spaces:
Sleeping
Sleeping
File size: 19,798 Bytes
a12ef5d 84cf73f a12ef5d 26e8a2f a12ef5d 26e8a2f a12ef5d 26e8a2f a12ef5d 26e8a2f a12ef5d 26e8a2f a12ef5d 26e8a2f a12ef5d 26e8a2f a12ef5d |
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 |
import pandas as pd
import numpy as np
from datetime import datetime
from data import extract_model_data
import gradio as gr
def get_time_series_summary_dfs(historical_df: pd.DataFrame) -> dict:
"""Return dataframes for historical summary plots (failure rates, AMD tests, NVIDIA tests)."""
# Group by date to get daily statistics
daily_stats = []
dates = sorted(historical_df['date'].unique())
for date in dates:
date_data = historical_df[historical_df['date'] == date]
amd_passed = date_data['success_amd'].sum() if 'success_amd' in date_data.columns else 0
amd_failed = (date_data['failed_multi_no_amd'].sum() + date_data['failed_single_no_amd'].sum()) if 'failed_multi_no_amd' in date_data.columns else 0
amd_skipped = date_data['skipped_amd'].sum() if 'skipped_amd' in date_data.columns else 0
amd_total = amd_passed + amd_failed + amd_skipped
amd_failure_rate = (amd_failed / amd_total * 100) if amd_total > 0 else 0
nvidia_passed = date_data['success_nvidia'].sum() if 'success_nvidia' in date_data.columns else 0
nvidia_failed = (date_data['failed_multi_no_nvidia'].sum() + date_data['failed_single_no_nvidia'].sum()) if 'failed_multi_no_nvidia' in date_data.columns else 0
nvidia_skipped = date_data['skipped_nvidia'].sum() if 'skipped_nvidia' in date_data.columns else 0
nvidia_total = nvidia_passed + nvidia_failed + nvidia_skipped
nvidia_failure_rate = (nvidia_failed / nvidia_total * 100) if nvidia_total > 0 else 0
daily_stats.append({
'date': date,
'amd_failure_rate': amd_failure_rate,
'nvidia_failure_rate': nvidia_failure_rate,
'amd_passed': amd_passed,
'amd_failed': amd_failed,
'amd_skipped': amd_skipped,
'nvidia_passed': nvidia_passed,
'nvidia_failed': nvidia_failed,
'nvidia_skipped': nvidia_skipped
})
# Failure rate dataframe
failure_rate_data = []
for i, stat in enumerate(daily_stats):
amd_change = stat['amd_failure_rate'] - daily_stats[i-1]['amd_failure_rate'] if i > 0 else 0
nvidia_change = stat['nvidia_failure_rate'] - daily_stats[i-1]['nvidia_failure_rate'] if i > 0 else 0
failure_rate_data.extend([
{'date': stat['date'], 'failure_rate': stat['amd_failure_rate'], 'platform': 'AMD', 'change': amd_change},
{'date': stat['date'], 'failure_rate': stat['nvidia_failure_rate'], 'platform': 'NVIDIA', 'change': nvidia_change}
])
failure_rate_df = pd.DataFrame(failure_rate_data)
# AMD tests dataframe
amd_data = []
for i, stat in enumerate(daily_stats):
passed_change = stat['amd_passed'] - daily_stats[i-1]['amd_passed'] if i > 0 else 0
failed_change = stat['amd_failed'] - daily_stats[i-1]['amd_failed'] if i > 0 else 0
skipped_change = stat['amd_skipped'] - daily_stats[i-1]['amd_skipped'] if i > 0 else 0
amd_data.extend([
{'date': stat['date'], 'count': stat['amd_passed'], 'test_type': 'Passed', 'change': passed_change},
{'date': stat['date'], 'count': stat['amd_failed'], 'test_type': 'Failed', 'change': failed_change},
{'date': stat['date'], 'count': stat['amd_skipped'], 'test_type': 'Skipped', 'change': skipped_change}
])
amd_df = pd.DataFrame(amd_data)
# NVIDIA tests dataframe
nvidia_data = []
for i, stat in enumerate(daily_stats):
passed_change = stat['nvidia_passed'] - daily_stats[i-1]['nvidia_passed'] if i > 0 else 0
failed_change = stat['nvidia_failed'] - daily_stats[i-1]['nvidia_failed'] if i > 0 else 0
skipped_change = stat['nvidia_skipped'] - daily_stats[i-1]['nvidia_skipped'] if i > 0 else 0
nvidia_data.extend([
{'date': stat['date'], 'count': stat['nvidia_passed'], 'test_type': 'Passed', 'change': passed_change},
{'date': stat['date'], 'count': stat['nvidia_failed'], 'test_type': 'Failed', 'change': failed_change},
{'date': stat['date'], 'count': stat['nvidia_skipped'], 'test_type': 'Skipped', 'change': skipped_change}
])
nvidia_df = pd.DataFrame(nvidia_data)
return {
'failure_rates_df': failure_rate_df,
'amd_tests_df': amd_df,
'nvidia_tests_df': nvidia_df,
}
def get_model_time_series_dfs(historical_df: pd.DataFrame, model_name: str) -> dict:
"""Return dataframes for a specific model's historical plots (AMD, NVIDIA)."""
model_data = historical_df[historical_df.index.str.lower() == model_name.lower()]
if model_data.empty:
empty_df = pd.DataFrame({'date': [], 'count': [], 'test_type': [], 'change': []})
return {'amd_df': empty_df.copy(), 'nvidia_df': empty_df.copy()}
dates = sorted(model_data['date'].unique())
amd_data = []
nvidia_data = []
for i, date in enumerate(dates):
date_data = model_data[model_data['date'] == date]
row = date_data.iloc[0]
amd_passed = row.get('success_amd', 0)
amd_failed = row.get('failed_multi_no_amd', 0) + row.get('failed_single_no_amd', 0)
amd_skipped = row.get('skipped_amd', 0)
prev_row = model_data[model_data['date'] == dates[i-1]].iloc[0] if i > 0 and not model_data[model_data['date'] == dates[i-1]].empty else None
amd_passed_change = amd_passed - (prev_row.get('success_amd', 0) if prev_row is not None else 0)
amd_failed_change = amd_failed - (prev_row.get('failed_multi_no_amd', 0) + prev_row.get('failed_single_no_amd', 0) if prev_row is not None else 0)
amd_skipped_change = amd_skipped - (prev_row.get('skipped_amd', 0) if prev_row is not None else 0)
amd_data.extend([
{'date': date, 'count': amd_passed, 'test_type': 'Passed', 'change': amd_passed_change},
{'date': date, 'count': amd_failed, 'test_type': 'Failed', 'change': amd_failed_change},
{'date': date, 'count': amd_skipped, 'test_type': 'Skipped', 'change': amd_skipped_change}
])
nvidia_passed = row.get('success_nvidia', 0)
nvidia_failed = row.get('failed_multi_no_nvidia', 0) + row.get('failed_single_no_nvidia', 0)
nvidia_skipped = row.get('skipped_nvidia', 0)
if prev_row is not None:
prev_nvidia_passed = prev_row.get('success_nvidia', 0)
prev_nvidia_failed = prev_row.get('failed_multi_no_nvidia', 0) + prev_row.get('failed_single_no_nvidia', 0)
prev_nvidia_skipped = prev_row.get('skipped_nvidia', 0)
else:
prev_nvidia_passed = prev_nvidia_failed = prev_nvidia_skipped = 0
nvidia_data.extend([
{'date': date, 'count': nvidia_passed, 'test_type': 'Passed', 'change': nvidia_passed - prev_nvidia_passed},
{'date': date, 'count': nvidia_failed, 'test_type': 'Failed', 'change': nvidia_failed - prev_nvidia_failed},
{'date': date, 'count': nvidia_skipped, 'test_type': 'Skipped', 'change': nvidia_skipped - prev_nvidia_skipped}
])
return {'amd_df': pd.DataFrame(amd_data), 'nvidia_df': pd.DataFrame(nvidia_data)}
def create_time_series_summary_gradio(historical_df: pd.DataFrame) -> dict:
"""Create time-series visualization for overall failure rates over time using Gradio native plots."""
if historical_df.empty or 'date' not in historical_df.columns:
# Return empty plots
empty_df = pd.DataFrame({'date': [], 'failure_rate': [], 'platform': []})
return {
'failure_rates': gr.LinePlot(empty_df, x="date", y="failure_rate", color="platform", title="No historical data available", tooltip=["failure_rate", "date", "change"]),
'amd_tests': gr.LinePlot(empty_df, x="date", y="failure_rate", color="platform", title="No historical data available", tooltip=["count", "date", "change"]),
'nvidia_tests': gr.LinePlot(empty_df, x="date", y="failure_rate", color="platform", title="No historical data available", tooltip=["count", "date", "change"])
}
# Group by date to get daily statistics
daily_stats = []
dates = sorted(historical_df['date'].unique())
for date in dates:
date_data = historical_df[historical_df['date'] == date]
# Calculate AMD stats - use the correct column names from the data structure
amd_passed = date_data['success_amd'].sum() if 'success_amd' in date_data.columns else 0
amd_failed = (date_data['failed_multi_no_amd'].sum() + date_data['failed_single_no_amd'].sum()) if 'failed_multi_no_amd' in date_data.columns else 0
amd_skipped = date_data['skipped_amd'].sum() if 'skipped_amd' in date_data.columns else 0
amd_total = amd_passed + amd_failed + amd_skipped
amd_failure_rate = (amd_failed / amd_total * 100) if amd_total > 0 else 0
# Calculate NVIDIA stats - use the correct column names from the data structure
nvidia_passed = date_data['success_nvidia'].sum() if 'success_nvidia' in date_data.columns else 0
nvidia_failed = (date_data['failed_multi_no_nvidia'].sum() + date_data['failed_single_no_nvidia'].sum()) if 'failed_multi_no_nvidia' in date_data.columns else 0
nvidia_skipped = date_data['skipped_nvidia'].sum() if 'skipped_nvidia' in date_data.columns else 0
nvidia_total = nvidia_passed + nvidia_failed + nvidia_skipped
nvidia_failure_rate = (nvidia_failed / nvidia_total * 100) if nvidia_total > 0 else 0
daily_stats.append({
'date': date,
'amd_failure_rate': amd_failure_rate,
'nvidia_failure_rate': nvidia_failure_rate,
'amd_passed': amd_passed,
'amd_failed': amd_failed,
'amd_skipped': amd_skipped,
'nvidia_passed': nvidia_passed,
'nvidia_failed': nvidia_failed,
'nvidia_skipped': nvidia_skipped
})
# Create failure rate data
failure_rate_data = []
for i, stat in enumerate(daily_stats):
# Calculate change from previous point
amd_change = 0
nvidia_change = 0
if i > 0:
amd_change = stat['amd_failure_rate'] - daily_stats[i-1]['amd_failure_rate']
nvidia_change = stat['nvidia_failure_rate'] - daily_stats[i-1]['nvidia_failure_rate']
failure_rate_data.extend([
{'date': stat['date'], 'failure_rate': stat['amd_failure_rate'], 'platform': 'AMD', 'change': amd_change},
{'date': stat['date'], 'failure_rate': stat['nvidia_failure_rate'], 'platform': 'NVIDIA', 'change': nvidia_change}
])
failure_rate_df = pd.DataFrame(failure_rate_data)
# Create AMD test results data
amd_data = []
for i, stat in enumerate(daily_stats):
# Calculate change from previous point for each test type
passed_change = 0
failed_change = 0
skipped_change = 0
if i > 0:
passed_change = stat['amd_passed'] - daily_stats[i-1]['amd_passed']
failed_change = stat['amd_failed'] - daily_stats[i-1]['amd_failed']
skipped_change = stat['amd_skipped'] - daily_stats[i-1]['amd_skipped']
amd_data.extend([
{'date': stat['date'], 'count': stat['amd_passed'], 'test_type': 'Passed', 'change': passed_change},
{'date': stat['date'], 'count': stat['amd_failed'], 'test_type': 'Failed', 'change': failed_change},
{'date': stat['date'], 'count': stat['amd_skipped'], 'test_type': 'Skipped', 'change': skipped_change}
])
amd_df = pd.DataFrame(amd_data)
# Create NVIDIA test results data
nvidia_data = []
for i, stat in enumerate(daily_stats):
# Calculate change from previous point for each test type
passed_change = 0
failed_change = 0
skipped_change = 0
if i > 0:
passed_change = stat['nvidia_passed'] - daily_stats[i-1]['nvidia_passed']
failed_change = stat['nvidia_failed'] - daily_stats[i-1]['nvidia_failed']
skipped_change = stat['nvidia_skipped'] - daily_stats[i-1]['nvidia_skipped']
nvidia_data.extend([
{'date': stat['date'], 'count': stat['nvidia_passed'], 'test_type': 'Passed', 'change': passed_change},
{'date': stat['date'], 'count': stat['nvidia_failed'], 'test_type': 'Failed', 'change': failed_change},
{'date': stat['date'], 'count': stat['nvidia_skipped'], 'test_type': 'Skipped', 'change': skipped_change}
])
nvidia_df = pd.DataFrame(nvidia_data)
return {
'failure_rates': gr.LinePlot(
failure_rate_df,
x="date",
y="failure_rate",
color="platform",
color_map={"AMD": "#FF6B6B", "NVIDIA": "#4ECDC4"},
title="Overall Failure Rates Over Time",
tooltip=["failure_rate", "date", "change"],
height=300,
x_label_angle=45,
y_title="Failure Rate (%)"
),
'amd_tests': gr.LinePlot(
amd_df,
x="date",
y="count",
color="test_type",
color_map={"Passed": "#4CAF50", "Failed": "#E53E3E", "Skipped": "#FFA500"},
title="AMD Test Results Over Time",
tooltip=["count", "date", "change"],
height=300,
x_label_angle=45,
y_title="Number of Tests"
),
'nvidia_tests': gr.LinePlot(
nvidia_df,
x="date",
y="count",
color="test_type",
color_map={"Passed": "#4CAF50", "Failed": "#E53E3E", "Skipped": "#FFA500"},
title="NVIDIA Test Results Over Time",
tooltip=["count", "date", "change"],
height=300,
x_label_angle=45,
y_title="Number of Tests"
)
}
def create_model_time_series_gradio(historical_df: pd.DataFrame, model_name: str) -> dict:
"""Create time-series visualization for a specific model using Gradio native plots."""
if historical_df.empty or 'date' not in historical_df.columns:
# Return empty plots
empty_df = pd.DataFrame({'date': [], 'count': [], 'test_type': []})
return {
'amd_plot': gr.LinePlot(empty_df, x="date", y="count", color="test_type", title=f"{model_name.upper()} - AMD Results Over Time", tooltip=["count", "date", "change"]),
'nvidia_plot': gr.LinePlot(empty_df, x="date", y="count", color="test_type", title=f"{model_name.upper()} - NVIDIA Results Over Time", tooltip=["count", "date", "change"])
}
# Filter data for the specific model (model_name is the index)
model_data = historical_df[historical_df.index.str.lower() == model_name.lower()]
if model_data.empty:
# Return empty plots
empty_df = pd.DataFrame({'date': [], 'count': [], 'test_type': []})
return {
'amd_plot': gr.LinePlot(empty_df, x="date", y="count", color="test_type", title=f"{model_name.upper()} - AMD Results Over Time", tooltip=["count", "date", "change"]),
'nvidia_plot': gr.LinePlot(empty_df, x="date", y="count", color="test_type", title=f"{model_name.upper()} - NVIDIA Results Over Time", tooltip=["count", "date", "change"])
}
# Group by date
dates = sorted(model_data['date'].unique())
amd_data = []
nvidia_data = []
for i, date in enumerate(dates):
date_data = model_data[model_data['date'] == date]
if not date_data.empty:
# Get the first row for this date (should be only one)
row = date_data.iloc[0]
# AMD data - use the correct column names from the data structure
amd_passed = row.get('success_amd', 0)
amd_failed = row.get('failed_multi_no_amd', 0) + row.get('failed_single_no_amd', 0)
amd_skipped = row.get('skipped_amd', 0)
# Calculate change from previous point
passed_change = 0
failed_change = 0
skipped_change = 0
if i > 0:
prev_date_data = model_data[model_data['date'] == dates[i-1]]
if not prev_date_data.empty:
prev_row = prev_date_data.iloc[0]
prev_amd_passed = prev_row.get('success_amd', 0)
prev_amd_failed = prev_row.get('failed_multi_no_amd', 0) + prev_row.get('failed_single_no_amd', 0)
prev_amd_skipped = prev_row.get('skipped_amd', 0)
passed_change = amd_passed - prev_amd_passed
failed_change = amd_failed - prev_amd_failed
skipped_change = amd_skipped - prev_amd_skipped
amd_data.extend([
{'date': date, 'count': amd_passed, 'test_type': 'Passed', 'change': passed_change},
{'date': date, 'count': amd_failed, 'test_type': 'Failed', 'change': failed_change},
{'date': date, 'count': amd_skipped, 'test_type': 'Skipped', 'change': skipped_change}
])
# NVIDIA data - use the correct column names from the data structure
nvidia_passed = row.get('success_nvidia', 0)
nvidia_failed = row.get('failed_multi_no_nvidia', 0) + row.get('failed_single_no_nvidia', 0)
nvidia_skipped = row.get('skipped_nvidia', 0)
# Calculate change from previous point for NVIDIA
nvidia_passed_change = 0
nvidia_failed_change = 0
nvidia_skipped_change = 0
if i > 0:
prev_date_data = model_data[model_data['date'] == dates[i-1]]
if not prev_date_data.empty:
prev_row = prev_date_data.iloc[0]
prev_nvidia_passed = prev_row.get('success_nvidia', 0)
prev_nvidia_failed = prev_row.get('failed_multi_no_nvidia', 0) + prev_row.get('failed_single_no_nvidia', 0)
prev_nvidia_skipped = prev_row.get('skipped_nvidia', 0)
nvidia_passed_change = nvidia_passed - prev_nvidia_passed
nvidia_failed_change = nvidia_failed - prev_nvidia_failed
nvidia_skipped_change = nvidia_skipped - prev_nvidia_skipped
nvidia_data.extend([
{'date': date, 'count': nvidia_passed, 'test_type': 'Passed', 'change': nvidia_passed_change},
{'date': date, 'count': nvidia_failed, 'test_type': 'Failed', 'change': nvidia_failed_change},
{'date': date, 'count': nvidia_skipped, 'test_type': 'Skipped', 'change': nvidia_skipped_change}
])
amd_df = pd.DataFrame(amd_data)
nvidia_df = pd.DataFrame(nvidia_data)
return {
'amd_plot': gr.LinePlot(
amd_df,
x="date",
y="count",
color="test_type",
color_map={"Passed": "#4CAF50", "Failed": "#E53E3E", "Skipped": "#FFA500"},
title=f"{model_name.upper()} - AMD Results Over Time",
x_label_angle=45,
y_title="Number of Tests",
height=300,
tooltip=["count", "date", "change"]
),
'nvidia_plot': gr.LinePlot(
nvidia_df,
x="date",
y="count",
color="test_type",
color_map={"Passed": "#4CAF50", "Failed": "#E53E3E", "Skipped": "#FFA500"},
title=f"{model_name.upper()} - NVIDIA Results Over Time",
x_label_angle=45,
y_title="Number of Tests",
height=300,
tooltip=["count", "date", "change"]
)
}
|