Spaces:
Sleeping
Sleeping
File size: 17,476 Bytes
21b40a1 3415945 21b40a1 3415945 21b40a1 ed8f460 129bb0f 78b8e12 ed8f460 9b08332 8c6aef7 9b08332 e1ac208 78b8e12 9b08332 ed8f460 8c6aef7 9b08332 8c6aef7 e1ac208 9b08332 e1ac208 ed8f460 9b08332 ed8f460 3415945 9b08332 21b40a1 e1ac208 9b08332 e1ac208 9b08332 e1ac208 9b08332 9edc0a9 ed8f460 e1ac208 9b08332 21b40a1 9b08332 21b40a1 9b08332 129bb0f 9b08332 129bb0f 9b08332 ed8f460 129bb0f 3415945 9b08332 21b40a1 e1ac208 8c6aef7 21b40a1 3415945 78b8e12 9b08332 78b8e12 3415945 9edc0a9 78b8e12 3415945 129bb0f 9b08332 e1ac208 9edc0a9 e1ac208 9b08332 e1ac208 9b08332 129bb0f e1ac208 9b08332 e1ac208 9b08332 129bb0f e1ac208 9b08332 e1ac208 9b08332 e1ac208 9b08332 e1ac208 129bb0f e1ac208 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 129bb0f 9b08332 e1ac208 129bb0f 9b08332 129bb0f 21b40a1 129bb0f 9b08332 129bb0f 9b08332 9edc0a9 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9edc0a9 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 833d0fe 9b08332 78b8e12 9b08332 833d0fe 9b08332 9edc0a9 9b08332 3415945 129bb0f 3415945 9b08332 833d0fe 9b08332 833d0fe 9b08332 21b40a1 833d0fe 9b08332 833d0fe 9b08332 9edc0a9 9b08332 9edc0a9 9b08332 129bb0f 9b08332 21b40a1 9edc0a9 129bb0f 9b08332 21b40a1 e1ac208 21b40a1 78b8e12 9b08332 78b8e12 e1ac208 78b8e12 129bb0f 78b8e12 129bb0f 78b8e12 833d0fe 78b8e12 833d0fe e1ac208 78b8e12 21b40a1 9b08332 78b8e12 9edc0a9 9b08332 78b8e12 9b08332 78b8e12 833d0fe 9b08332 78b8e12 9edc0a9 9b08332 9edc0a9 9b08332 78b8e12 e1ac208 78b8e12 6228329 e1ac208 21b40a1 78b8e12 |
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 |
import io
import requests
import pandas as pd
import numpy as np
import gradio as gr
import json
import pickle
import os
from typing import Dict, List, Any
from flask import Flask, request, jsonify
from flask_cors import CORS
import threading
os.environ["TF_ENABLE_ONEDNN_OPTS"] = "0"
os.environ["TF_CPP_MIN_LOG_LEVEL"] = "2"
print("π Starting Eco Finder API...")
# Initialize Flask app
flask_app = Flask(__name__)
CORS(flask_app) # Enable CORS for all routes
# Configuration
try:
import tensorflow as tf
print(f"β
TensorFlow version: {tf.__version__}")
from tensorflow.keras.models import load_model
TENSORFLOW_AVAILABLE = True
except ImportError as e:
print(f"β TensorFlow not available: {e}")
TENSORFLOW_AVAILABLE = False
# Load resources
def load_resources():
try:
with open("feature_stats.json", "r") as f:
feature_stats = json.load(f)
print("β
Feature stats loaded")
with open("scaler.pkl", "rb") as f:
scaler = pickle.load(f)
print("β
Scaler loaded")
with open("label_encoder.pkl", "rb") as f:
label_encoder = pickle.load(f)
print("β
Label encoder loaded")
model = None
if TENSORFLOW_AVAILABLE:
model = load_model("modulo_tabular.h5")
print("β
Model loaded")
return model, scaler, label_encoder, feature_stats
except Exception as e:
print(f"β Error loading resources: {str(e)}")
feature_stats = {
"feature_columns": [
"koi_period",
"koi_duration",
"koi_depth",
"koi_prad",
"koi_srad",
"koi_teq",
"koi_steff",
"koi_slogg",
"koi_smet",
"koi_kepmag",
"koi_model_snr",
"koi_num_transits",
],
"train_medians": {
"koi_period": 10.0,
"koi_duration": 5.0,
"koi_depth": 1000.0,
"koi_prad": 2.0,
"koi_srad": 1.0,
"koi_teq": 1000.0,
"koi_steff": 6000.0,
"koi_slogg": 4.5,
"koi_smet": 0.0,
"koi_kepmag": 12.0,
"koi_model_snr": 10.0,
"koi_num_transits": 3.0,
},
}
return None, None, None, feature_stats
# Load resources
model, scaler, label_encoder, feature_stats = load_resources()
feature_columns = feature_stats.get("feature_columns", [])
train_medians = feature_stats.get("train_medians", {})
BASE = "https://exoplanetarchive.ipac.caltech.edu/cgi-bin/nstedAPI/nph-nstedAPI"
# ==================== FLASK API ENDPOINTS ====================
@flask_app.route('/')
def home():
return jsonify({
"message": "Eco Finder API - Exoplanet Classification",
"version": "1.0.0",
"endpoints": {
"health": "/health (GET)",
"predict": "/predict (POST)",
"predict_batch": "/predict-batch (POST)",
"features": "/features (GET)"
}
})
@flask_app.route('/health', methods=['GET'])
def health():
return jsonify({
"status": "healthy",
"model_loaded": model is not None,
"tensorflow_available": TENSORFLOW_AVAILABLE,
"features_count": len(feature_columns)
})
@flask_app.route('/predict', methods=['POST'])
def api_predict():
"""REST API endpoint for single prediction"""
try:
data = request.get_json()
if not data:
return jsonify({"error": "No JSON data provided"}), 400
# Use default values if parameters are missing
features = {}
for feature in feature_columns:
features[feature] = data.get(feature, train_medians.get(feature, 0))
# Make prediction
result = predict_single(features)
if "error" in result:
return jsonify(result), 500
return jsonify(result)
except Exception as e:
return jsonify({"error": str(e)}), 500
@flask_app.route('/predict-batch', methods=['POST'])
def api_predict_batch():
"""REST API endpoint for batch predictions"""
try:
data = request.get_json()
if not data or 'objects' not in data:
return jsonify({"error": "No 'objects' array provided"}), 400
predictions = []
for obj in data['objects']:
features = {}
for feature in feature_columns:
features[feature] = obj.get(feature, train_medians.get(feature, 0))
result = predict_single(features)
predictions.append(result)
return jsonify({"predictions": predictions})
except Exception as e:
return jsonify({"error": str(e)}), 500
@flask_app.route('/features', methods=['GET'])
def api_features():
"""Get information about available features"""
return jsonify({
"feature_columns": feature_columns,
"train_medians": train_medians,
"feature_descriptions": {
"koi_period": "Orbital period (days)",
"koi_duration": "Transit duration (hours)",
"koi_depth": "Transit depth (ppm)",
"koi_prad": "Planetary radius (Earth radii)",
"koi_srad": "Stellar radius (Solar radii)",
"koi_teq": "Equilibrium temperature (K)",
"koi_steff": "Stellar effective temperature (K)",
"koi_slogg": "Stellar surface gravity (log g)",
"koi_smet": "Stellar metallicity ([Fe/H])",
"koi_kepmag": "TESS magnitude",
"koi_model_snr": "Signal-to-noise ratio",
"koi_num_transits": "Number of transits"
}
})
# ==================== PREDICTION FUNCTIONS ====================
def predict_single(features: Dict) -> Dict:
"""Function to predict a single object"""
try:
if model is None or scaler is None or label_encoder is None:
return {"error": "Model not available"}
# Create feature array
input_features = []
for feature in feature_columns:
value = features.get(feature, train_medians.get(feature, 0))
input_features.append(float(value))
# Predict
input_array = np.array([input_features])
X_input = scaler.transform(input_array)
if TENSORFLOW_AVAILABLE:
probs = model.predict(X_input, verbose=0)[0]
else:
probs = np.random.dirichlet(np.ones(3), size=1)[0]
pred_idx = np.argmax(probs)
pred_label = label_encoder.inverse_transform([pred_idx])[0]
return {
"prediction": pred_label,
"probabilities": {
"CONFIRMED": float(probs[0]),
"CANDIDATE": float(probs[1]),
"FALSE_POSITIVE": float(probs[2]),
},
"input_features": dict(zip(feature_columns, input_features)),
}
except Exception as e:
return {"error": str(e)}
def predict_from_dict(
koi_period: float,
koi_duration: float,
koi_depth: float,
koi_prad: float,
koi_srad: float,
koi_teq: float,
koi_steff: float,
koi_slogg: float,
koi_smet: float,
koi_kepmag: float,
koi_model_snr: float,
koi_num_transits: float,
) -> Dict:
"""Wrapper that takes individual parameters and converts them to dict"""
features = {
"koi_period": koi_period,
"koi_duration": koi_duration,
"koi_depth": koi_depth,
"koi_prad": koi_prad,
"koi_srad": koi_srad,
"koi_teq": koi_teq,
"koi_steff": koi_steff,
"koi_slogg": koi_slogg,
"koi_smet": koi_smet,
"koi_kepmag": koi_kepmag,
"koi_model_snr": koi_model_snr,
"koi_num_transits": koi_num_transits,
}
return predict_single(features)
def predict_toi_realtime():
"""Function for real-time TOI"""
try:
if model is None or scaler is None or label_encoder is None:
return "β Model not available"
# Query exoplanet API
where = (
"(tfopwg_disp like 'PC' or tfopwg_disp like 'APC') "
"and (pl_orbper is not null or tce_period is not null)"
)
params = {"table": "toi", "where": where, "format": "csv"}
resp = requests.get(BASE, params=params, timeout=60)
resp.raise_for_status()
toi_df = pd.read_csv(io.StringIO(resp.text))
if toi_df.empty:
return "β No TOI objects found"
# Take sample
toi_sample = toi_df.sample(min(3, len(toi_df)), random_state=7)
toi_sample.columns = [c.strip().lower() for c in toi_sample.columns]
# Synonym mapping
candidates_map = {
"koi_period": ["pl_orbper", "tce_period", "orbper", "period"],
"koi_duration": [
"pl_trandurh",
"tce_duration",
"tran_dur",
"trandur",
"duration",
"dur",
],
"koi_depth": ["pl_trandep", "tce_depth", "depth", "trandep"],
"koi_prad": ["pl_rade", "prad", "rade", "planet_radius"],
"koi_srad": ["st_rad", "srad", "stellar_radius", "star_radius"],
"koi_teq": ["pl_eqt", "teq", "equilibrium_temp"],
"koi_steff": ["st_teff", "teff", "stellar_teff", "effective_temp"],
"koi_slogg": ["st_logg", "logg", "slogg"],
"koi_smet": ["st_met", "feh", "metallicity", "smet"],
"koi_kepmag": ["st_tmag", "tmag", "kepmag", "koi_kepmag"],
"koi_model_snr": ["tce_model_snr", "model_snr", "snr"],
"koi_num_transits": [
"tce_num_transits",
"num_transits",
"ntransits",
"tran_count",
],
}
def first_present(candidates, cols_set):
for name in candidates:
if name in cols_set:
return name
for name in candidates:
found = [c for c in cols_set if name in c]
if found:
return found[0]
return None
cols_set = set(toi_sample.columns)
results = []
for idx, row in toi_sample.iterrows():
# Prepare features
features = {}
for feat in feature_columns:
src = first_present(candidates_map.get(feat, []), cols_set)
if src and src in row and pd.notna(row[src]):
features[feat] = float(row[src])
else:
features[feat] = train_medians.get(feat, 0)
# Predict
result = predict_single(features)
if "error" not in result:
results.append(
{
"TOI": row.get("toi", f"TOI-{idx}"),
"Disposition": row.get("tfopwg_disp", "Unknown"),
"Prediction": result["prediction"],
"P(Confirmed)": f"{result['probabilities']['CONFIRMED']:.3f}",
"P(Candidate)": f"{result['probabilities']['CANDIDATE']:.3f}",
"P(False Positive)": f"{result['probabilities']['FALSE_POSITIVE']:.3f}",
}
)
if not results:
return "β Could not generate predictions"
result_df = pd.DataFrame(results)
return f"**TOI Predictions:**\n\n{result_df.to_markdown(index=False)}"
except Exception as e:
return f"β Error: {str(e)}"
def predict_manual(
period,
duration,
depth,
prad,
srad,
teq,
steff,
slogg,
smet,
kepmag,
snr,
num_transits,
):
"""Function for manual prediction in Gradio"""
try:
result = predict_from_dict(
period,
duration,
depth,
prad,
srad,
teq,
steff,
slogg,
smet,
kepmag,
snr,
num_transits,
)
if "error" in result:
return f"β {result['error']}"
output = f"**Prediction:** {result['prediction']}\n\n**Probabilities:**\n"
for clase, prob in result["probabilities"].items():
output += f"- {clase}: {prob:.3f}\n"
return output
except Exception as e:
return f"β Error: {str(e)}"
# ==================== GRADIO INTERFACE ====================
with gr.Blocks(theme=gr.themes.Soft(), title="Eco Finder API") as demo:
gr.Markdown("# π Eco Finder API")
gr.Markdown("Exoplanet classifier with REST API")
with gr.Tab("π― API Documentation"):
gr.Markdown("""
## REST API Endpoints
### Health Check
**GET** `/health`
```bash
curl -X GET "https://your-domain/health"
```
### Single Prediction
**POST** `/predict`
```bash
curl -X POST "https://your-domain/predict" \\
-H "Content-Type: application/json" \\
-d '{
"koi_period": 10.0,
"koi_duration": 5.0,
"koi_depth": 1000.0,
"koi_prad": 2.0,
"koi_srad": 1.0,
"koi_teq": 1000.0,
"koi_steff": 6000.0,
"koi_slogg": 4.5,
"koi_smet": 0.0,
"koi_kepmag": 12.0,
"koi_model_snr": 10.0,
"koi_num_transits": 3.0
}'
```
### Batch Predictions
**POST** `/predict-batch`
```bash
curl -X POST "https://your-domain/predict-batch" \\
-H "Content-Type: application/json" \\
-d '{
"objects": [
{"koi_period": 10.0, "koi_duration": 5.0, ...},
{"koi_period": 15.0, "koi_duration": 6.0, ...}
]
}'
```
### Features Information
**GET** `/features`
```bash
curl -X GET "https://your-domain/features"
```
### JavaScript Example
```javascript
async function predictExoplanet(features) {
const response = await fetch('/predict', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body: JSON.stringify(features)
});
return await response.json();
}
// Usage
const result = await predictExoplanet({
koi_period: 10.0,
koi_duration: 5.0,
// ... all parameters
});
```
""")
with gr.Tab("π§ Test Interface"):
gr.Markdown("Test the prediction model with this interface")
with gr.Row():
with gr.Column():
period = gr.Number(label="koi_period", value=10.0)
duration = gr.Number(label="koi_duration", value=5.0)
depth = gr.Number(label="koi_depth", value=1000.0)
prad = gr.Number(label="koi_prad", value=2.0)
with gr.Column():
srad = gr.Number(label="koi_srad", value=1.0)
teq = gr.Number(label="koi_teq", value=1000.0)
steff = gr.Number(label="koi_steff", value=6000.0)
slogg = gr.Number(label="koi_slogg", value=4.5)
with gr.Column():
smet = gr.Number(label="koi_smet", value=0.0)
kepmag = gr.Number(label="koi_kepmag", value=12.0)
snr = gr.Number(label="koi_model_snr", value=10.0)
num_transits = gr.Number(label="koi_num_transits", value=3.0)
test_btn = gr.Button("π Test Prediction")
test_output = gr.JSON()
test_btn.click(
fn=predict_from_dict,
inputs=[
period,
duration,
depth,
prad,
srad,
teq,
steff,
slogg,
smet,
kepmag,
snr,
num_transits,
],
outputs=test_output,
)
with gr.Tab("π Real-time TOI"):
gr.Markdown("Real-time TOI object predictions")
toi_btn = gr.Button("π Analyze TOI")
toi_output = gr.Markdown()
toi_btn.click(predict_toi_realtime, outputs=toi_output)
# ==================== APPLICATION STARTUP ====================
def run_flask():
"""Run Flask app on port 5000"""
flask_app.run(host='0.0.0.0', port=5000, debug=False)
def run_gradio():
"""Run Gradio app on port 7860"""
demo.launch(server_name='0.0.0.0', server_port=7860, share=True)
if __name__ == "__main__":
print("π Application started successfully!")
print("π Gradio Interface available at: http://0.0.0.0:7860")
print("π REST API available at: http://0.0.0.0:5000")
print("π API Documentation:")
print(" GET /health")
print(" POST /predict")
print(" POST /predict-batch")
print(" GET /features")
# Start both servers
flask_thread = threading.Thread(target=run_flask)
flask_thread.daemon = True
flask_thread.start()
# Run Gradio in main thread
run_gradio() |