Spaces:
Running
Running
File size: 9,008 Bytes
38691ae |
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 |
"""
Chargement et gestion des données pour l'application Rakuten.
Ce module gère le chargement des données d'entraînement et de test,
avec fallback sur des données mock si les fichiers ne sont pas disponibles.
"""
import pandas as pd
import numpy as np
from pathlib import Path
from typing import Optional, Tuple, Dict
import sys
sys.path.insert(0, str(Path(__file__).parent.parent))
from config import RAW_DATA_DIR, PROCESSED_DATA_DIR, IMAGES_DIR
from utils.category_mapping import CATEGORY_CODES, CATEGORY_MAPPING
# =============================================================================
# Chemins des fichiers de données
# =============================================================================
X_TRAIN_PATH = RAW_DATA_DIR / "X_train_update.csv"
Y_TRAIN_PATH = RAW_DATA_DIR / "Y_train_CVw08PX.csv"
X_TEST_PATH = RAW_DATA_DIR / "X_test_update.csv"
IMAGES_TRAIN_DIR = IMAGES_DIR / "images" / "image_train"
IMAGES_TEST_DIR = IMAGES_DIR / "images" / "image_test"
def is_data_available() -> bool:
"""Vérifie si les données réelles sont disponibles."""
return X_TRAIN_PATH.exists() and Y_TRAIN_PATH.exists()
def load_training_data() -> Tuple[Optional[pd.DataFrame], Optional[pd.DataFrame]]:
"""
Charge les données d'entraînement (X et Y).
Returns:
Tuple (X_train, Y_train) ou (None, None) si non disponible
"""
if not is_data_available():
return None, None
try:
X_train = pd.read_csv(X_TRAIN_PATH, index_col=0)
Y_train = pd.read_csv(Y_TRAIN_PATH, index_col=0)
return X_train, Y_train
except Exception as e:
print(f"Erreur chargement données: {e}")
return None, None
def load_test_data() -> Optional[pd.DataFrame]:
"""Charge les données de test."""
if not X_TEST_PATH.exists():
return None
try:
return pd.read_csv(X_TEST_PATH, index_col=0)
except Exception:
return None
def get_category_distribution(Y_train: Optional[pd.DataFrame] = None) -> pd.DataFrame:
"""
Retourne la distribution des catégories.
Si Y_train est None, retourne des données mock réalistes basées
sur les statistiques connues du dataset Rakuten.
"""
if Y_train is not None:
# Données réelles
dist = Y_train['prdtypecode'].value_counts().reset_index()
dist.columns = ['category_code', 'count']
dist['category_code'] = dist['category_code'].astype(str)
else:
# Données mock réalistes (basées sur distribution connue)
mock_distribution = {
"2583": 8560, "1280": 7550, "2705": 6850, "2522": 6100,
"1560": 5800, "2060": 5500, "1920": 5200, "1140": 4800,
"1180": 4500, "1300": 4200, "2403": 4000, "2582": 3800,
"1281": 3600, "1320": 3400, "40": 3200, "1302": 3000,
"2280": 2800, "2220": 2600, "10": 2400, "1160": 2200,
"50": 2000, "60": 1800, "1301": 1600, "2585": 1400,
"1940": 1200, "2462": 1000, "2905": 800
}
dist = pd.DataFrame([
{"category_code": code, "count": count}
for code, count in mock_distribution.items()
])
# Ajouter les noms de catégories
dist['category_name'] = dist['category_code'].apply(
lambda x: CATEGORY_MAPPING.get(str(x), ("Inconnu", "", ""))[0]
)
dist['category_full'] = dist['category_code'].apply(
lambda x: CATEGORY_MAPPING.get(str(x), ("", "Inconnu", ""))[1]
)
dist['emoji'] = dist['category_code'].apply(
lambda x: CATEGORY_MAPPING.get(str(x), ("", "", "❓"))[2]
)
# Calculer les pourcentages
total = dist['count'].sum()
dist['percentage'] = (dist['count'] / total * 100).round(2)
return dist.sort_values('count', ascending=False).reset_index(drop=True)
def get_text_statistics(X_train: Optional[pd.DataFrame] = None) -> Dict:
"""
Calcule les statistiques sur les textes.
Returns:
Dict avec les statistiques texte
"""
if X_train is not None:
# Statistiques réelles
designation_lengths = X_train['designation'].str.len()
description_lengths = X_train['description'].fillna('').str.len()
# Langues (si disponible)
if 'lang' in X_train.columns:
lang_dist = X_train['lang'].value_counts().to_dict()
else:
lang_dist = None
return {
"total_products": len(X_train),
"designation": {
"mean_length": designation_lengths.mean(),
"min_length": designation_lengths.min(),
"max_length": designation_lengths.max(),
"median_length": designation_lengths.median(),
},
"description": {
"mean_length": description_lengths.mean(),
"min_length": description_lengths.min(),
"max_length": description_lengths.max(),
"non_empty_pct": (description_lengths > 0).mean() * 100,
},
"languages": lang_dist,
"is_mock": False
}
else:
# Statistiques mock réalistes
return {
"total_products": 84916,
"designation": {
"mean_length": 48.5,
"min_length": 3,
"max_length": 250,
"median_length": 42,
},
"description": {
"mean_length": 312.8,
"min_length": 0,
"max_length": 5000,
"non_empty_pct": 78.5,
},
"languages": {
"fr": 72500,
"en": 8200,
"de": 2100,
"es": 1200,
"it": 600,
"other": 316
},
"is_mock": True
}
def get_sample_products(
X_train: Optional[pd.DataFrame] = None,
Y_train: Optional[pd.DataFrame] = None,
category_code: str = None,
n_samples: int = 5
) -> pd.DataFrame:
"""
Retourne des exemples de produits.
Args:
X_train: DataFrame des features
Y_train: DataFrame des labels
category_code: Code catégorie pour filtrer (optionnel)
n_samples: Nombre d'exemples à retourner
"""
if X_train is not None and Y_train is not None:
# Données réelles
data = X_train.join(Y_train)
if category_code:
data = data[data['prdtypecode'] == int(category_code)]
return data.sample(min(n_samples, len(data)))
else:
# Données mock
mock_products = [
{"designation": "Harry Potter à l'école des sorciers", "description": "Roman fantastique de J.K. Rowling", "prdtypecode": "10"},
{"designation": "Console PlayStation 5", "description": "Console de jeux nouvelle génération Sony", "prdtypecode": "60"},
{"designation": "Piscine gonflable ronde 3m", "description": "Piscine familiale pour jardin", "prdtypecode": "2583"},
{"designation": "Figurine Dragon Ball Z Goku", "description": "Figurine collector 25cm", "prdtypecode": "1140"},
{"designation": "Canapé convertible 3 places", "description": "Canapé-lit confortable gris", "prdtypecode": "1560"},
{"designation": "Lot cartes Pokemon Pikachu", "description": "Collection 50 cartes Pokemon rares", "prdtypecode": "1160"},
{"designation": "Tondeuse à gazon électrique", "description": "Tondeuse 1800W coupe 42cm", "prdtypecode": "2582"},
{"designation": "Puzzle 1000 pièces paysage", "description": "Puzzle adulte montagne", "prdtypecode": "1281"},
{"designation": "Croquettes chien adulte 15kg", "description": "Alimentation premium pour chien", "prdtypecode": "2220"},
{"designation": "Perceuse visseuse sans fil", "description": "Perceuse 18V avec 2 batteries", "prdtypecode": "2585"},
]
df = pd.DataFrame(mock_products)
if category_code:
df = df[df['prdtypecode'] == category_code]
return df.head(n_samples)
def get_dataset_summary() -> Dict:
"""Retourne un résumé du dataset."""
X_train, Y_train = load_training_data()
if X_train is not None:
return {
"train_samples": len(X_train),
"test_samples": len(load_test_data()) if load_test_data() is not None else "N/A",
"num_categories": 27,
"features": ["designation", "description", "productid", "imageid"],
"is_mock": False
}
else:
return {
"train_samples": 84916,
"test_samples": 13812,
"num_categories": 27,
"features": ["designation", "description", "productid", "imageid"],
"is_mock": True
}
|