Spaces:
Sleeping
Sleeping
File size: 11,551 Bytes
b1af7f1 |
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 |
import sys
import os
import numpy as np
import pandas as pd
from geopy.distance import geodesic
from sklearn.preprocessing import LabelEncoder, StandardScaler
from src.constants.training_pipeline import TARGET_COLUMN
from src.constants.training_pipeline import DATA_TRANSFORMATION_IMPUTER_PARAMS
from src.entity.artifact_entity import (
DataTransformationArtifact,
DataValidationArtifact
)
from sklearn.pipeline import Pipeline
from src.entity.config_entity import DataTransformationConfig
from src.exception.exception import DeliveryTimeException
from src.logging.logger import logging
from src.utils.main_utils.utils import save_numpy_array_data, save_object
import joblib
class DataTransformation:
def __init__(self, data_validation_artifact:DataValidationArtifact,
data_transformation_config:DataTransformationConfig):
try:
self.data_validation_artifact:DataValidationArtifact=data_validation_artifact
self.data_transformation_config:DataTransformationConfig=data_transformation_config
except Exception as e:
raise DeliveryTimeException(e, sys)
@staticmethod
def read_data(file_path) -> pd.DataFrame:
try:
return pd.read_csv(file_path)
except Exception as e:
raise DeliveryTimeException(e, sys)
@staticmethod
def extract_time_taken(val):
try:
if isinstance(val, str) and " " in val:
return int(val.split(" ")[1])
elif isinstance(val, (int, float)):
return int(val)
else:
return np.nan # fallback
except:
return np.nan
def preprocess_data(self, df):
try:
df.rename(columns={'Weatherconditions': 'Weather_conditions'}, inplace=True)
if 'Time_taken(min)' in df.columns:
df['Time_taken(min)'] = df['Time_taken(min)'].apply(self.extract_time_taken)
df['Weather_conditions'] = df['Weather_conditions'].apply(lambda x: x.split(' ')[1].strip() if pd.notnull(x) else x)
df['City_code'] = df['Delivery_person_ID'].str.split("RES", expand=True)[0]
df.drop(['ID', 'Delivery_person_ID'], axis=1, inplace=True)
df['Delivery_person_Age'] = pd.to_numeric(df['Delivery_person_Age'], errors='coerce').astype('float64')
df['Delivery_person_Ratings'] = pd.to_numeric(df['Delivery_person_Ratings'], errors='coerce').astype('float64')
df['multiple_deliveries'] = pd.to_numeric(df['multiple_deliveries'], errors='coerce').astype('float64')
df['Order_Date'] = pd.to_datetime(df['Order_Date'], format="%d-%m-%Y")
df.replace('NaN', float(np.nan), regex=True, inplace=True)
df['Delivery_person_Age'] = df['Delivery_person_Age'].fillna(np.random.choice(df['Delivery_person_Age'].dropna()))
df['Weather_conditions'] = df['Weather_conditions'].fillna(np.random.choice(df['Weather_conditions'].dropna()))
df['City'] = df['City'].fillna(df['City'].mode()[0])
df['Festival'] = df['Festival'].fillna(df['Festival'].mode()[0])
df['multiple_deliveries'] = df['multiple_deliveries'].fillna(df['multiple_deliveries'].mode()[0])
df['Road_traffic_density'] = df['Road_traffic_density'].fillna(df['Road_traffic_density'].mode()[0])
df['Delivery_person_Ratings'] = df['Delivery_person_Ratings'].fillna(df['Delivery_person_Ratings'].median())
logging.info("Data preprocessing completed for DataFrame")
return df
except Exception as e:
raise DeliveryTimeException(e, sys)
def label_encoding(self, df):
try:
categorical_columns = df.select_dtypes(include='object').columns
for col in categorical_columns:
le = LabelEncoder()
df[col] = le.fit_transform(df[col])
joblib.dump(le, 'final_model/label_encoder.pkl')
logging.info("Label encoding completed for DataFrame")
return df
except Exception as e:
raise DeliveryTimeException(e, sys)
def feature_engineering(self, df):
try:
logging.info(f"Starting feature engineering. Initial DataFrame shape: {df.shape}")
# Drop rows with missing coordinates that will break geodesic distance calculation
df.dropna(subset=[
'Restaurant_latitude', 'Restaurant_longitude',
'Delivery_location_latitude', 'Delivery_location_longitude'
], inplace=True)
# --- Temporal features from Order_Date ---
df["day"] = df.Order_Date.dt.day
df["month"] = df.Order_Date.dt.month
df["quarter"] = df.Order_Date.dt.quarter
df["year"] = df.Order_Date.dt.year
df["day_of_week"] = df.Order_Date.dt.day_of_week.astype(int)
df["is_month_start"] = df.Order_Date.dt.is_month_start.astype(int)
df["is_month_end"] = df.Order_Date.dt.is_month_end.astype(int)
df["is_quarter_start"] = df.Order_Date.dt.is_quarter_start.astype(int)
df["is_quarter_end"] = df.Order_Date.dt.is_quarter_end.astype(int)
df["is_year_start"] = df.Order_Date.dt.is_year_start.astype(int)
df["is_year_end"] = df.Order_Date.dt.is_year_end.astype(int)
df["is_weekend"] = np.where(df["day_of_week"].isin([5, 6]), 1, 0)
# --- Order preparation time ---
df["Time_Orderd"] = pd.to_timedelta(df["Time_Orderd"].fillna("00:00:00"))
df["Time_Order_picked"] = pd.to_timedelta(df["Time_Order_picked"].fillna("00:00:00"))
df["Time_Ordered_formatted"] = df["Order_Date"] + df["Time_Orderd"]
df["Time_Order_picked_base"] = df["Order_Date"] + df["Time_Order_picked"]
mask = df["Time_Order_picked"] < df["Time_Orderd"]
df["Time_Order_picked_formatted"] = df["Time_Order_picked_base"].copy()
df.loc[mask, "Time_Order_picked_formatted"] += pd.Timedelta(days=1)
df["order_prepare_time"] = (
df["Time_Order_picked_formatted"] - df["Time_Ordered_formatted"]
).dt.total_seconds() / 60
df["order_prepare_time"] = df["order_prepare_time"].fillna(df["order_prepare_time"].median())
# Drop intermediate time columns
df.drop([
"Time_Orderd", "Time_Order_picked", "Time_Ordered_formatted",
"Time_Order_picked_base", "Time_Order_picked_formatted", "Order_Date"
], axis=1, inplace=True)
# --- Label encoding for categorical features (done before numeric operations) ---
df = self.label_encoding(df)
# --- Geodesic distance calculation ---
restaurant_coords = df[["Restaurant_latitude", "Restaurant_longitude"]].to_numpy()
delivery_coords = df[["Delivery_location_latitude", "Delivery_location_longitude"]].to_numpy()
df["distance"] = np.array([
geodesic(restaurant, delivery).kilometers
for restaurant, delivery in zip(restaurant_coords, delivery_coords)
])
# --- Derived composite features ---
df["distance_traffic"] = df["distance"] * df["Road_traffic_density"]
df["distance_deliveries"] = df["distance"] * df["multiple_deliveries"]
df["prep_traffic"] = df["order_prepare_time"] * df["Road_traffic_density"]
df["age_ratings"] = df["Delivery_person_Age"] * df["Delivery_person_Ratings"]
df["prep_distance"] = df["order_prepare_time"] * df["distance"]
# --- Outlier capping ---
for col in ["distance", "order_prepare_time", "Delivery_person_Age", "multiple_deliveries"]:
upper_limit = df[col].quantile(0.99)
df[col] = np.where(df[col] > upper_limit, upper_limit, df[col])
# --- Select final features ---
selected_features = [
"multiple_deliveries", "Road_traffic_density", "Vehicle_condition",
"Delivery_person_Ratings", "distance_deliveries", "Weather_conditions",
"Festival", "distance_traffic", "distance", "Delivery_person_Age",
"prep_traffic", "City", "Time_taken(min)"
]
clean_df = df[selected_features]
# --- Remove rows with any NaN or inf values ---
clean_df = clean_df.replace([np.inf, -np.inf], np.nan)
clean_df = clean_df.dropna()
logging.info("Feature engineering completed successfully.")
logging.info(f"Final cleaned DataFrame shape: {clean_df.shape}")
return clean_df
except Exception as e:
raise DeliveryTimeException(e, sys)
def initiate_data_transformation(self)->DataTransformationArtifact:
logging.info("Entered initialize_data_transformation method of Data transformation class")
try:
logging.info("Starting data transformation")
# Reading train and test data
logging.info("Reading train and test data")
train_df = DataTransformation.read_data(self.data_validation_artifact.valid_train_file_path)
test_df = DataTransformation.read_data(self.data_validation_artifact.valid_test_file_path)
preprocessed_train_df = self.preprocess_data(train_df)
preprocessed_test_df = self.preprocess_data(test_df)
feature_engineered_train_df = self.feature_engineering(preprocessed_train_df)
feature_engineered_test_df = self.feature_engineering(preprocessed_test_df)
# # Training dataframe
# input_feature_train_df=feature_engineered_train_df.drop(columns=[TARGET_COLUMN], axis=1)
# target_feature_train_df=feature_engineered_train_df[TARGET_COLUMN]
# target_feature_train_df= target_feature_train_df.replace(-1, 0)
# ## Testing dataframe
# input_feature_test_df=feature_engineered_test_df.drop(columns=[TARGET_COLUMN], axis=1)
# target_feature_test_df = feature_engineered_test_df[TARGET_COLUMN]
# target_feature_test_df = target_feature_test_df.replace(-1, 0)
logging.info(f"Shape after train feature engineering {feature_engineered_train_df.shape}")
logging.info(f"Shape after test feature engineering {feature_engineered_test_df.shape}")
train_arr=np.c_[feature_engineered_train_df]
test_arr = np.c_[feature_engineered_test_df]
## Save the numpy array data
save_numpy_array_data(self.data_transformation_config.transformed_train_file_path, array=train_arr,)
save_numpy_array_data(self.data_transformation_config.transformed_test_file_path, array=test_arr)
## Preparing artifacts
data_transformation_artifact=DataTransformationArtifact(
transformed_object_file_path=self.data_transformation_config.transformed_object_file_path,
transformed_train_file_path=self.data_transformation_config.transformed_train_file_path,
transformed_test_file_path=self.data_transformation_config.transformed_test_file_path
)
return data_transformation_artifact
except Exception as e:
raise DeliveryTimeException(e, sys)
|