File size: 2,964 Bytes
3677ab8 | 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 | """Time-based feature computation: rolling windows, inter-order stats."""
import pandas as pd
from src.config import FEATURE_WINDOWS
from src.utils.logger import get_logger
logger = get_logger(__name__)
def compute_time_features(
orders: pd.DataFrame,
reference_date: pd.Timestamp,
windows: list[int] = FEATURE_WINDOWS,
) -> pd.DataFrame:
"""Compute time-based features per customer.
Features produced:
- avg_inter_order_days – mean gap between consecutive orders
- std_inter_order_days – std dev of that gap (0 for single-order customers)
- orders_last_Nd – order count in the last N days (per ``windows``)
- spend_last_Nd – total spend in the last N days
- ratio_recent_historical – orders_last_30d / (orders_last_90d + 1)
Args:
orders: Full orders fact table.
reference_date: Features are computed on orders strictly before this date.
windows: List of rolling-window sizes in days.
Returns:
DataFrame with one row per customer.
"""
hist = orders[orders["order_date"] < reference_date].sort_values(["customer_id", "order_date"])
# Inter-order gap stats
hist = hist.copy()
hist["prev_date"] = hist.groupby("customer_id")["order_date"].shift(1)
hist["inter_days"] = (hist["order_date"] - hist["prev_date"]).dt.days
inter_stats = (
hist.groupby("customer_id")["inter_days"]
.agg(avg_inter_order_days="mean", std_inter_order_days="std")
.reset_index()
)
inter_stats["avg_inter_order_days"] = inter_stats["avg_inter_order_days"].fillna(0).round(2)
inter_stats["std_inter_order_days"] = inter_stats["std_inter_order_days"].fillna(0).round(2)
result = inter_stats.copy()
# Rolling-window counts and spend
for w in sorted(windows):
w_start = reference_date - pd.Timedelta(days=w)
window_orders = hist[hist["order_date"] >= w_start]
cnt = window_orders.groupby("customer_id").size().reset_index(name=f"orders_last_{w}d")
spend = (
window_orders.groupby("customer_id")["total_value"]
.sum()
.reset_index()
.rename(columns={"total_value": f"spend_last_{w}d"})
)
result = result.merge(cnt, on="customer_id", how="left")
result[f"orders_last_{w}d"] = result[f"orders_last_{w}d"].fillna(0).astype(int)
result = result.merge(spend, on="customer_id", how="left")
result[f"spend_last_{w}d"] = result[f"spend_last_{w}d"].fillna(0.0).round(2)
# Ratio recent vs historical
if "orders_last_30d" in result.columns and "orders_last_90d" in result.columns:
result["ratio_recent_historical"] = (
result["orders_last_30d"] / (result["orders_last_90d"] + 1)
).round(4)
else:
result["ratio_recent_historical"] = 0.0
logger.info(f"compute_time_features: {len(result):,} customers, windows={windows}")
return result
|