LeadTimeAnalysisEvents

Lead times between order placement and delivery events, their statistics, and abnormally long deliveries.

ts_shape.events.supplychain.lead_time_analysis · Production guide · API reference

Example

import pandas as pd
from ts_shape.events.supplychain.lead_time_analysis import LeadTimeAnalysisEvents

orders = pd.DataFrame({"systime": pd.date_range("2024-01-01", periods=8, freq="1D"), "uuid": "po_placed",
                       "value_string": [f"PO-{i}" for i in range(8)]})
deliveries = orders.assign(uuid="po_received",
                           systime=orders["systime"] + pd.to_timedelta([3, 4, 4, 5, 5, 12, 4, 4], unit="D"))
df = pd.concat([orders, deliveries], ignore_index=True)

lt = LeadTimeAnalysisEvents(df)
lead_times = lt.lead_times("po_placed", "po_received")
print(lt.lead_time_statistics("po_placed", "po_received"))
late = lt.detect_lead_time_anomalies("po_placed", "po_received", threshold_factor=2.0)

Methods

Method	Returns
lead_times(order_uuid, delivery_uuid)	per order: order_time, delivery_time, lead_time_seconds, lead_time_hours, order_id
lead_time_statistics(order_uuid, delivery_uuid)	one row: mean_hours, std_hours, min_hours, max_hours, p95_hours, count
detect_lead_time_anomalies(order_uuid, delivery_uuid, threshold_factor=2.0)	lead times above mean + threshold_factor × std: order_time, delivery_time, lead_time_hours, z_score (systime = delivery time)

Notes

• The order identifier is read from value_string (value_column= to change). When the deliveries carry identifiers found among the orders, each order is matched to the first delivery with the same identifier at or after it, and unmatched orders are left out; otherwise each order, in time order, is paired with the first not yet used delivery at or after it. Orders and deliveries left without a partner are dropped with a DataQualityWarning; a lead time is never negative.
• For per-supplier statistics, use one signal pair per supplier or filter the frame first.
• calculate_lead_times is the deprecated name of lead_times.