Skip to content

DemandPatternEvents¤

Analyze demand patterns from a timeseries demand signal.

Module: ts_shape.events.supplychain.demand_pattern Guide: Production Guide

When to Use¤

Use to understand demand patterns for production planning. Identifies spikes, seasonal patterns, and trends that affect scheduling and inventory management. Works with order counts, request volumes, or any signal representing demand intensity over time.

Quick Example¤

from ts_shape.events.supplychain.demand_pattern import DemandPatternEvents

analyzer = DemandPatternEvents(
    df=orders_df,
    timestamp_col="timestamp",
    value_col="order_quantity"
)

# Aggregate demand per period
daily_demand = analyzer.demand_by_period(period="1D", agg="sum")

# Detect abnormal demand spikes
spikes = analyzer.detect_demand_spikes(
    window="7D", threshold_std=2.5
)
print(f"Detected {len(spikes)} demand spikes")

# Analyze day-of-week and hour-of-day seasonality
seasonality = analyzer.seasonality_summary()
print("Busiest day:", seasonality["day_of_week"].idxmax())
print("Peak hour:", seasonality["hour_of_day"].idxmax())

Key Methods¤

Method Purpose Returns
demand_by_period() Aggregate demand per time period DataFrame with total/mean demand per period
detect_demand_spikes() Flag abnormal demand levels DataFrame of spike events with magnitude
seasonality_summary() Day-of-week and hour-of-day patterns DataFrame with average demand by time component

Tips & Notes¤

Remove known promotions before seasonality analysis

Promotional events create artificial spikes that distort seasonality_summary(). Filter out known promotion dates or use detect_demand_spikes() first to identify and exclude outliers before analyzing seasonal patterns.

Related modules

See Also¤