@logistics-ts/core

Foundational layer of logistics-ts: canonical record types, CSV/DB-row loaders, time bucketization, shared numerics, the Explained<T> result wrapper, and synthetic demo data. Zero runtime dependencies.

Install

npm i @logistics-ts/core

What's in it

• Records: DemandRecord, StockRecord, LeadTimeRecord, plus loadDemand/loadStock/loadLeadTimes to map arbitrary rows onto them.
• bucketize: turns raw demand into a dense, zero-filled per-item time series — the shape every forecasting/classification function expects.
• Numerics: mean, standardDeviation, variance, coefficientOfVariation, averageDemandInterval, inverseNormalCdf, normalCdf/normalPdf, nelderMead — hand-rolled and golden-tested against authoritative values, not a wrapped stats library.
• generateExampleData: a reproducible synthetic { demand, stock, leadTimes } catalogue, for demos and tests.
• Explained<T> / explain(): the result-wrapper contract used across every logistics-ts package — { value, method, inputs, reasoning, citations?, warnings? }.

Quick start

import { bucketize, generateExampleData, mean, standardDeviation } from '@logistics-ts/core'

const { demand } = generateExampleData({ items: 3, seed: 1 })
const series = bucketize(demand, 'month')[0]
const quantities = series.buckets.map((b) => b.quantity)

console.log(mean(quantities), standardDeviation(quantities))

In the umbrella package

@logistics-ts/core is re-exported as the core namespace from logistics-ts. Install this scoped package directly when you only need the data model and numerics — no forecasting, classification, or inventory logic.

Links

• Full docs, API map, and examples
• Other logistics-ts packages

License

MIT