Track D – Business Statistics & Forecasting for Accountants

Track D teaches business statistics and forecasting to learners who work with accounting-shaped data (bookkeeping, AP/AR, payroll, staff accounting, finance ops).

The core idea is simple:

Don’t just calculate results — engineer them. We treat statistical analysis like production software:

  • reproducible inputs and deterministic outputs,

  • automated checks (tests) so you can trust what you share,

  • “audit-friendly” artifacts (tables, figures, memos) that support decisions.

Run Track D as a Workbook (PyPI-only)

If you’re a student following Track D, the easiest path is the Track D Workbook. It installs from PyPI and generates a self-contained folder (scripts + tests + data).

python -m venv .venv
# Windows (Git Bash)
source .venv/Scripts/activate
python -m pip install --upgrade pip
python -m pip install "pystatsv1[workbook]"

pystatsv1 workbook init --track d --dest track_d_workbook
cd track_d_workbook

# Start here
pystatsv1 workbook run d00_peek_data
pystatsv1 workbook run d01

# Optional: run the included smoke test
pystatsv1 workbook check business_smoke

The workbook ships canonical datasets (seed=123) under data/synthetic/ and writes chapter outputs to outputs/track_d/. See Track D Workbook: Business Statistics for Accounting Data for the dataset map, output conventions, and tips.

The running case and data

Track D uses a synthetic, accounting-shaped dataset family generated by the PyStatsV1 simulator scripts (sometimes referred to as LedgerLab).

There are two dataset “sizes” used across Track D:

  • LedgerLab core datasets (Chapters 1–3): small ledgers used to teach double-entry, the GL, and statements as summaries. You will see folders like data/synthetic/ledgerlab_ch01.

  • North Shore Outfitters (NSO v1) running case (Chapters 4+): a richer, realistic small business scenario used for operational + financial analysis, written to data/synthetic/nso_v1.

Everything is generated locally, so you can modify assumptions and rerun the book.

Reproducibility quick start

Inputs and outputs follow two conventions:

  • Inputs live in data/synthetic/... (generated datasets)

  • Outputs live in outputs/track_d/ (chapter artifacts)

To (re)generate the NSO v1 dataset:

make business-nso-sim
make business-validate

To run an individual chapter (example: Chapter 15):

make business-ch15

Chapter map (high level)

Track D is designed as a practical progression:

  • Ch 1–3: accounting as measurement; double-entry; statements as summary statistics

  • Ch 4–6: accounting subsystems + controls (inventory, payroll/taxes, reconciliations)

  • Ch 7–9: make data analysis-ready; descriptive stats; reporting style contract

  • Ch 10–13: probability, sampling, hypothesis tests, controlled comparisons

  • Ch 14: regression driver analysis (plus deep-dive appendices 14A–14E)

  • Ch 15: forecasting foundations + forecast hygiene (baseline methods + backtesting)

Where to start

Start with Ch 01 — Accounting as a measurement system.