Track D Student Edition (Workbook Landing)

This page is the front door for Track D (Business Statistics with an accounting case study).

If you do only one thing first, do this:

  1. Follow the workbook quickstart.

  2. Run d00_peek_data to see the data.

Big picture

If you want a quick map of how the Track D pieces fit together, read:

  1. Run d01 to see the accounting invariants.

  2. Use the “skill map” below to keep your bearings.

Note

Track D is designed to make you a better analyst of accounting data. You will learn how to trust the numbers (quality control), summarize them (statements and KPIs), and make decisions with them (inference, regression, forecasting, scenarios).

Where to start (PyPI-only)

Install and create the Track D workbook:

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

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

Then run these two “confidence builders”:

pystatsv1 workbook run d00_peek_data
pystatsv1 workbook run d01

Helpful pages in the Workbook docs

These pages live inside the workbook documentation subtree (they build cleanly on their own):

What you are building (the pipeline)

Track D is a repeatable analysis workflow. You are not just “running scripts.” You are learning how to take messy accounting-like events and turn them into a decision-ready story.

Here’s the mental model:

Events (sales, bills, payroll, inventory, loans)
    ↓  (recording rules + checks)
General Ledger (journal entries)
    ↓  (postings → trial balance)
Financial Statements (IS, BS, CF)
    ↓  (descriptive stats, visual checks)
Decisions (risk, sampling, tests, regression)
    ↓  (forecasts, scenarios, governance)
Communicate (clear memo + reproducible outputs)

The goal: trustworthy numbers + clear decisions.

Why Track D makes you a better analyst

Track D trains three “analyst superpowers” that matter in real accounting and finance work:

  1. Data integrity (trust the numbers) - You learn the invariants that must hold (balanced entries, consistent statements). - You learn how to spot red flags (duplicates, missingness, impossible values, broken joins).

  2. Decision discipline (answers with uncertainty) - You learn how to quantify risk (probability). - You learn how to estimate and test (sampling + hypothesis testing). - You learn how to model drivers (regression) without overclaiming.

  3. Communication (results people can act on) - You learn how to tell a coherent story from the data. - You learn to separate “signal” from “noise” and explain limits honestly. - You learn reproducible workflows that other people can audit.

Skill map (D00–D23)

Use this map to understand why each group of chapters exists.

Phase 0 — On-ramp (see the data)

  • D00: Setup/reset datasets and peek at the datasets (what tables exist, what they look like)

Phase 1 — Accounting foundations (what the numbers mean)

  • D01–D06: journal entries, chart of accounts, statements logic, reconciliations, and quality control

Phase 2 — Data preparation (make the dataset analysis-ready)

  • D07: build analysis tables, document joins/keys/grain, verify quality checks

Phase 3 — Describe performance + report responsibly

  • D08–D09: descriptive statistics and reporting conventions (what to compute, how to present it)

Phase 4 — Statistics for decisions (business lens)

  • D10–D14: probability/risk, sampling/estimation, hypothesis testing, correlation vs causation, regression drivers

Phase 5 — Forecasting + governance

  • D15–D23: forecasting hygiene, seasonality, drivers, cash flow, integrated scenarios, and communication/governance

How to use Track D week-to-week

A good weekly rhythm:

  1. Run the chapter script (pystatsv1 workbook run dXX).

  2. Open what it writes in outputs/track_d/ (tables + summaries). If you are not sure what a file is for, use Track D Outputs Guide.

  3. Answer the chapter questions in words (what changed, why, what action follows).

  4. Run the smoke checks (pystatsv1 workbook check business_smoke).

pystatsv1 workbook run d08
pystatsv1 workbook check business_smoke

Common “student mistakes” and what to do

“I ran it, but I don’t know what it means.” - Start at Track D Lab 0/1 (PyPI-only) and follow the interpretation prompts. - Re-run d00_peek_data and read the previews slowly.

“My outputs differ from the handout / screenshots.” - Confirm you are using the canonical datasets (seed=123) under data/synthetic/. - If you edited anything in data/synthetic/, reset:

pystatsv1 workbook run d00_setup_data --force

“I want to apply this to my own data.” - That’s the endgame. After you complete the basics, you’ll use a “bring your own data” playbook (coming next in the workbook docs) that shows how to map real exports (QuickBooks/bank/invoices) into the same workflow.

What “good” looks like by the end

By the end of Track D you should be able to:

  • Explain how accounting events become analysis tables (and what can go wrong).

  • Produce a monthly trial balance and statements and sanity-check them.

  • Compute KPIs and explain what drives changes (not just “the number changed”).

  • Use estimation, tests, and regression to support a recommendation.

  • Produce a simple forecast and scenario analysis with clear assumptions.

  • Write a short memo that a manager could actually use.

How to apply Track D to your own data

For a general starting point, see My Own Data: a mini-guide.

For Track D specifically:

Next page

When you’re ready, jump to: