ChakataStat User Guide

A task-oriented manual for people running analyses — separate from the project overview at chakatastat.app, which describes the architecture. The goal: a non-developer can learn ChakataStat end to end from this guide, without reading the source.

Status: written. The pages below are the user manual. They are organised so a brand-new user can read one short page for the whole picture, then drop into detailed task pages as needed, with the advanced features kept in their own section. Worked examples use the bundled sample dataset (File → Open Sample Dataset); screenshots, as they are added, live under images/.

Start here

• Quick Start — install to your first result in about ten minutes, on the bundled sample dataset. The fastest way in.
• ChakataStat at a glance — the whole application in one read: what it is, every part of the window, and what each menu does.
• Getting Started — install, the workbench tour, and your first analysis in five minutes.
• Tutorial Workbook — a ten-module course on the bundled sample health study (see its codebook), for self-study or classroom training, with exercises and the reproducible syntax for every module.

Core tasks (the detail)

• Working with Data — importing/exporting CSV, Excel and Parquet; the Data View and Variable View; value labels, missing values and measurement levels; the .ckd document; find & replace; the data dictionary; bulk rename and variable sets.
• Case Semantics & Transforms — Select / Weight / Split cases, and the Compute / Recode / Rank / Sort / Merge / Transpose transforms.
• Analyzing Data — every procedure, grouped as the menu is. When to use it, how to fill the dialog, how to read the output:
  • Descriptive Statistics
  • Compare Means
  • Correlation & Regression
  • Meta-Analysis
  • Nonparametric Tests
  • Scale, Dimension Reduction & Classify
  • Confirmatory Factor Analysis
  • Survival & Time Series
  • Custom Tables
• Graphs — Histogram, Scatterplot, Boxplot, the Chart Builder and analysis-attached plots.
• Settings — appearance, connectivity, the AI assistant and Advanced mode.

Advanced features (their own section)

Optional power-user and developer features, in advanced/:

• The AI Assistant — connect a model backend and ask for analyses in plain language.
• Reproducibility — the .cks syntax console, the command palette, headless batch runs (--run), output export to HTML/Markdown/LaTeX, and Export as Python.
• Direct Engine Access — Advanced mode, the engine diagnostics panel, and driving ChakataStat over MCP.

Reference

• Choosing a procedure — "what do I have, what do I want to know?" → which analysis.
• Glossary — ChakataStat and statistics terms.

In-app entry points

This guide is reachable from Help → Documentation (and F1) and from the first-run welcome / empty-state surfaces.

Contributing to this guide

Screenshots live under images/ and are referenced with relative paths. Keep each procedure page self-contained and worked against the bundled sample dataset (File → Open Sample Dataset) so examples are reproducible.