ChakataStat
Download

ChakataStat at a glance

This page is the whole application in one read. It is deliberately short: it tells you what ChakataStat is, what each part of the window does, and where to go next for detail. If you have never opened the app before, read this first, then follow Getting Started for your first analysis.

What ChakataStat is

ChakataStat is a desktop application for statistical data analysis. It pairs a spreadsheet-style data editor with a fast statistics engine, so you can type or import a table of data, describe and model it, draw charts, and save the whole thing as a single document — without writing any code.

Two things make it more than a spreadsheet:

  • A real statistics engine. Everything from a frequency table to ARIMA forecasting, factor analysis and Cox regression runs natively, on a background worker so the window never freezes.
  • One command for every surface. Every analysis you run from a menu can also be run by typing a command, asked for in plain language from the built-in AI assistant, or driven by an external program. They are all the same underlying operation, so anything you do by clicking can be reproduced exactly later.

The workbench, part by part

When you open ChakataStat you see the workbench — one window with these pieces:

Part What it is Where it is
Menu bar File, Edit, View, Data, Transform, Analyze, Graphs, Help Top
Toolbar One-click buttons for the most common actions Under the menu
View tabs Switch between Data, Variable, Output and Syntax Left of the work area
Work area The active view — your data, your variable definitions, or your results Centre
AI assistant panel Ask for analyses in plain language Right (toggle with Ctrl+Shift+A)
Status bar Case count, active filters, engine state Bottom

The four views

You spend all your time in one of these tabs. Switch with the tabs or Ctrl+1…Ctrl+4.

  • Data View (Ctrl+1) — the spreadsheet. Rows are cases (observations); columns are variables. This is where you type, paste and edit data.
  • Variable View (Ctrl+2) — one row per variable, where you set each variable's name, type, label, value labels, measurement level and missing-value rules. Defining variables well here makes every analysis and chart read better.
  • Output View (Ctrl+3) — the log of results. Every analysis and chart you run appends a titled card here. You can copy, export, or clear it.
  • Syntax console (Ctrl+4) — advanced. The editable command stream that records what you have done and lets you re-run it. See Reproducibility.

(Earlier versions had a fifth, Advanced-mode SQL tab; it has been removed — see Direct engine access.)

What you can do — the menus in one line each

  • File — new dataset, open/save .ckd documents, import/export CSV, Excel and Parquet, import/export a data dictionary, export the output log.
  • Edit — undo/redo, cut/copy/paste, select all, find & replace (Ctrl+F).
  • View — switch views, open the Command Palette (Ctrl+Shift+P), toggle the AI panel, and (in Advanced mode) Engine Diagnostics.
  • Data — define variable properties, sort, select / weight / split cases, merge files, transpose, bulk-rename variables, define variable sets.
  • Transform — compute a new variable from an expression, recode, rank cases.
  • Analyze — the statistics: Descriptives, Compare Means, Correlate, Regression, the General Linear and Mixed models, Nonparametric tests, Scale, Dimension Reduction, Classify, Meta-Analysis, Survival, Time Series and custom Tables.
  • Graphs — Histogram, Scatterplot, Boxplot and a free-form Chart Builder.
  • Help — documentation and the About box.

How a session usually flows

  Open or import data ─▶ Define variables ─▶ (optionally) filter / weight / split
        │                                                  │
        ▼                                                  ▼
   Transform / compute new variables ───────────▶  Run analyses & graphs
                                                          │
                                                          ▼
                                          Read the Output log ─▶ Export / save .ckd

Everything you run is also recorded as a re-runnable command in the Syntax console, so a finished analysis can be replayed on new data with one click.

The advanced features, in brief

These have their own sections because they reward a closer read:

  • The AI assistant — connect a model (Claude or any OpenAI-compatible endpoint) and ask for analyses in plain English; results come back as real tables and charts in the chat.
  • Reproducibility — the .cks syntax stream, the command palette, headless batch runs (--run), exporting the output log to HTML/Markdown/LaTeX, and exporting a session as a Python script.
  • Direct engine access — the engine diagnostics panel, Advanced mode, and driving ChakataStat from another program over MCP.

Where to go next