KernelStats
Open source & free forever

Statistical analysis
for everyone

Everything Minitab, JMP, and SAS do — without the $8,700 price tag. Built by a publishing statistician. Runs in your browser.

Launch Kernel → View on GitHub
132
Statistical methods
10
Model families
7
ML algorithms
$0
Forever
Every tool a statistician needs.
Nothing they don't.
From data cleaning to GARCH volatility models — Kernel covers the full analysis pipeline with plain-English interpretations at every step.
🩺

Variable Doctor

15 automated diagnostics detect coded missing values, numeric-as-text, date columns, outliers, and ID columns. One-click fixes with full undo.

🧪

20+ hypothesis tests

Parametric and nonparametric — t-tests, ANOVA, chi-square, Mann-Whitney, Kruskal-Wallis. With a Test Finder wizard that recommends the right test.

📈

Regression & GLMs

10 model families: linear, logistic, probit, multinomial, ordinal, Poisson, negative binomial, zero-inflated, hurdle, and bias-reduced.

📉

Time series & GARCH

ARIMA with auto-selection, GARCH(1,1), GJR-GARCH, EGARCH. Conditional volatility plots, news impact curves, and multi-step forecasting.

🤖

Machine learning pipeline

7 algorithms with auto-detection of classification vs regression. Model comparison dashboard, ROC curves, confusion matrices, variable importance.

🫀

Survival analysis

Kaplan-Meier curves with log-rank tests, Cox proportional hazards, hazard ratio forest plots, concordance index, and Schoenfeld PH tests.

🏭

Industry templates

SPC control charts with Cp/Cpk, RFM customer segmentation, Pareto analysis, clinical heatmaps, volcano plots. Ready for quality, business, and healthcare.

🎓

Education layer

Plain-English interpretations for every test result. Built to teach statistics, not just compute it. Perfect for classrooms and self-learners.

📝

Professional reports

Generate Word or HTML reports with data overview, diagnostics, analysis results, and visualizations. One click from analysis to deliverable.

The features you need.
Without the enterprise invoice.
See how Kernel compares to the commercial tools your organization is already paying for.
Feature Kernel Minitab JMP SAS
Annual cost Free $1,700 $1,785 $8,700
Open source
Runs in browser
Variable Doctor (auto-diagnostics)
Test Finder wizard
GARCH / volatility models
Machine learning pipeline
Survival analysis
Control charts / SPC
Plain-English interpretations
Report generation
From raw data to insights in minutes
1

Upload

CSV, Excel, SPSS, Stata, SAS, or paste data directly

2

Diagnose

Variable Doctor auto-detects 15 types of data issues

3

Analyze

Run tests, fit models, build ML pipelines

4

Report

Generate professional Word or HTML reports

Built by a statistician,
for statisticians.
Anjana Yatawara, Ph.D.

Anjana Yatawara, Ph.D.

Assistant Professor of Mathematics & Statistics
California State University, Bakersfield

KernelStats was born from a simple frustration: students and researchers shouldn't need $8,700/yr software to run a t-test. I built KernelStats to be the tool I wished existed when I started my career -- powerful enough for publication-grade analysis, intuitive enough for a first-year statistics student.

My research spans financial volatility modeling (GARCH, GJR-GARCH, EGARCH), environmental justice, and AI in statistics education. I've published in Environmental Research: Health, ACS ES&T Air, and IEEE FIE, with papers targeting the Journal of Business & Economic Statistics and Journal of Forecasting.

yatawara.com GitHub Google Scholar

Ready to try Kernel?

No installation. No account. No credit card.
Just open the app and start analyzing.

Launch Kernel → Star on GitHub ⭐

Coming soon: Kernel Pro — AI-powered analysis with Claude integration, team collaboration, and enterprise deployment.