OptiGein Design Docs

OptiGein Design

Plan an experiment, run it, and let the data tell you where the best settings are. This guide walks through every screen of the software, from picking a design to reporting the optimum you found.

OptiGein Design is a Design of Experiments (DOE) application for Windows and Linux, built by Geinforce Technology Private Limited. It covers the whole cycle of an experiment: choosing a design that answers your question with as few runs as possible, entering the results, fitting and checking a model, and then reading the optimum settings off that model.

New to Design of Experiments? Start with A tour of the software, then follow the response surface tutorial. Between them they use every screen you are likely to need, on one small worked example.
Prefer to read offline or print it? The whole guide is also available as a single PDF user manual, with a contents list, an index and every screenshot. It is generated from these pages, so the two never disagree.

Start here

Install the software

What the installer does, where files go, the .ogdx file type, and how to update later.

Activate your license

Enter your keys or sign in, work offline, move to a new computer, or renew when the period ends.

Take the tour

The welcome screen, the design wizard, the run sheet, and the analysis workspace.

Pick the right design

Screening, characterisation or optimisation? A short decision guide with run counts.

What the software does

Build a design

Sixteen design types are grouped into four families. Factorial designs sort the important factors from the unimportant ones. Response surface designs map curvature and locate an optimum. Mixture designs handle recipes, where the ingredients must add up to a fixed total. Custom designs cover everything else, including a blank sheet and importing data you already have.

The New Design screen listing every design family
Every design the software can build, grouped by what it is good for.

Analyse the results

Type your measurements into the run sheet, press Start Analysis, and the software fits a model, suggests the order that the data supports, and lays out the ANOVA table with plain language notes telling you what each number means. Residual diagnostics tell you whether the model can be trusted.

The ANOVA tab with fit statistics and the fitted equation
The ANOVA table, fit statistics and the fitted equation, all on one screen.

Find the best settings

Set a goal for each response, such as maximise yield while holding cost in range, and the optimizer searches the fitted models for the settings that satisfy all of them at once. The graphical route shades the region where every response stays inside its limits.

Contour graph showing the optimum point
The optimum marked on the response contours.

How this guide is organised

If you want toRead
Install and activate the softwareInstalling and Activation
Understand what each screen is forA tour of the software
Decide which design fits your questionWhich design do I need?
Look up the settings of one design typeThe four design pages, starting with Factorial
Read an ANOVA table with confidenceFit summary, model and ANOVA
Check whether your model is trustworthyDiagnostics
Explain the model to somebody elseModel graphs and Exporting
Get numbers for the best settingsNumerical optimization
Follow a complete worked exampleResponse surface tutorial
Watch it done on screenVideo tutorials
Move to a newer versionKeeping the software up to date

Getting help

If something in the software does not behave the way this guide describes, please tell us. Include the design type and, if you can, the saved .ogdx file.