PCA software · Mac & Windows
Decide how many components to keep on evidence rather than on a default. Eigenvalues, variance explained per component, the full loadings table, a scree plot and a biplot with loading vectors — the whole picture, every time. Native on Mac and Windows, bought once. Free to try on your own data before you decide.
Most of the established packages are sold as an annual subscription — often quoted only after a sales call, and multivariate methods are frequently a priced add-on on top. Whatever the number turns out to be, it returns every year.
Five years, five bills — and no software left if you stop.
Bought once. The version you buy keeps working forever, online or off, on Mac and Windows alike.
$249 for Professional — once, and it’s yours.
| Capability | In Stratum |
|---|---|
| Extraction basis | Correlation or covariance matrix, your choice |
| Sampling adequacy | KMO overall and per-variable MSA, with Bartlett’s test of sphericity — so you can say whether the data was worth reducing |
| Eigenvalues | Per component, with variance explained and cumulative share |
| Loadings | Full table — every variable on every component |
| Scree plot | Variance explained per component, plotted |
| Biplot | PC1 against PC2 with loading vectors over the scores |
| Component scores | Written back as columns for downstream analysis — regression, clustering or a chart |
| Degenerate cases | Constant variables and singular matrices reported by name rather than returning a silent result |
| Linked selection | Lasso a cluster of scores on the biplot and those rows highlight in the datasheet and in every other chart |
| Related multivariate | Clustering, correspondence analysis, MANOVA |
| Numerical validation | Published validation record — the outside tool named for every statistic |
| Pricing | One-time (Professional $249) — no subscription |
PCA — components extracted on a correlation or covariance basis, your choice, because standardizing changes the answer.
Scree plot & variance explained — eigenvalues and the cumulative share per component, so the cut is yours rather than an unseen default.
Loadings table — every variable’s weight on every component, in full, so a component gets a name instead of a number.
Biplot — PC1 against PC2 with loading vectors drawn over the scores, so which variables drive which axis is visible at a glance.
Component scores — written back as columns, ready for a regression, a cluster analysis or a chart.
Clustering — k-means and hierarchical with a dendrogram, on the components or on the raw variables.
Correspondence Analysis — row and column coordinates, inertia by dimension and a symmetric biplot, for contingency tables.
Regression, ANOVA, trees — the rest of the toolkit in the same app, so PCA is not a silo.
Point-and-click throughout — and equally native on a Mac, where most multivariate tools are Windows-first.
PCA is in Stratum’s Professional edition together with clustering, correspondence analysis, MANOVA, decision trees and SPC — on top of everything in Standard, including descriptive statistics, ANOVA, correlation and contingency, and linear and multiple regression with full diagnostics. Import your .csv, .xlsx or .sav and try it on your own data first.
Yes — eigenvalues with variance explained and cumulative share per component, and the full loadings table rather than a truncated top-few. Understanding PCA loadings walks through reading one.
Either, and you choose. Standardizing changes which variables dominate, so it is a decision the analysis surfaces rather than makes for you.
The scree plot and the cumulative variance table are both reported, so you can cut on the elbow or on a variance threshold. Interpreting a PCA works through an example.
Yes — component scores are written back as columns, so they feed straight into a regression, a clustering run or a chart.
Yes — Stratum is a native app for both macOS 14+ and Windows 10/11, so PCA works the same on either.
No — Stratum is a one-time purchase. PCA is in Professional ($249), and you can upgrade from Standard later.
Stratum