PCA
- Form
- Linear.
- Score
- Variance share.
- Components
- A fixed number of components.
Loading…
PCA keeps the directions of greatest variance as a linear summary of the columns. Factor analysis keeps latent factors and a noise variance for each column. These runs use no rotation, and the published share is the column variance in the squared loadings.
Choose the fit the table needs.
This is a choice, not a measured winner.
A new row uses the fitted PCA axes. A later row uses the fitted factor analysis model.
PCA keeps more original neighborhoods on Iris Flowers, Wine (UCI wine recognition), and Customer Segments; PCA still reports the variance those two axes keep: 95.8% on Iris Flowers, 55.4% on Wine (UCI wine recognition), 90.5% on Customer Segments; Factor analysis reports the column variance in the squared loadings: 84.6% on Iris Flowers, 48.4% on Wine (UCI wine recognition), 83.6% on Customer Segments, and mean noise variance 0.154 on Iris Flowers, 0.516 on Wine (UCI wine recognition), 0.164 on Customer Segments.
Do you need the directions of greatest variance?
A variance summary of the columns, or latent factors with a noise variance for each column.
PCA on Iris Flowers (150 rows, 4 features) keeps 95.8% of the variance in two axes, trustworthiness 0.974. Factor analysis uses no rotation. The fit stopped after 125 iterations. It reports 84.6% of column variance in the squared loadings, mean noise variance 0.154, and trustworthiness 0.917. Points are colored by held-out species. PCA keeps more of the original neighborhoods in this picture.


PCA on Wine (UCI wine recognition) (178 rows, 13 features) keeps 55.4% of the variance in two axes, trustworthiness 0.871. Factor analysis uses no rotation. The fit stopped after 23 iterations. It reports 48.4% of column variance in the squared loadings, mean noise variance 0.516, and trustworthiness 0.852. Points are colored by held-out wine_class. PCA keeps more of the original neighborhoods in this picture.


PCA on Customer Segments (200 rows, 5 features) keeps 90.5% of the variance in two axes, trustworthiness 0.927. Factor analysis uses no rotation. The fit stopped after 13 iterations. It reports 83.6% of column variance in the squared loadings, mean noise variance 0.164, and trustworthiness 0.910. Points are colored by held-out segment. PCA keeps more of the original neighborhoods in this picture.


| Dataset | Method | Trustworthiness | Also reported |
|---|---|---|---|
| Iris Flowers | PCA | 0.974 | 95.8% variance in 2 components |
| Iris Flowers | Factor analysis | 0.917 | 84.6% of column variance in the squared loadings, mean noise variance 0.154 |
| Wine (UCI wine recognition) | PCA | 0.871 | 55.4% variance in 2 components |
| Wine (UCI wine recognition) | Factor analysis | 0.852 | 48.4% of column variance in the squared loadings, mean noise variance 0.516 |
| Customer Segments | PCA | 0.927 | 90.5% variance in 2 components |
| Customer Segments | Factor analysis | 0.910 | 83.6% of column variance in the squared loadings, mean noise variance 0.164 |
PCA keeps the directions of greatest variance as a linear summary of the columns. Factor analysis keeps latent factors and a noise variance for each column. These runs use no rotation, and the published share is the column variance in the squared loadings.
Method: StandardScaler, drop constant columns. Comparison code version 8fbe69461da4 (scripts/seo/build_compare_runs.py). scikit-learn 1.4.2. Seed 42. Factor analysis was fit on 2026-11-11. PCA numbers are the published compare run computed on 2026-11-03. The class column was held out of every fit. Iris Flowers: UCI Machine Learning Repository, Iris (Fisher, 1936). License: CC BY 4.0. Wine (UCI wine recognition): UCI Machine Learning Repository, wine recognition. License: CC BY 4.0. Customer Segments: Synthetic customer records bundled with this site. License: Generated for demos; no external license.
Results computed, published, and last updated on 2026-11-11.
The comparison sentences are filled from these published runs by a fixed template, so the scores can be checked against the same datasets in the tool.
Maintainer: nikolaou1982.
Contact: nick1982100@gmail.com.