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. Truncated SVD keeps the largest singular directions of the scaled table. These runs use the randomized algorithm with 5 iterations, and the published share is the explained variance of that fit.
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 truncated SVD.
Trustworthiness is within 0.01 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; Truncated SVD reports the same explained variance on these three tables.
Do you need the variance those axes keep?
PCA reports that share. Truncated SVD reports the explained variance of the randomized fit.
PCA on Iris Flowers (150 rows, 4 features) keeps 95.8% of the variance in two axes, trustworthiness 0.974. Truncated SVD with the randomized algorithm and 5 iterations reports 95.8% explained variance and trustworthiness 0.974. Points are colored by held-out species. The two maps are within 0.01 trustworthiness.


PCA on Wine (UCI wine recognition) (178 rows, 13 features) keeps 55.4% of the variance in two axes, trustworthiness 0.871. Truncated SVD with the randomized algorithm and 5 iterations reports 55.4% explained variance and trustworthiness 0.871. Points are colored by held-out wine_class. The two maps are within 0.01 trustworthiness.


PCA on Customer Segments (200 rows, 5 features) keeps 90.5% of the variance in two axes, trustworthiness 0.927. Truncated SVD with the randomized algorithm and 5 iterations reports 90.5% explained variance and trustworthiness 0.927. Points are colored by held-out segment. The two maps are within 0.01 trustworthiness.


| Dataset | Method | Trustworthiness | Also reported |
|---|---|---|---|
| Iris Flowers | PCA | 0.974 | 95.8% variance in 2 components |
| Iris Flowers | Truncated SVD | 0.974 | 95.8% explained variance from truncated SVD |
| Wine (UCI wine recognition) | PCA | 0.871 | 55.4% variance in 2 components |
| Wine (UCI wine recognition) | Truncated SVD | 0.871 | 55.4% explained variance from truncated SVD |
| Customer Segments | PCA | 0.927 | 90.5% variance in 2 components |
| Customer Segments | Truncated SVD | 0.927 | 90.5% explained variance from truncated SVD |
PCA keeps the directions of greatest variance as a linear summary of the columns. Truncated SVD keeps the largest singular directions of the scaled table. These runs use the randomized algorithm with 5 iterations, and the published share is the explained variance of that fit.
Method: StandardScaler, drop constant columns. Comparison code version 8fbe69461da4 (scripts/seo/build_compare_runs.py). scikit-learn 1.4.2. Seed 42. Truncated SVD 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.