## Statistics / PCA

### Description

Calculates principal components for a set of samples.

### Parameters

- Variability to explain (0...100) [80]
- Should data be scaled (yes,no) [no]
- Should data be centered (yes, no) [yes]

### Details

Performs principal component analysis for samples. Principal components
are retained until the explained variability exceeds the specified value.
The output of the analysis can be visualized using the 3D scatterplot in the
visualization panel. To do this, select the pca.tsv spreadsheet and change
the visualization method to "3D scatterplot for PCA".

### Output

The following output files are produced:

- pca.tsv containing the principal components for the samples (these are not expression values).
- variance.tsv where the row "proportion of variance" indicates how much of the variance is explained by each component.
- loadings.tsv containing the component loadings. The higher the absolute value, the more that gene contributes to the difference in the samples along that principal component.