Principal Component Analysis (PCA)

Principal component analysis is a popular tool for performing dimensionality reduction in a dataset. PCA performs a linear transformation of a dataset (having possibly correlated variables) to a dimension of linearly uncorrelated variables (called principal components). This transformation aims to maximize the variance of the data. In practice, you would select a subset of the principal components to represent your dataset in a reduced dimension.

The Principal Component Analysis card provides a visual representation of a dataset in a reduced dimension.


The PCA card displays a scree plot of eigenvalues for each principal component and the cumulative explained variance (in percentage). The card also displays a scatter plot of the data projected onto the first two principal components and a heatmap that shows the composition of all the principal components.

You can use the PCA menu (⋮) to configure the visualization of the heatmap by toggling the values and colors on and off or choosing to show absolute values.