Principal component analysis (PCA) is a widely used tool to handle high-dimension data. As typically high-dimension data, grain-size distributions also could be processed by PCA. QGrain has integrated the basic PCA to extract the major variations of GSDs. The interface is very simple:
- Just click the
Load Dataset
button to load your GSDs. - Select the number of PCs.
- Click the
Perform
button to execute the PCA algorithm. - Check the explained variances of PCs (the percentages in the legend) to determine whether to add/reduce a component. If the sum of them is less than 90%, it means the number is not enough. If some minor components have very small variances, consider reducing the number.
- If you are sure the number is appropriate, click the
Save
button to save the PCA result to an Excel file.