v0.7.8

13/03/2024

• Added prc() which enables easy building of precision-recall curves from 'nestedcv' models and repeatcv() results.
• Added predict method for cva.glmnet.
• Removed magrittr as an imported package. The standard R pipe |> can be used instead.
• Added metrics() which gives additional performance metrics for binary classification models such as F1 score, Matthew's correlation coefficient and precision recall AUC.
• Added pls_filter() which uses partial least squares regression to filter features.
• Enabled parallelisation over repeats in repeatedcv() leading to significant improvement in speed.