This is a collection of functions that I used for longitudinal biomarker analyses in cohort studies.
For more information on publications in which these functions are used, please visit my personal website.
The most important functions in this package are those that perform permutation testing between many different variables, using the permutation version of the spearman and wilcoxon tests as implemented in the coin package (see here and here) (coin::spearman_test and coin::wilcoxon_test, respectively). While permutation tests can be computational heavy, by using these non-parametric tests we do not have to make assumptions on the underlying distributions, hence making the tests more robust and reliable.
Other functions included in this package are functions facilitating prediction using random Forest, and functions that help to create publication-ready summary tables.
Important: This is still a work in progress, use at your own risk. Failures can be expected. More functions will be added in the future. Furthermore, I have to add more unit tests to test robustness of some functions.
To do:
- Extend this readme file and description.
- Merge the function association_study with association_study_long?
- Improve random forest function, implement unit tests
- Create package vignettes