A MATLAB/Octave function for adjusting p-values for multiple comparisons. Given a set of p-values, returns p-values adjusted using one of several methods: 'holm', 'hochberg', 'hommel', 'bonferroni', 'BH', 'BY', 'fdr', 'sidak' or 'none'.
This is an implementation of the p.adjust R function, the documentation of which can be found at http://www.inside-r.org/r-doc/stats/p.adjust. Contrary to the R function, this function does not handle missing values, and adds one additional correction method, 'sidak', as described in https://en.wikipedia.org/wiki/%C5%A0id%C3%A1k_correction.
If you use this script in your research please cite the following paper:
* Fachada N, Rosa AC. (2018). micompm: A MATLAB/Octave toolbox for multivariate independent comparison of observations. Journal of Open Source Software. 3(23):430. https://doi.org/10.21105/joss.00430
Fachada N, Rosa AC. (2018). micompm: A MATLAB/Octave toolbox for multivariate independent comparison of observations. Journal of Open Source Software. 3(23):430. https://doi.org/10.21105/joss.00430
Dear Giuseppe, thank you for the feedback. I've uploaded a corrected version. Best regards.
This is a very good idea.
However ... it is an error in the Holm's algorithm.
Try it with these test values:
rawpval=[0.0020 0.0045 0.0060 0.0080 0.0085 ...
0.0090 0.0175 0.0250 0.1055 0.5350];
Fix paper reference
- Add paper reference
Fix bug in Holm method (acknowledgements to Giuseppe Altieri for finding bug)
Update description, add Sidak method.
Fix typo in description.