Why only "Pearson's" method for correlation? about inspectdf

[alastairrushworth]

Hi @gd047, the short answer is that stat tests for comparing pairs of spearman & kendal require a different type of test - maybe a bootstrap. That's not necessarily a huge problem & I agree it would be useful. I'll add this to the list of features to add to the next release. Thanks!

Just for the record, I use corr.test {psych} and holm's adjustment for multiple tests.

[alastairrushworth]

Hi @gd047, a wee update. As of v0.0.4 (now on CRAN), you can request Kendall or Spearman correlation using

df %>% inspect_cor(method = "kendall")

For now, only the point estimates are returned or plotted. Tests and confidence intervals are not implemented yet, I'll leave this issue open until that is added.

[alastairrushworth]

Intervals and tests (based on normal approximation to the Fisher z-transform) for Kendall and Spearman correlations have been added to the development version. This will be rolled into the next CRAN release. Thanks again!