R-scripts to test for correlation between traits under directional evolution

They require the packages ape, ade4, caper, lmtest, tseries

All the functions take as arguments a phylogeny "phy" which has to be provided as an object of type "phylo" and one or two numeric vectors X (and Y) storing the values of the trait(s) at the tips of "phy" (their indices have to be consistent with those of the tips of "phy").

Trend-detection tests

SR_trend.test takes as argument a phylogeny phy and a vector X and returns Pearson's correlation test between the tip values and their times in order to detect a trend in the trait evolution and the results of the diagnostic tests for least squares regression validity conditions (Durbin-Watson's, Harrison-McCabe's and Jarque-Bera's tests).

hR_trend.test takes as argument phy and X and returns the test of the nullity of the slope when regressing the independent contrasts on the "h_k", in order to detect a trend in the trait evolution. This function also provides the results of the diagnostic tests for least squares regression validity conditions (Durbin-Watson's, Harrison-McCabe's and Jarque-Bera's tests).

Correlation tests

SR_corr.test, IC_corr.test, PGLS_corr.test, DC_corr.test, PGLSt_corr.test and MR_corr.test take as argument a phylogeny phy and two vectors X and Y and implement the SR, IC, DC and MR tests as described in the manuscript. They return the corresponding test of the nullity of the slope in the associated regression in order to test for correlation. These functions also provide the results of the diagnostic tests for least squares regression validity conditions (Durbin-Watson's, Harrison-McCabe's and Jarque-Bera's tests).

The script in the file "simulations.R" compute all the figures of the manuscript.