Closed aphalo closed 6 months ago
The convergence failure was in mvtnorm::pmvnorm()
and making maxpts
in the call to summary()
increase with the number of levels in the factor seems to solve the problem, at least up to 10 levels and Tukey contrasts. Fixed.
Add code to catch the error from package 'multcomp' (seems to be failure of convergence) and handle this in a way visible in the plot, or simply not show the test results.
The problem was revelead by tests. See #54.