I am experiencing an issue when I use Frechet analysis of variance for data consisting of SPD matrices. The issue is that for samples that I can verify independently to have very similar Frechet means, I get p-values << 0.05 much more often than the nominal rate.
I was wondering if you have ever notice similar behavior. I'm happy to share my code if needed, though this is more of a general question.
Hi,
Your package is great, thanks for your hard work.
I am experiencing an issue when I use Frechet analysis of variance for data consisting of SPD matrices. The issue is that for samples that I can verify independently to have very similar Frechet means, I get p-values << 0.05 much more often than the nominal rate.
I was wondering if you have ever notice similar behavior. I'm happy to share my code if needed, though this is more of a general question.
Thanks,
NEV