ENH: Setting diagonal to 1 in corr function

[dhalpern]

Is your feature request related to a problem?

I often want to use the corr function to also compute distances between columns (e.g. df.corr(method = spicy.spatial.distance.euclidean). However, I noticed that the .corr function always sets the i=j case to 1 (which is incorrect for the distance case) (see here).

Describe the solution you'd like

I realize using this function to compute distances is sort of a non-standard usage but is there any reason to hardcode it? Would it make more sense as a parameter that could be set?

[simonjayhawkins]

Thanks @dhalpern for the report

see also #25732, #25781

closing as duplicate