guess_material() was not considering the '_he' in the material name, which was causing that efficiency and smearing files were not found, and those matrices were set to ones. This was producing no error in the ar40kt_he case, but it was not properly running with he eff and smear files.
for the case of wc100kt30prct_he detector, the oxygen files don't contain '_he' in the channel, but ibd and nu-e ES do. So for the oxygen case, the file name was correctly found, but there was a missmatch with the rates binning, which was set only for the non he case.
This should close issue #205
The problem was the following: