Cat_boost encoder
xuanqing94 opened this issue · 2 comments
xuanqing94 commented
That does't look like an interpolation of mean of global vs. mean of subgroup. I think it should be something like:
level_means = ((colmap['sum'] + self._mean * self.a) / (colmap['count'] + self.a)).where(level_notunique, self._mean)
glevv commented
Yep, this commit forget to add multiplication by a. It's a bug, good catch.
Also could be that this line also needs fix?
PaulWestenthanner commented
fixed by #339
Thanks :)