Yue Zhao

could you provide a script and OS env for this? it is hard to guess, but personally I have never encountered this issue in windows and macOS in the last...

you can replace it by 0 or whichever you want https://numpy.org/doc/stable/reference/generated/numpy.nan_to_num.html Just like sklearn, pyod does not handle missing values. Users are welcomed to convert the data to fully numerical...

I see the problem. I save the enriched feature space. In other words, the entire X plus new features are saved. This will take lots of space. I could introduce...

Hi there, Thanks for reporting this issue. I thought I manually added that layer for symmetry purposes but may be redundant (or even detrimental). I know autoencoders are totally fine...

noted! thanks for mentioning this

Hi there, Sorry for responding late. Unfortunately, this function has not been implemented. One temporary workaround is to turn your categorical into numerical (not a good idea though). Will update...

this is a valid point. I fixed the problem in combo https://github.com/yzhao062/combo/commit/5d3066ebf94be9012bc55f7e5e48fa3338bf2e60 will update back when 0.1.3. is published. or you could override the combo/models/score_comb.py line 71 as check_parameter(n_buckets, 2,...

Sorry for being late on this. I recall some people mentioned that pickle may work. Haven't investigate. Should possibly do some experiment

is there a minimal example to show this? like our knn_exmaple using synthetic datasets?

   This is interesting. Looks like it is fine on my end with https://github.com/yzhao062/pyod/blob/master/examples/knn_example.py The code is  Is that possible something wrong with the sklearn version?