Error running `Bootstrap n-grams` about pkdocclassifier

[fgh95]

Hi, @quang-nnguyen thank you very much for trying to reproduce our results. Your comments are very useful!

I looked into the integer type issue and added a conversion into int64 although this should not be the cause of the error you saw (pandas takes care of these different types when doing the comparison). It seems that it was trying to do bootstrap with the test features which is incorrect, so I filtered any input files with the 'test' string on them. Finally, I added an overwrite flag that should allow you to decided whether you want to overwrite result files if present in the output directory. This should allow you to reproduce our results but skip sections that are computationally intensive (if needed). Here is the commit modifying each of these things: b2829cd

Let me know if things are working now