Question about scaling and feature selection

[jessicaliu09]

Hi,

Because our data is zero-inflated, we want to use MaxAbsScaler. So when we turn off the scaling, it has error during the feature selection step. See errors below.

So we want to turn off scaling and keep feature selection, how should we set up the parameters? Thanks.

INFO: Applying collective feature selection...
Traceback (most recent call last):
File "/N/slate/shiwliu/STREAMLINE/0.1.STREAMLINE-Notebook.py", line 492, in
f_sel.run(run_parallel=False)
File "/N/slate/shiwliu/STREAMLINE/streamline/runners/feature_runner.py", line 324, in run
job_obj.run()
File "/N/slate/shiwliu/STREAMLINE/streamline/featurefns/selection.py", line 79, in run
self.select_features(selected_feature_lists,
File "/N/slate/shiwliu/STREAMLINE/streamline/featurefns/selection.py", line 254, in select_features
if len(union_list) > max_features_to_keep: # Apply further filtering if more than max features remains
TypeError: '>' not supported between instances of 'int' and 'NoneType'