python - Match Exactly column name in a data frame -
i have following
for col in features: my_features = pd.merge(my_features,drug.ix[:,[col]], left_index = true,right_index=true)
which produces following
my_features.columns out[373]: index([u'**pcbd1_x**', u'**pcbd1_y**', u'klk8', u'tnfsf13 /// tnfsf12-', u'ripk5', u'svil'], dtype='object')
while should match features
features out[375]: ['pcbd1', 'klk8', 'tnfsf13 /// tnfsf12-', 'ripk5', 'svil']
how can change "col" match when creating data frame above .in other words avoid bold column matches. regular expression problem.
update
i trying extract columns listed in "features" "drug". before
for col in features: my_features = pd.merge(my_features,drug.ix[:,[col]], left_index = true,right_index=true)enter code here
i have following line create structure my_features
my_features = pd.dataframe( drug.ix[:,features[0]] , index = drug.index , columns=[features[0]])
so whole code like
my_features = pd.dataframe( drug.ix[:,features[0]] , index = drug.index , columns=[features[0]]) iterfeatures = iter(features) next(iterfeatures) col in features: my_features = pd.merge(my_features,drug.ix[:,[col]], left_index = true,right_index=true)
you can use pandas.dataframe.filter(). has regex argument [as "like" argument]. http://pandas.pydata.org/pandas-docs/stable/generated/pandas.dataframe.filter.html
Comments
Post a Comment