Merge pull request #3 from StatguyUser/0.0.4

Optimized code for faster processing

Author: StatguyUser

build/lib/TextFeatureSelection.py

@@ -164,11 +164,13 @@ def getvalues_multiclass(self,unique_words,calc_df):
  label=[]
  word_presence=[]
 
+ #get binary pandas series for word if it is present row-wise or not
+ word_presence=calc_df['input_doc_list'].str.contains('\\b'+word+'\\b')
+
  for calc_base_label in set(self.target):
  ##get binary pandas series for label if it is present row-wise or not 
  label=calc_df['target']==calc_base_label
- #get binary pandas series for word if it is present row-wise or not
- word_presence=calc_df['input_doc_list'].str.contains('\\b'+word+'\\b')
+
  ##check if word count is existing and labels have value, to be sure if any regex error for word.
  if sum(word_presence) and sum(label):
  A,B,C,D,N=self.custom_cross_tab(label,word_presence)

dist/TextFeatureSelection-0.0.3-py3-none-any.whl

@@ -12,7 +12,7 @@
 
 setup(
  name='TextFeatureSelection',
- version='0.0.3',
+ version='0.0.4',
  description='Implementation of various algorithms for feature selection for text features, based on filter method',
  long_description=long_description,
  long_description_content_type='text/markdown', # This is important!

dist/TextFeatureSelection-0.0.3.tar.gz

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: TextFeatureSelection
-Version: 0.0.3
+Version: 0.0.4
 Summary: Implementation of various algorithms for feature selection for text features, based on filter method
 Home-page: https://github.com/StatguyUser/TextFeatureSelection
 Author: StatguyUser

dist/TextFeatureSelection-0.0.4-py3-none-any.whl

@@ -164,11 +164,13 @@ def getvalues_multiclass(self,unique_words,calc_df):
  label=[]
  word_presence=[]
 
+ #get binary pandas series for word if it is present row-wise or not
+ word_presence=calc_df['input_doc_list'].str.contains('\\b'+word+'\\b')
+
  for calc_base_label in set(self.target):
  ##get binary pandas series for label if it is present row-wise or not 
  label=calc_df['target']==calc_base_label
- #get binary pandas series for word if it is present row-wise or not
- word_presence=calc_df['input_doc_list'].str.contains('\\b'+word+'\\b')
+
  ##check if word count is existing and labels have value, to be sure if any regex error for word.
  if sum(word_presence) and sum(label):
  A,B,C,D,N=self.custom_cross_tab(label,word_presence)