finalised categorical and lineage col classifications

pnca_config.py

@@ -5,29 +5,22 @@ Created on Sat May 28 05:25:30 2022
 
 @author: tanu
 """
-import os, sys
 
-def MyGlobalVars():
-    global gene
-    global drug 
-    global homedir
-    gene  = 'pncA'
-    drug  = 'pyrazinamide'
-    homedir = os.path.expanduser("~")
-    
-MyGlobalVars()
+import os
 
-os.chdir(homedir + "/git/ML_AI_training/")
+gene  = 'pncA'
+drug  = 'pyrazinamide'
+total_mtblineage_u = 8
 
-# my function
+
+homedir = os.path.expanduser("~")
+os.chdir( homedir + '/git/ML_AI_training/')
+
+from UQ_ML_data import *
+setvars(gene,drug)
+from UQ_ML_data import *
+
+# from YC run_all_ML: run locally
 from UQ_MultModelsCl import MultModelsCl
-from UQ_pnca_ML.py import *
-# from YC run_all_ML
-
-# YC_resD2 = run_all_ML(input_pd=X, target_label=y, blind_test_input_df=X_bts, blind_test_target=y_bts, preprocess = True, var_type = 'mixed')
-
-# CVResultsDF = YC_resD2['CrossValResultsDF']
-# CVResultsDF.sort_values(by=['matthew'], ascending=False, inplace=True)
-# BTSResultsDF = YC_resD2['BlindTestResultsDF']
-# BTSResultsDF.sort_values(by=['matthew'], ascending=False, inplace=True)
 
+print('TESTING cmd:', Counter(y))