added target_count_numbers.py

scripts/ml/target_count_numbers.py

@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Created on Sun Jul 10 11:19:27 2022
+
+@author: tanu
+"""
+#========
+# pnca
+#========
+n_pnca_complete =  len(pnca_df)
+n_pnca_na       = pnca_df['dst'].isna().sum()
+n_pnca_actual   = len(pnca_df) - pnca_na
+
+#complete: 424
+#NA: 355
+#actual: 69
+
+#========
+# embb
+#========
+n_embb_complete = len(embb_df)
+n_embb_na       = embb_df['dst'].isna().sum()
+n_embb_actual   = len(embb_df) - a2
+#complete: 858
+#NA: 573
+#actual: 315
+    
+#========
+# katg 
+#========
+n_katg_complete = len(katg_df)
+n_katg_na       = katg_df['dst'].isna().sum()
+n_katg_actual   = len(katg_df) - katg_na
+#complete: 817
+#NA: 457
+#actual: 360
+    
+#========
+# rpob
+#========
+n_rpob_complete = len(rpob_df)
+n_rpob_na       = rpob_df['dst'].isna().sum()
+n_rpob_actual   = len(rpob_df) - rpob_na
+#complete: 1132
+#NA: 800
+#actual: 332
+    
+#========    
+# gid
+#========
+n_gid_complete  = len(gid_df)
+n_gid_na        = gid_df['dst'].isna().sum()
+n_gid_actual    = len(gid_df) -gid_na
+#complete: 531
+#NA: 473
+#actual: 58
+    
+#========        
+# alr
+#========
+n_alr_complete = len(alr_df)
+n_alr_na       = alr_df['dst'].isna().sum()
+n_alr_actual   = len(alr_df) - alr_na
+#complete: 271
+#NA: 268
+#actual: 3