CU-8699049kf: Fix a few v2 vs v2 access issues

Author: mart-r

medcat/2_train_model/1_unsupervised_training/unsupervised_medcattraining.py

@@ -28,13 +28,14 @@
 df = cs.DataFrame(index=cogstack_indices, columns=text_columns)  # type: ignore
 
 cat = CAT.load_model_pack(model_pack_path+model_pack_name)
-cat.cdb.print_stats()
-cat.train(data_iterator=df[text_columns].iterrows(),
-          nepochs=1,
-          fine_tune=True,
-          progress_print=10000,
-          is_resumed=False)
+print(cat.cdb.get_basic_info())
+cat.trainer.train_unsupervised(
+    data_iterator=df[text_columns].iterrows(),
+    nepochs=1,
+    fine_tune=True,
+    progress_print=10000,
+    is_resumed=False)
 
-cat.cdb.print_stats()
+print(cat.cdb.get_basic_info())
 
-cat.create_model_pack(save_dir_path=model_pack_path, model_pack_name=output_modelpack_name)
+cat.save_model_pack(save_dir_path=model_pack_path, model_pack_name=output_modelpack_name)

medcat/2_train_model/1_unsupervised_training/unsupervised_training.py

@@ -1,3 +1,4 @@
+from medcat.cat import logger as cat_logger
 from medcat.cat import CAT
 import pandas as pd
 import os
@@ -44,25 +45,25 @@
 # Load modelpack
 print('Loading modelpack')
 cat = CAT.load_model_pack(model_pack_path)
-cat.log.addHandler(logging.StreamHandler()) # add console output
+cat_logger.addHandler(logging.StreamHandler()) # add console output
 
 print('STATS:')
-cat.cdb.print_stats()
+print(cat.cdb.get_basic_info())
 
 # CHANGE AS NEEDED - if the number of spligt files is different
 all_data_files = [f'split_notes_5M_{nr}.csv' for nr in range(1, 20)]  # file containing training material.
 for i, data_file in enumerate(all_data_files):
     # Load training data
     print('Load data for', i, 'from', data_file)
     data = pd.read_csv(os.path.join(data_dir, data_file))
-    cat.train(data.text.values, progress_print=100)
+    cat.trainer.train_unsupervised(data.text.values, progress_print=100)
 
     print('Stats now, after', i)
-    cat.cdb.print_stats()
+    print(cat.cdb.get_basic_info())
 
     # save modelpack
-    cat.create_model_pack(save_dir_path=model_dir, model_pack_name=f"{output_modelpack}_{i}")
+    cat.save_model_pack(save_dir_path=model_dir, model_pack_name=f"{output_modelpack}_{i}")
 
 # save modelpack - ALL
-cat.create_model_pack(save_dir_path=model_dir, model_pack_name=output_modelpack)
+cat.save_model_pack(save_dir_path=model_dir, model_pack_name=output_modelpack)
 

medcat/3_run_model/run_model.py

@@ -59,7 +59,7 @@
 else:
     snomed_filter = set(cat.cdb.cui2info.keys())
 
-cat.config.linking.filters.cuis = snomed_filter
+cat.config.components.linking.filters.cuis = snomed_filter
 del snomed_filter
 
 # build query, change as appropriate

medcat/compare_models/compare.py

@@ -154,8 +154,8 @@ def get_diffs_for(model_pack_path_1: str,
             if show_progress:
                 print("After adding children from 2nd model have a total of",
                       len(cui_filter), "CUIs")
-        cat1.config.linking.filters.cuis = cui_filter
-        cat2.config.linking.filters.cuis = cui_filter
+        cat1.config.components.linking.filters.cuis = cui_filter
+        cat2.config.components.linking.filters.cuis = cui_filter
     ann_diffs = get_per_annotation_diffs(cat1, cat2, documents, keep_raw=keep_raw,
                                          doc_limit=doc_limit)
     if show_progress: