Add dynamic_ncols=True to all tqdm calls to make things cleaner.

avocado/augment.py

@@ -449,7 +449,8 @@ def augment_dataset(self, augment_name, dataset, num_augments,
         """
         augmented_objects = []
 
-        for reference_object in tqdm(dataset.objects, desc='Object'):
+        for reference_object in tqdm(dataset.objects, desc='Object',
+                                     dynamic_ncols=True):
             if include_reference:
                 augmented_objects.append(reference_object)
 

avocado/classifier.py

@@ -283,7 +283,8 @@ def predict(self, dataset):
 
         predictions = 0
 
-        for classifier in tqdm(self.classifiers, desc='Classifier'):
+        for classifier in tqdm(self.classifiers, desc='Classifier',
+                               dynamic_ncols=True):
             fold_scores = classifier.predict_proba(
                 features, raw_score=True,
                 num_iteration=classifier.best_iteration_

avocado/dataset.py

@@ -407,7 +407,7 @@ def extract_raw_features(self, featurizer):
         """
         list_raw_features = []
         object_ids = []
-        for obj in tqdm(self.objects, desc='Object'):
+        for obj in tqdm(self.objects, desc='Object', dynamic_ncols=True):
             obj_features = featurizer.extract_raw_features(obj)
             list_raw_features.append(obj_features.values())
             object_ids.append(obj.metadata['object_id'])

scripts/avocado_augment

@@ -79,7 +79,8 @@ if __name__ == "__main__":
     else:
         # Process all chunks
         print("Processing the dataset in %d chunks..." % args.num_chunks)
-        for chunk in tqdm(range(args.num_chunks), desc='Chunk'):
+        for chunk in tqdm(range(args.num_chunks), desc='Chunk',
+                          dynamic_ncols=True):
             process_chunk(augmentor, chunk, args, verbose=False)
 
     print("Done!")

scripts/avocado_download_plasticc

@@ -134,7 +134,8 @@ def update_plasticc_observations(observations):
 def preprocess_observations(input_path, output_path, chunk_size=10**6):
     """Preprocess an observations table and write it out."""
     for chunk in tqdm(pd.read_csv(input_path, chunksize=chunk_size),
-                      desc="    %s" % os.path.basename(input_path)):
+                      desc="    %s" % os.path.basename(input_path),
+                      dynamic_ncols=True):
         chunk = update_plasticc_observations(chunk)
         avocado.utils.write_dataframe(output_path, chunk, 'observations',
                                       append=True, index_chunk_column=False)

scripts/avocado_featurize

@@ -69,7 +69,8 @@ if __name__ == "__main__":
     else:
         # Process all chunks
         print("Processing the dataset in %d chunks..." % args.num_chunks)
-        for chunk in tqdm(range(args.num_chunks), desc='Chunk'):
+        for chunk in tqdm(range(args.num_chunks), desc='Chunk',
+                          dynamic_ncols=True):
             process_chunk(featurizer, chunk, args, verbose=False)
 
     print("Done!")

scripts/avocado_predict

@@ -63,7 +63,8 @@ if __name__ == "__main__":
     else:
         # Process all chunks
         print("Processing the dataset in %d chunks..." % args.num_chunks)
-        for chunk in tqdm(range(args.num_chunks), desc='Chunk'):
+        for chunk in tqdm(range(args.num_chunks), desc='Chunk',
+                          dynamic_ncols=True):
             process_chunk(classifier, chunk, args, verbose=False)
 
     print("Done!")