feat: apply ruff and fix main problems

Signed-off-by: Luca Foppiano <luca@foppiano.org>

Author: lfoppiano

delft/applications/citationClassifier.py

@@ -1,10 +1,11 @@
-import json
-from delft.utilities.Utilities import split_data_and_labels
-from delft.textClassification.reader import load_citation_sentiment_corpus
-from delft.textClassification import Classifier
 import argparse
+import json
 import time
+
+from delft.textClassification import Classifier
 from delft.textClassification.models import architectures
+from delft.textClassification.reader import load_citation_sentiment_corpus
+from delft.utilities.Utilities import split_data_and_labels
 
 list_classes = ["negative", "neutral", "positive"]
 
@@ -194,7 +195,7 @@ def classify(texts, output_format, architecture="gru", embeddings_name=None, tra
     if architecture not in architectures:
         print("unknown model architecture, must be one of " + str(architectures))
 
-    if transformer == None and embeddings_name == None:
+    if transformer is None and embeddings_name is None:
         # default word embeddings
         embeddings_name = "glove-840B"
 

delft/applications/dataseerClassifier.py

@@ -1,16 +1,18 @@
-import json
-from delft.utilities.Utilities import split_data_and_labels, t_or_f
-from delft.textClassification.reader import load_dataseer_corpus_csv
-from delft.textClassification.reader import vectorize as vectorizer
-from delft.textClassification import Classifier
 import argparse
+import json
 import time
-from delft.textClassification.models import architectures
+
 import numpy as np
 
+from delft.textClassification import Classifier
+from delft.textClassification.models import architectures
+from delft.textClassification.reader import load_dataseer_corpus_csv
+from delft.textClassification.reader import vectorize as vectorizer
+from delft.utilities.Utilities import split_data_and_labels, t_or_f
+
 """
-    Classifier for deciding if a sentence introduce a dataset or not, and prediction of the 
-    dataset type. 
+    Classifier for deciding if a sentence introduce a dataset or not, and prediction of the
+    dataset type.
 """
 
 
@@ -198,8 +200,8 @@ def train(
 
         model_name = 'dataseer-' + the_class + "_" + architecture
 
-        model = Classifier(model_name, architecture=architecture, list_classes=datatypes_list_subclasses[the_class], max_epoch=max_epoch, 
-            fold_number=fold_count, patience=patience, use_roc_auc=True, embeddings_name=embeddings_name, 
+        model = Classifier(model_name, architecture=architecture, list_classes=datatypes_list_subclasses[the_class], max_epoch=max_epoch,
+            fold_number=fold_count, patience=patience, use_roc_auc=True, embeddings_name=embeddings_name,
             batch_size=batch_size, class_weights=class_weights, early_stop=early_stop, transformer_name=transformer)
 
         if fold_count == 1:
@@ -560,7 +562,6 @@ def train_eval_cascaded(embeddings_name, fold_count, architecture="gru", transfo
     model_binary.eval(x_test, y_test)
 
     x_test_binary = x_test
-    y_test_binary = y_test
 
     # second, the first level datatype taxonomy for sentences classified as dataset
     (
@@ -610,7 +611,6 @@ def train_eval_cascaded(embeddings_name, fold_count, architecture="gru", transfo
 
     # eval by cascading
     result_binary = model_binary.predict(x_test_binary, output_format="default")
-    result_first = model_first.predict(x_test, output_format="default")
 
     # select sequences classified as dataset
     result_intermediate = np.asarray([np.argmax(line) for line in result_binary])
@@ -810,7 +810,7 @@ def build_prior_class_distribution():
     if architecture not in architectures:
         print("unknown model architecture, must be one of " + str(architectures))
 
-    if transformer == None and embeddings_name == None:
+    if transformer is None and embeddings_name is None:
         # default word embeddings
         embeddings_name = "glove-840B"
 

delft/applications/datasetTagger.py

@@ -1,9 +1,9 @@
-import os
 import argparse
 import json
+import os
 import time
-import numpy as np
 
+import numpy as np
 from sklearn.model_selection import train_test_split
 
 from delft.sequenceLabelling import Sequence

delft/applications/grobidTagger.py

@@ -308,9 +308,9 @@ def train(
     )
 
     if incremental:
-        if input_model_path != None:
+        if input_model_path is not None:
             model.load(input_model_path)
-        elif output_path != None:
+        elif output_path is not None:
             model.load(output_path)
         else:
             model.load()
@@ -423,9 +423,9 @@ def train_eval(
     )
 
     if incremental:
-        if input_model_path != None:
+        if input_model_path is not None:
             model.load(input_model_path)
-        elif output_path != None:
+        elif output_path is not None:
             model.load(output_path)
         else:
             model.load()
@@ -798,7 +798,7 @@ class Tasks:
             someTexts.append("2023 July the 22nd")
         elif model == "citation":
             someTexts.append(
-                "N. Al-Dhahir and J. Cioffi, \“On the uniform ADC bit precision and clip level computation for a Gaussian signal,\” IEEE Trans. Signal Processing, pp. 434–438, Feb. 1996."
+                r"N. Al-Dhahir and J. Cioffi, \“On the uniform ADC bit precision and clip level computation for a Gaussian signal,\” IEEE Trans. Signal Processing, pp. 434–438, Feb. 1996."
             )
             someTexts.append(
                 "T. Steinherz, E. Rivlin, N. Intrator, Off-line cursive script word recognition—a survey, Int. J. Doc. Anal. Recognition 2(3) (1999) 1–33."

delft/applications/insultTagger.py

@@ -1,10 +1,10 @@
-import os
-import json
-from delft.sequenceLabelling import Sequence
-from delft.sequenceLabelling.reader import load_data_and_labels_xml_file
 import argparse
+import json
+import os
 import time
 
+from delft.sequenceLabelling import Sequence
+from delft.sequenceLabelling.reader import load_data_and_labels_xml_file
 from delft.utilities.Utilities import t_or_f
 
 
@@ -224,7 +224,7 @@ def annotate(
     wandb = args.wandb
     num_workers = args.num_workers
 
-    if transformer == None and embeddings_name == None:
+    if transformer is None and embeddings_name is None:
         # default word embeddings
         embeddings_name = "glove-840B"
 

delft/applications/licenseClassifier.py

@@ -1,23 +1,25 @@
+import argparse
 import json
-from delft.utilities.Utilities import split_data_and_labels, t_or_f
-from delft.utilities.numpy import shuffle_triple_with_view
-from delft.textClassification.reader import vectorize as vectorizer
+import time
+
+import numpy as np
+
+from delft.textClassification import Classifier
+from delft.textClassification.models import architectures
 from delft.textClassification.reader import (
     load_texts_and_classes_pandas_no_id,
 )
-from delft.textClassification import Classifier
-import argparse
-import time
-from delft.textClassification.models import architectures
-import numpy as np
+from delft.textClassification.reader import vectorize as vectorizer
+from delft.utilities.numpy import shuffle_triple_with_view
+from delft.utilities.Utilities import split_data_and_labels, t_or_f
 
 """
     Two multiclass classifiers to be used in combination with Grobid to classify a license/copyrights section
-    extracted from a scientific article into two dimensions:  
+    extracted from a scientific article into two dimensions:
     - copyright owner: publisher, authors or undecidable (changed from NA to avoid issues with pandas)
-    - license associated to the article file: explicit copyrights (no restriction), creative commons licenses 
+    - license associated to the article file: explicit copyrights (no restriction), creative commons licenses
       (CC-0, CC-BY, CC-BY-NC, etc.), other, or undecidable (changed from NA to avoid issues with pandas)
-    
+
     Note: when the license is undecidable, this means normal copyrights when a copyright owner exists.
 """
 
@@ -690,7 +692,7 @@ def report_training_copyrights(y):
     if architecture not in architectures:
         print("unknown model architecture, must be one of " + str(architectures))
 
-    if transformer == None and embeddings_name == None:
+    if transformer is None and embeddings_name is None:
         # default word embeddings
         embeddings_name = "glove-840B"
 

delft/applications/nerTagger.py

@@ -1,16 +1,18 @@
+import argparse
 import os
+import time
+
 import numpy as np
+from sklearn.model_selection import train_test_split
+
 from delft.sequenceLabelling import Sequence
-from delft.utilities.Utilities import stats, t_or_f
-from delft.utilities.numpy import shuffle_arrays
 from delft.sequenceLabelling.reader import (
     load_data_and_labels_conll,
     load_data_and_labels_lemonde,
     load_data_and_labels_ontonotes,
 )
-from sklearn.model_selection import train_test_split
-import argparse
-import time
+from delft.utilities.numpy import shuffle_arrays
+from delft.utilities.Utilities import stats, t_or_f
 
 
 def configure(
@@ -642,8 +644,6 @@ def annotate(
     if not os.path.isfile(file_in):
         raise ValueError("the provided input file is not valid")
 
-    annotations = []
-
     if (dataset_type == "conll2003") and (lang == "en"):
         # load model
         model_name = "ner-en-conll2003-" + architecture

delft/applications/onnx_export.py

@@ -13,12 +13,12 @@
     python -m delft.applications.onnx_export --model MODEL_NAME --output OUTPUT_DIR
 """
 
-import os
-import json
 import argparse
+import json
+import os
 
-import torch
 import numpy as np
+import torch
 
 from delft.sequenceLabelling.wrapper import Sequence
 

delft/applications/softwareClassifier.py

@@ -1,21 +1,22 @@
-import json
-from delft.utilities.Utilities import split_data_and_labels
-from delft.textClassification.reader import load_software_use_corpus_json
-from delft.textClassification import Classifier
 import argparse
+import json
 import time
+
+from delft.textClassification import Classifier
 from delft.textClassification.models import architectures
+from delft.textClassification.reader import load_software_use_corpus_json
+from delft.utilities.Utilities import split_data_and_labels
 
 """
     This binary classifier is used in combination with a software mention recognition model, for characterizing
-    the nature of the citation of software in scientific and technical literature. 
+    the nature of the citation of software in scientific and technical literature.
     This classifier predicts if the software introduced by a software mention in a sentence is used
-    or not by the described work. 
+    or not by the described work.
 
     For the software mention recognizer, see https://github.com/ourresearch/software-mentions
     and grobidTagger.py in the present project DeLFT.
 
-    Best architecture/model is fine-tuned SciBERT. 
+    Best architecture/model is fine-tuned SciBERT.
 """
 
 list_classes = ["not_used", "used"]
@@ -225,7 +226,7 @@ def classify(texts, output_format, embeddings_name=None, architecture="gru", tra
     if architecture not in architectures:
         print("unknown model architecture, must be one of " + str(architectures))
 
-    if transformer == None and embeddings_name == None:
+    if transformer is None and embeddings_name is None:
         # default word embeddings
         embeddings_name = "glove-840B"
 

delft/applications/softwareContextClassifier.py

@@ -1,19 +1,21 @@
+import argparse
 import json
-from delft.utilities.Utilities import split_data_and_labels
-from delft.utilities.numpy import concatenate_or_none, shuffle_triple_with_view
+import time
+
+import numpy as np
+
+from delft.textClassification import Classifier
+from delft.textClassification.models import architectures
 from delft.textClassification.reader import (
     load_software_context_corpus_json,
     load_software_dataset_context_corpus_json,
 )
-from delft.textClassification import Classifier
-import argparse
-import time
-from delft.textClassification.models import architectures
-import numpy as np
+from delft.utilities.numpy import concatenate_or_none, shuffle_triple_with_view
+from delft.utilities.Utilities import split_data_and_labels
 
 """
     A multiclass classifier to be used in combination with a software mention recognition model, for characterizing
-    the nature of the mention of software in scientific and technical literature. 
+    the nature of the mention of software in scientific and technical literature.
     This classifier predicts if the software introduced by a software mention in a sentence is likely:
     - used or not by the described work (class used)
     - a creation of the described work (class creation)
@@ -22,7 +24,7 @@
     For the software mention recognizer, see https://github.com/ourresearch/software-mentions
     and grobidTagger.py in the present project DeLFT.
 
-    Best architecture/model is fine-tuned SciBERT. 
+    Best architecture/model is fine-tuned SciBERT.
 """
 
 list_classes = ["used", "creation", "shared"]
@@ -375,7 +377,7 @@ def report_training_contexts(y):
     if architecture not in architectures:
         print("unknown model architecture, must be one of " + str(architectures))
 
-    if transformer == None and embeddings_name == None:
+    if transformer is None and embeddings_name is None:
         # default word embeddings
         embeddings_name = "glove-840B"