Daniel/fix code style and pylint exit code (#336)

* Added html output for pylint and flake8

* style fixes

* fixed pylint exit code; more style fixes

* disabled subrocess.run exception on child failure

* write style results to external storage dir

* create output dir if missing

* invoke ansi2html through python instead of shell

* added flake-html to reqs

* move ansi2hml.py to nlp_architect.utils

Author: danielkorat

.gitignore

@@ -7,6 +7,7 @@
 .venv
 .styleenv
 .coverage
+flake8.txt
 /build/
 generated
 /dist/
@@ -22,7 +23,4 @@ doc/build/**
 .idea/
 .nlp_architect_env/
 src/
-pylint.html
-pylint.txt
-flake8.txt
-tests/fixtures/data/chunker/(
+tests/fixtures/data/chunker/

doc/source/api.rst

@@ -45,7 +45,7 @@ to train the model weights, perform inference, and save/load the model.
     nlp_architect.models.chunker.SequenceChunker
     nlp_architect.models.intent_extraction.Seq2SeqIntentModel
     nlp_architect.models.intent_extraction.MultiTaskIntentModel
-    nlp_architect.models.matchlstm_ansptr.MatchLSTM_AnswerPointer
+    nlp_architect.models.matchlstm_ansptr.MatchLSTMAnswerPointer
     nlp_architect.models.memn2n_dialogue.MemN2N_Dialog
     nlp_architect.models.most_common_word_sense.MostCommonWordSense
     nlp_architect.models.ner_crf.NERCRF

doc/source/cross_doc_coref.rst

@@ -130,66 +130,4 @@ See code example below for running a full cross document coreference evaluation
 Code Example
 ============
 
-.. code:: python
-
-    # Configure which sieves you would like to run, the order, constrain and threshold,
-    event_config = EventConfig()
-
-    event_config.sieves_order = [
-        (SieveType.STRICT, RelationType.SAME_HEAD_LEMMA, 0.0),
-        (SieveType.VERY_RELAX, RelationType.WIKIPEDIA_DISAMBIGUATION, 0.1),
-        (SieveType.VERY_RELAX, RelationType.WORD_EMBEDDING_MATCH, 0.7),
-        (SieveType.RELAX, RelationType.SAME_HEAD_LEMMA_RELAX, 0.5),
-    ]
-
-    event_config.gold_mentions_file = '<Replace with your event mentions json file>'
-
-    entity_config = EntityConfig()
-
-    entity_config.sieves_order = [
-        (SieveType.STRICT, RelationType.SAME_HEAD_LEMMA, 0.0),
-        (SieveType.VERY_RELAX, RelationType.WIKIPEDIA_REDIRECT_LINK, 0.1),
-        (SieveType.VERY_RELAX, RelationType.WIKIPEDIA_DISAMBIGUATION, 0.1),
-        (SieveType.VERY_RELAX, RelationType.WORD_EMBEDDING_MATCH, 0.7),
-        (SieveType.VERY_RELAX, RelationType.REFERENT_DICT, 0.5)
-    ]
-
-    entity_config.gold_mentions_file = '<Replace with your entity mentions json file>'
-
-    # CDCResources hold default attribute values that might need to be change,
-    # (using the defaults values in this example), use to configure attributes
-    # such as resources files location, output directory, resources init methods and other.
-    # check in class and see if any attributes require change in your set-up
-    resource_location = CDCResources()
-
-    # create a new cross doc resources, with all needed semantic relation models
-    resources = CDCSettings(resource_location, event_config, entity_config)
-
-    # run event evaluation
-    event_clusters = None
-    if event_config.run_evaluation:
-        # entry point for the event evaluation process
-        event_clusters = run_event_coref(resources)
-
-    # run entity evaluation in the same way
-    entity_clusters = None
-    if entity_config.run_evaluation:
-        # entry point for the entity evaluation process
-        entity_clusters = run_entity_coref(resources)
-
-    print('-=Cross Document Coref Results=-')
-    print('-=Event Clusters Mentions=-')
-    for event_cluster in event_clusters.clusters_list:
-        print(event_cluster.coref_chain)
-        for event_mention in event_cluster.mentions:
-            print(event_mention.mention_id)
-            print(event_mention.tokens_str)
-
-    print('-=Entity Clusters Mentions=-')
-    for entity_cluster in entity_clusters.clusters_list:
-        print(entity_cluster.coref_chain)
-        for entity_mention in entity_cluster.mentions:
-            print(entity_mention.mention_id)
-            print(entity_mention.tokens_str)
-
-You can find the above example at this location: ``examples/cross_doc_coref/cross_doc_coref_sieves.py``
+You can find code example for running the system at: ``examples/cross_doc_coref/cross_doc_coref_sieves.py``

examples/memn2n_dialogue/interactive_utils.py

@@ -46,12 +46,12 @@ def interactive_loop(model, babi):
         line_in = text_input('>>> ').strip().lower()
         if not line_in:
             line_in = "<SILENCE>"
-        if line_in == 'exit' or line_in == 'quit':
+        if line_in in ('exit', 'quit'):
             break
         if line_in == 'help':
             print_help()
             continue
-        if line_in == 'restart' or line_in == 'clear':
+        if line_in in ('restart', 'clear'):
             context = []
             response = None
             time_feat = 1

examples/most_common_word_sense/inference.py

@@ -23,13 +23,12 @@
 from nltk.corpus import wordnet as wn
 from termcolor import colored
 
+from examples.most_common_word_sense.feature_extraction import extract_synset_data, \
+    extract_features_envelope
+from examples.most_common_word_sense.prepare_data import read_inference_input_examples_file
 from nlp_architect.models.most_common_word_sense import MostCommonWordSense
 from nlp_architect.utils.io import validate_existing_filepath, check_size
 
-from prepare_data import read_inference_input_examples_file
-from feature_extraction import extract_features_envelope
-from feature_extraction import extract_synset_data
-
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.DEBUG)
 

examples/most_common_word_sense/prepare_data.py

@@ -26,8 +26,8 @@
 import gensim
 import numpy as np
 from sklearn.model_selection import train_test_split
-from feature_extraction import extract_features_envelope
 
+from examples.most_common_word_sense.feature_extraction import extract_features_envelope
 from nlp_architect.utils.io import validate_existing_filepath, check_size, validate_parent_exists
 
 logger = logging.getLogger(__name__)

examples/np_semantic_segmentation/data.py

@@ -244,7 +244,7 @@ def load_data_from_file(self):
         # count num of feature vectors
         num_feats = len(reader_list)
         # is_y_labels is for inference - if the inference data is labeled y_labels are extracted
-        self.is_y_labels = True if (len(reader_list[0]) == self.feature_vec_dim + 1) else False
+        self.is_y_labels = len(reader_list[0]) == self.feature_vec_dim + 1
         X_feature_matrix = numpy.zeros((num_feats, self.feature_vec_dim))
         Y_labels_vec = []
         cntr = 0

examples/np_semantic_segmentation/inference.py

@@ -18,7 +18,7 @@
 import csv
 import os
 
-from data import NpSemanticSegData, absolute_path
+from .data import NpSemanticSegData, absolute_path
 
 from nlp_architect.models.np_semantic_segmentation import NpSemanticSegClassifier
 from nlp_architect.utils.io import validate_existing_filepath, validate_parent_exists

examples/np_semantic_segmentation/preprocess_tratz2011.py

@@ -21,7 +21,7 @@
 import csv
 import os
 
-from data import absolute_path
+from .data import absolute_path
 
 from nlp_architect.utils.io import validate_existing_directory
 

examples/np_semantic_segmentation/train.py

@@ -22,20 +22,20 @@
 from nlp_architect.utils.io import validate_existing_filepath, validate_parent_exists, validate
 
 
-def train_mlp_classifier(dataset, model_file_path, num_epochs, callback_args=None):
+def train_mlp_classifier(dataset, model_file_path, epochs, callback_args=None):
     """
     Train the np_semantic_segmentation mlp classifier
     Args:
         model_file_path (str): model path
-        num_epochs (int): number of epochs
+        epochs (int): number of epochs
         callback_args (dict): callback_arg
         dataset: NpSemanticSegData object containing the dataset
 
     Returns:
         print error_rate, test_accuracy_rate and precision_recall_rate evaluation from the model
 
     """
-    model = NpSemanticSegClassifier(num_epochs, callback_args)
+    model = NpSemanticSegClassifier(epochs, callback_args)
     input_dim = dataset.train_set_x.shape[1]
     model.build(input_dim)
     # run fit