Refactoring: drop pandas usage.

bpsad.py

@@ -12,224 +12,195 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-"""Brazilian Portuguese Sentiment Analysis Datasets (BPSAD)"""
+"""BPSAD -- Brazilian Portuguese Sentiment Analysis Datasets"""
 
 import csv
-import re
-import pandas as pd
-import json
 import os
 import datasets
+import sys
 
-# Functions
-def get_text(text):
-    preproc_text = []
-    for sentence in text:
-        preproc_sentence = re.findall("'([^']*)'", sentence)
-        preproc_sentence = ' '.join(preproc_sentence)
-        preproc_text.append(preproc_sentence)
-    return preproc_text
-
-
-def get_kfold(text, label, kfold_ref, kfolds):
-    output_dictionary = {}
-    boolean_vec = [kfold_ref[i] 
-        in kfolds for i in range(len(kfold_ref))]
-    output_dictionary['text'] = [text[i] for i in range(len(text)) if boolean_vec[i]]
-    output_dictionary['label'] = [int(label[i]) for i in range(len(label)) if boolean_vec[i]]
-    output_dictionary['kfold'] = [kfold_ref[i] for i in range(len(text)) if boolean_vec[i]]
-    return output_dictionary 
-
-
-def load_bpsad_p(address):
-    table =  pd.read_csv(address, low_memory = False)
-    
-    # We'll get 'review_text_tokenized' and 'polarity'
-    text = table['review_text_tokenized'].to_list()
-    text = get_text(text)
-    label = table['polarity'].to_list()
-    # label = [int(i) for i in table['polarity'].to_list()]
-    kfold = table['kfold_polarity'].to_list()
-    # Removing nan instances from polarity
-    data_train = get_kfold(text, label, kfold, [1,2,3,4,5,6,7,8])
-    data_dev = get_kfold(text, label, kfold, [9])
-    data_test = get_kfold(text, label, kfold, [10])
-    return data_train, data_dev, data_test
-
-
-def load_bpsad_r(address):
-    table =  pd.read_csv(address, low_memory = False)
-    
-    # We'll get 'review_text_tokenized' and 'polarity'
-    text = table['review_text_tokenized'].to_list()
-    text = get_text(text)
-    label = table['rating'].to_list()
-    # label = [int(i) for i in table['rating'].to_list()]
-    kfold = table['kfold_polarity'].to_list()
-    # Removing nan instances from polarity
-    data_train = get_kfold(text, label, kfold, [1,2,3,4,5,6,7,8])
-    data_dev = get_kfold(text, label, kfold, [9])
-    data_test = get_kfold(text, label, kfold, [10])
-    return data_train, data_dev, data_test
-
-
-_HOMEPAGE = "https://www.kaggle.com/datasets/fredericods/ptbr-sentiment-analysis-datasets"
-
-_DESCRIPTION = """
-The Brazilian Portuguese Sentiment Analysis Dataset (BPSAD) is composed by the
-concatenation of 5 differents sources (Olist, B2W Digital, Buscapé, UTLC-Apps and
-UTLC-Movies), each one is composed by evaluation sentences classified according
-to the polarity (0: negative; 1: positive) and ratings (1, 2, 3, 4 and 5 stars).
-"""
+csv.field_size_limit(sys.maxsize)
 
-_CITATION = r"""
-@misc{corpusCarolinaV1.1,
-    title={
-        Brazilian Portuguese Sentiment Analysis Datasets},
-    author={
-        Dias, Frederico},
-    howpublished={
-        \url{https://www.kaggle.com/datasets/fredericods/ptbr-sentiment-analysis-datasets}},
-    year={
-        2021},
-    note={Version 1},
-}
-"""
 
-_LICENSE = """
-"""
+_HOMEPAGE = """\
+https://www.kaggle.com/datasets/fredericods/ptbr-sentiment-analysis-datasets"""
+
 
-_MANUAL_DOWNLOAD_INSTRUCTIONS = """
-data = datasets.load_dataset(
-    path = 'BPSAD.py',
-    name = 'Polarity'/'Rating',
-    data_dir = 'path to concatenated.csv')
+_DESCRIPTION = """\
+The Brazilian Portuguese Sentiment Analysis Dataset (BPSAD) is composed
+by the concatenation of 5 differents sources (Olist, B2W Digital, Buscapé,
+UTLC-Apps and UTLC-Movies), each one is composed by evaluation sentences
+classified according to the polarity (0: negative; 1: positive) and ratings
+(1, 2, 3, 4 and 5 stars)."""
+
+
+_CITATION = """\
+@inproceedings{souza2021sentiment,
+    author={
+        Souza, Frederico Dias and
+        Baptista de Oliveira e Souza Filho, João},
+    booktitle={
+        2021 IEEE Latin American Conference on
+        Computational Intelligence (LA-CCI)}, 
+    title={
+        Sentiment Analysis on Brazilian Portuguese User Reviews}, 
+    year={2021},
+    pages={1-6},
+    doi={10.1109/LA-CCI48322.2021.9769838}
+}
 """
 
-class BPSADPolarity(datasets.GeneratorBasedBuilder):
-    """BPSAD: Polarity classification task for BPSAD dataset."""
 
-    VERSION = datasets.Version("1.0.0")
+_VERSION = datasets.Version("1.0.0")
+_LICENSE = ""
 
-    # This is an example of a dataset with multiple configurations.
-    # If you don't want/need to define several sub-sets in your dataset,
-    # just remove the BUILDER_CONFIG_CLASS and the BUILDER_CONFIGS attributes.
+        
 
-    # If you need to make complex sub-parts in the datasets with configurable options
-    # You can create your own builder configuration class to store attribute, inheriting from datasets.BuilderConfig
-    # BUILDER_CONFIG_CLASS = MyBuilderConfig
+class BPSAD(datasets.GeneratorBasedBuilder):
+    """BPSAD dataset."""
 
-    # You will be able to load one or the other configurations in the following list with
-    # data = datasets.load_dataset('my_dataset', 'first_domain')
-    # data = datasets.load_dataset('my_dataset', 'second_domain')
     BUILDER_CONFIGS = [
-        datasets.BuilderConfig(name="Polarity", version=VERSION, description="Polarity classification of the Brazilian Portuguese Sentiment Analysis Datasets (BPSAD)"),
-        datasets.BuilderConfig(name="Rating", version=VERSION, description="Rating classification of the Brazilian Portuguese Sentiment Analysis Datasets (BPSAD)"),
+        datasets.BuilderConfig(
+            name="polarity",
+            description="Polarity classification dataset."
+        ),
+        datasets.BuilderConfig(
+            name="rating",
+            description="Rating classification dataset."
+        ),
     ]
 
-    DEFAULT_CONFIG_NAME = "Polarity"  # It's not mandatory to have a default configuration. Just use one if it make sense.
+    @property
+    def manual_download_instructions(self):
+        return (
+            "To use this dataset you have to download it manually:\n"
+            "   1. Download the `concatenated` file from `{_HOMEPAGE}`.\n"
+            "   2. Extract the file inside `[PATH_TO_FILE]`.\n"
+            "   3. Load the dataset using the command:\n"
+            "       datasets.load_dataset("
+            "\"lm4pt/bpsad\", name=..., data_dir=\"[PATH_TO_FILE]\")\n\n"
+            "Possible names are: `polarity` and `rating`."
+        )
 
     def _info(self):
-        features = datasets.Features(
-            {
-                "text": datasets.Value("string"),
-                "label": datasets.Value("int8"),
-                "kfold": datasets.Value("int8")
-                # These are the features of your dataset like images, labels ...
-            }
-        )
+        # Note:
+        # DEFAULT_CONFIG_NAME is not working and returns the value `default`.
+        # Also, it is better to set the config name explicitly.
+
+        if self.config.name not in ['polarity', 'rating']:
+            raise ValueError((
+                f"`{self.config.name}` is not a valid config name. Possible "
+                "values are `polarity` and `rating`. Make sure to pass via "
+                "`datasets.load_dataset('lm4pt/bpsad', name=...)`"
+            ))
+
+        if self.config.name == "polarity":
+            features = datasets.Features({
+                "review_text": datasets.Value("string"),
+                "polarity": datasets.Value("int8"),
+            })
+        else:
+            features = datasets.Features({
+                "review_text": datasets.Value("string"),
+                "rating": datasets.Value("int8"),
+            })
+
 
         return datasets.DatasetInfo(
-            # This is the description that will appear on the datasets page.
             description=_DESCRIPTION,
-            # This defines the different columns of the dataset and their types
-            features=features,  # Here we define them above because they are different between the two configurations
-            # If there's a common (input, target) tuple from the features, uncomment supervised_keys line below and
-            # specify them. They'll be used if as_supervised=True in builder.as_dataset.
-            # supervised_keys=("sentence", "label"),
-            # Homepage of the dataset for documentation
+            features=features,
             homepage=_HOMEPAGE,
-            # License for the dataset if available
-            license=_LICENSE,
-            # Citation for the dataset
             citation=_CITATION,
+            license=_LICENSE,
+            version=_VERSION,
         )
 
 
     def _split_generators(self, dl_manager):
         data_dir = os.path.abspath(os.path.expanduser(dl_manager.manual_dir))
-        # check if manual folder exists
+
+        # validates if dataset folder exists
         if not os.path.exists(data_dir):
-            raise FileNotFoundError(
-                f"{data_dir} does not exist. Make sure you insert a manual dir via `datasets.load_dataset('bpsad', data_dir=...)`. Manual download instructions: {_MANUAL_DOWNLOAD_INSTRUCTIONS})"
-            )
-        
+            raise FileNotFoundError((
+                data_dir + " does not exist. Make sure to pass the "
+                "parameter `data_dir` via `datasets.load_dataset`.\n"
+                "Manual download instructions:\n" +
+                self.manual_download_instructions
+            ))
+
         data_file = os.path.join(data_dir, "concatenated.csv")
+
         # check if dataset file exists
         if not os.path.exists(data_file):
-            raise FileNotFoundError(
-                f"{data_file} does not exist. Make sure you the downloaded data is inside the manual dir passed via `datasetts.load_dataset('bpsad', data_dir=...)`. Manual download instructions: {_MANUAL_DOWNLOAD_INSTRUCTIONS})"
-            )
-        if self.config.name == "Polarity":
-            data_train, data_dev, data_test = load_bpsad_p(data_file)
-        else:
-            data_train, data_dev, data_test = load_bpsad_r(data_file)
-
-        pd.DataFrame(data_train).to_csv(
-            os.path.join(data_dir, "train.csv"), index=False, header=False)
-        pd.DataFrame(data_dev).to_csv(
-            os.path.join(data_dir, "dev.csv"), index=False, header=False)
-        pd.DataFrame(data_test).to_csv(
-            os.path.join(data_dir, "test.csv"), index=False, header=False)
-
-        # with open(os.path.join(data_dir, "train.jsonl"),"w") as fname:
-        #     json.dump(data_train, fname)
-
-        # with open(os.path.join(data_dir, "dev.jsonl"), "w") as fname:
-        #     json.dump(data_dev, fname)
-
-        # with open(os.path.join(data_dir, "test.jsonl"), "w") as fname:
-        #     json.dump(data_test, fname)
+            raise FileNotFoundError((
+                data_file + " does not exist. " +
+                self.manual_download_instructions
+            ))
 
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
-                # These kwargs will be passed to _generate_examples
                 gen_kwargs={
-                    "filepath": os.path.join(data_dir, "train.csv"),
+                    "filepath": data_file,
                     "split": "train",
+                    'kfold_min': 1,
+                    'kfold_max': 8
                 },
             ),
             datasets.SplitGenerator(
                 name=datasets.Split.VALIDATION,
-                # These kwargs will be passed to _generate_examples
                 gen_kwargs={
-                    "filepath": os.path.join(data_dir, "dev.csv"),
+                    "filepath": data_file,
                     "split": "dev",
+                    'kfold_min': 9,
+                    'kfold_max': 9
                 },
             ),
             datasets.SplitGenerator(
                 name=datasets.Split.TEST,
-                # These kwargs will be passed to _generate_examples
                 gen_kwargs={
-                    "filepath": os.path.join(data_dir, "test.csv"),
-                    "split": "test"
+                    "filepath": data_file,
+                    "split": "test",
+                    'kfold_min': 10,
+                    'kfold_max': 10
                 },
             ),
         ]
 
 
-    # method parameters are unpacked from `gen_kwargs` as given in `_split_generators`
-    def _generate_examples(self, filepath, split):
-        with open(filepath, "r") as f:
-            reader = csv.reader(f)
-            # for key, row in enumerate(f):
-            for key, row in enumerate(reader):
-                # data = json.loads(row)
-                
-                # Yields examples as (key, example) tuples
-                yield key, {
-                    "text": row[0],
-                    "label": row[1],
-                    "kfold": row[2],
-                }
+    def _generate_examples(self, filepath, split, kfold_min, kfold_max):
+        # CSV columns
+        # 0 - original_index,
+        # 1 - review_text,
+        # 2 - review_text_processed,
+        # 3 - review_text_tokenized,
+        # 4 - polarity,
+        # 5 - rating,
+        # 6 - kfold_polarity,
+        # 7 - kfold_rating
+
+        with open(filepath) as csv_file:
+            csv_reader = csv.reader(csv_file, delimiter=',')
+            
+            # skip header
+            _ = next(csv_reader)
+
+            _id = 0
+            if self.config.name == 'polarity':
+                for row in csv_reader:
+                    kfold = int(row[7])
+                    if kfold_min <= kfold and kfold <= kfold_max:
+                        yield _id, {
+                            "review_text": row[2],
+                            "polarity": int(float(row[5])),
+                        }
+                        _id += 1
+            else:
+                for row in csv_reader:
+                    kfold = int(row[8])
+                    if kfold_min <= kfold and kfold <= kfold_max:
+                        yield _id, {
+                            "review_text": row[2],
+                            "rating": int(float(row[6])),
+                        }
+                        _id += 1