def _load_from_original_file(self):
# Load data from original
zipFile_path = self.DATASET_SPLIT_ROOT_FOLDER + self.DATASET_SUBFOLDER
try:
dataFile = zipfile.ZipFile(zipFile_path + "ml-10m.zip")
except (FileNotFoundError, zipfile.BadZipFile):
self._print("Unable to find data zip file. Downloading...")
download_from_URL(self.DATASET_URL, zipFile_path, "ml-10m.zip")
dataFile = zipfile.ZipFile(zipFile_path + "ml-10m.zip")
ICM_genre_path = dataFile.extract("ml-10M100K/movies.dat",
path=zipFile_path + "decompressed/")
ICM_tags_path = dataFile.extract("ml-10M100K/tags.dat",
path=zipFile_path + "decompressed/")
URM_path = dataFile.extract("ml-10M100K/ratings.dat",
path=zipFile_path + "decompressed/")
self._print("Loading Item Features Genres")
ICM_genres_dataframe, ICM_years_dataframe = _loadICM_genres_years(
ICM_genre_path, header=None, separator='::', genresSeparator="|")
self._print("Loading Item Features Tags")
ICM_tags_dataframe = _loadICM_tags(ICM_tags_path,
header=None,
separator='::')
ICM_all_dataframe = pd.concat(
[ICM_genres_dataframe, ICM_tags_dataframe])
self._print("Loading Interactions")
URM_all_dataframe, URM_timestamp_dataframe = _loadURM(URM_path,
header=None,
separator='::')
dataset_manager = DatasetMapperManager()
dataset_manager.add_URM(URM_all_dataframe, "URM_all")
dataset_manager.add_URM(URM_timestamp_dataframe, "URM_timestamp")
dataset_manager.add_ICM(ICM_genres_dataframe, "ICM_genres")
dataset_manager.add_ICM(ICM_years_dataframe, "ICM_year")
dataset_manager.add_ICM(ICM_tags_dataframe, "ICM_tags")
dataset_manager.add_ICM(ICM_all_dataframe, "ICM_all")
loaded_dataset = dataset_manager.generate_Dataset(
dataset_name=self._get_dataset_name(),
is_implicit=self.IS_IMPLICIT)
self._print("Cleaning Temporary Files")
shutil.rmtree(zipFile_path + "decompressed", ignore_errors=True)
self._print("Loading Complete")
return loaded_dataset
def _load_from_original_file(self):
# Load data from original
zipFile_path = self.DATASET_SPLIT_ROOT_FOLDER + self.DATASET_SUBFOLDER
try:
dataFile = zipfile.ZipFile(zipFile_path + "ml-1m.zip")
except (FileNotFoundError, zipfile.BadZipFile):
self._print("Unable to find data zip file. Downloading...")
download_from_URL(self.DATASET_URL, zipFile_path, "ml-1m.zip")
dataFile = zipfile.ZipFile(zipFile_path + "ml-1m.zip")
ICM_genre_path = dataFile.extract("ml-1m/movies.dat", path=zipFile_path + "decompressed/")
UCM_path = dataFile.extract("ml-1m/users.dat", path=zipFile_path + "decompressed/")
URM_path = dataFile.extract("ml-1m/ratings.dat", path=zipFile_path + "decompressed/")
self._print("Loading Interactions")
URM_all_dataframe, URM_timestamp_dataframe = _loadURM(URM_path, header=None, separator='::')
self._print("Loading Item Features genres")
ICM_genres_dataframe = _loadICM_genres(ICM_genre_path, header=None, separator='::', genresSeparator="|")
self._print("Loading User Features")
UCM_dataframe = pd.read_csv(filepath_or_buffer=UCM_path, sep="::", header=None, dtype={0:str, 1:str, 2:str, 3:str, 4:str})
UCM_dataframe.columns = ["UserID", "gender", "age_group", "occupation", "zip_code"]
# For each user a list of features
UCM_list = [[feature_name + "_" + str(UCM_dataframe[feature_name][index]) for feature_name in ["gender", "age_group", "occupation", "zip_code"]] for index in range(len(UCM_dataframe))]
UCM_dataframe = pd.DataFrame(UCM_list, index=UCM_dataframe["UserID"]).stack()
UCM_dataframe = UCM_dataframe.reset_index()[[0, 'UserID']]
UCM_dataframe.columns = ['FeatureID', 'UserID']
UCM_dataframe["Data"] = 1
dataset_manager = DatasetMapperManager()
dataset_manager.add_URM(URM_all_dataframe, "URM_all")
dataset_manager.add_URM(URM_timestamp_dataframe, "URM_timestamp")
dataset_manager.add_ICM(ICM_genres_dataframe, "ICM_genres")
dataset_manager.add_UCM(UCM_dataframe, "UCM_all")
loaded_dataset = dataset_manager.generate_Dataset(dataset_name=self._get_dataset_name(),
is_implicit=self.IS_IMPLICIT)
self._print("cleaning temporary files")
shutil.rmtree(zipFile_path + "decompressed", ignore_errors=True)
self._print("Loading Complete")
return loaded_dataset
def _load_from_original_file(self):
# Load data from original
URM_path = '../../data/data_train.csv'
ICM_path = '../../data/data_ICM_title_abstract.csv'
self._print("Loading Interactions")
URM_all_dataframe = _loadURM(URM_path, header=None, separator=',')
ICM_all_dataframe = _loadICM(ICM_path, header=None, separator=',')
# ICM_all_dataframe = pd.concat([ICM_genres_dataframe, ICM_tags_dataframe])
dataset_manager = DatasetMapperManager()
dataset_manager.add_URM(URM_all_dataframe, "URM_all")
dataset_manager.add_ICM(ICM_all_dataframe, "ICM_all")
loaded_dataset = dataset_manager.generate_Dataset(
dataset_name=self._get_dataset_name(),
is_implicit=self.IS_IMPLICIT)
return loaded_dataset
def _load_from_original_file(self):
# Load data from original
zipFile_path = self.DATASET_SPLIT_ROOT_FOLDER + self.DATASET_SUBFOLDER
try:
dataFile = zipfile.ZipFile(zipFile_path + "ml-100k.zip")
except (FileNotFoundError, zipfile.BadZipFile):
self._print("Unable to find data zip file. Downloading...")
download_from_URL(self.DATASET_URL, zipFile_path, "ml-100k.zip")
dataFile = zipfile.ZipFile(zipFile_path + "ml-100k.zip")
URM_path = dataFile.extract("ml-100k/u.data",
path=zipFile_path + "decompressed/")
self._print("Loading Interactions")
URM_all_dataframe, URM_timestamp_dataframe = _loadURM(URM_path,
header=None,
separator='\t')
dataset_manager = DatasetMapperManager()
dataset_manager.add_URM(URM_all_dataframe, "URM_all")
dataset_manager.add_URM(URM_timestamp_dataframe, "URM_timestamp")
loaded_dataset = dataset_manager.generate_Dataset(
dataset_name=self._get_dataset_name(),
is_implicit=self.IS_IMPLICIT)
self._print("Cleaning Temporary Files")
shutil.rmtree(zipFile_path + "decompressed", ignore_errors=True)
self._print("Loading Complete")
return loaded_dataset
