from src.preprocessing.transformator import get_df
from src.evaluation.compare import compare_classifiers
from src.classifier.sklearn import pipelines
from src.data_retrieval.helpers import database
import pandas as pd
def warn(*args, **kwargs):
pass
warnings.warn = warn
db = database.MongoDB()
df = get_df(list(db.get_articles()))
# models
feature_sets = [
'bg_bert', 'bg_xlm', 'bg_styl', 'bg_lsa', 'en_use', 'en_nela', 'en_bert',
'en_elmo'
]
features = [
('top_1', ['bg_lsa_title', 'bg_lsa_text']),
('top_2', ['bg_lsa_title', 'bg_lsa_text', 'en_elmo_title',
'en_elmo_text']),
('top_3', [
'bg_lsa_title', 'bg_lsa_text', 'en_elmo_title', 'en_elmo_text',
'en_use_title', 'en_use_text'
]),
('top_4', [
import numpy as np
import pandas as pd
from src.classifier.sklearn import pipelines
from src.evaluation.compare import compare_classifiers
from src.preprocessing.transformator import get_df
from sklearn.model_selection import cross_val_predict, GridSearchCV
from sklearn.linear_model import LogisticRegression
#db = database.MongoDB()
articles = list(db.get_articles())
df = get_df(articles)
clf = LogisticRegression()
feature_sets = [
#'bg_bert',
'bg_xlm',
# 'bg_styl',
# 'bg_lsa',
# 'en_use',
# 'en_nela',
# 'en_bert',
# 'en_elmo'
]
all_feats = []
for feature_set in feature_sets:
all_feats.append(feature_set + '_title')