return (self.__data__.groupby(['id', 'session']).category.first().isin(categories)). \
groupby('id').value_counts(normalize=normalize).rename(name)
if application:
applications = [application] if not isinstance(application, list) else application
return (self.__data__.groupby(['id', 'session']).application.first().isin(applications)). \
groupby('id').value_counts(normalize=normalize).rename(name)
if __name__ == "__main__":
###########
# EXAMPLE #
###########
hlp.hi()
hlp.set_param(log_level=3)
# Read sample data
data = hlp.add_dates(
pd.read_parquet(
path='../../data/glance/processed_appevents/0a0fe3ed-d788-4427-8820-8b7b696a6033_appevents.parquet'),
'appevents')
# Data path
data_path = '../../data/glance/appevents/0a0fe3ed-d788-4427-8820-8b7b696a6033_appevents.parquet'
# More sample data
data2 = pd.read_parquet(path='../../data/glance/appevents/0a9edba1-14e3-466a-8d0c-f8a8170cefc8_appevents.parquet')
data3 = pd.read_parquet(path='../../data/glance/appevents/0a48d1e8-ead2-404a-a5a2-6b05371200b1_appevents.parquet')
data4 = hlp.add_dates(pd.concat([data, data2, data3], sort=True), 'appevents')
import pandas as pd
from os.path import join
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import cross_validate
from sklearn.neural_network import MLPRegressor
from sklearn.ensemble import RandomForestRegressor, AdaBoostRegressor
from tqdm import tqdm
import matplotlib.pyplot as plt
import seaborn as sns
import xgboost as xg
if __name__ == '__main__':
# Set some parameters
hlp.hi('Implicit attitude & mobileDNA')
hlp.set_param(data_dir=join(hlp.DATA_DIR, 'implicit'), log_level=1)
# Get data
survey_data = pd.read_csv(join(hlp.DATA_DIR, 'data.csv'), sep='\t')
log_data = pd.read_csv(join(hlp.DATA_DIR, 'log_data.csv'),
sep=';').iloc[:, 1:]
# Build object
ae = Appevents(log_data,
add_categories=False,
add_date_annotation=False,
strip=True)
#ae = Appevents.from_pickle(path=join(hlp.DATA_DIR, 'implicit.ae'))
del log_data
import numpy as np
import mobiledna.core.help as hlp
import pandas as pd
from os.path import join
from mobiledna.core.appevents import Appevents
from mobiledna.core.notifications import Notifications
from tqdm import tqdm
import mobiledna.test.mdecline_features as mf
if __name__ == '__main__':
# Set some parameters
hlp.hi('Data merge')
orig_data_dir = hlp.DATA_DIR
hlp.set_param(log_level=3)
ae = Appevents.load_data(join(
hlp.DATA_DIR, 'mdecline_newest/m-decline_newest_appevents.csv'),
sep=';')
nf = Notifications.load(join(
hlp.DATA_DIR, 'mdecline_newest/m-decline_newest_notifications.csv'),
sep=';')
# Annotate (already scraped so set to False)
ae.add_category(scrape=False)
ae.add_time_of_day()
ae.add_date_type()
ae.strip(number_of_days=28, min_log_days=5)
# BUILD FEATURES
feature_list = []
apps = ae.get_applications()[:30].index.tolist()
from sklearn.model_selection import cross_validate, cross_val_predict, train_test_split, cross_val_score
from sklearn.dummy import DummyRegressor
from sklearn.neural_network import MLPRegressor
from sklearn.svm import SVR
from sklearn.ensemble import RandomForestRegressor, AdaBoostRegressor
from tqdm import tqdm
import matplotlib.pyplot as plt
import seaborn as sns
import xgboost as xg
from xgboost import XGBRegressor, plot_importance
import mobiledna.test.mdecline_features as mf
if __name__ == '__main__':
# Set some parameters
hlp.hi('M-decline age prediction')
hlp.set_param(data_dir=join(hlp.DATA_DIR, 'm-decline_pred'), log_level=1)
# Get data
ae = Appevents.from_pickle(join(hlp.DATA_DIR, 'ae_full_intersection.npy'))
age = pd.read_pickle(
path=join(hlp.DATA_DIR, 'full_age_vector_intersection.npy'))
age = age.sort_index()
sns.set_palette('Accent')
sns.set_style('white')
sns.distplot(age)
plt.show()
#age = age[age>35]
#selection_ids = list(age.index)
#ae.filter(users=selection_ids,inplace=True)
# m-decline grid search
import mobiledna.core.help as hlp
from os.path import join
import pandas as pd
import numpy as np
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import cross_validate
from xgboost import XGBRegressor, XGBRFRegressor, DMatrix, plot_importance,train, XGBClassifier
from sklearn.model_selection import RandomizedSearchCV, GridSearchCV, StratifiedKFold
import matplotlib.pyplot as plt
import seaborn as sns
if __name__ == '__main__':
# Set some parameters
hlp.hi('M-decline grid search')
hlp.set_param(data_dir=join(hlp.DATA_DIR, 'm-decline_pred'), log_level=1)
# Get data
age = pd.read_pickle(path=join(hlp.DATA_DIR, 'full_age_vector_intersection.npy'))
age = age.sort_index()
features = pd.read_csv(join(hlp.DATA_DIR, 'feature_matrix.csv')).set_index('id').sort_index()
X = features.values
y = age
bins = np.array([0,28,40,100])
inds = np.digitize(age,bins)
label_encoder = LabelEncoder()
y_new = label_encoder.fit_transform(inds)
xgb = XGBClassifier()