def test_churn_predictor(self):
# Arrange
time = [1453845953 + 20000 * x for x in range(500)]
user = [1, 2, 3, 4, 5] * 20 + [1, 2, 3, 4] * 25 + [1, 2, 3] * 100
actions = turicreate.SFrame({
"user_id": user,
"timestamp": time,
"action": [1, 2, 3, 4, 5] * 100,
})
def _unix_timestamp_to_datetime(x):
import datetime
return datetime.datetime.fromtimestamp(x)
actions["timestamp"] = actions["timestamp"].apply(
_unix_timestamp_to_datetime)
actions = turicreate.TimeSeries(actions, "timestamp")
# Act
m = turicreate.churn_predictor.create(actions)
actual = [x for x in dir(m) if not x.startswith("_")]
# Assert.
expected = [
"categorical_features",
"evaluate",
"extract_features",
"get_feature_importance",
"churn_period",
"grace_period",
"features",
"get",
"is_data_aggregated",
"_list_fields",
"_list_fields",
"lookback_periods",
"model_options",
"name",
"num_features",
"num_observations",
"num_users",
"numerical_features",
"predict",
"explain",
"processed_training_data",
"save",
"show",
"summary",
"time_boundaries",
"time_period",
"trained_model",
"trained_explanation_model",
"get_churn_report",
"get_activity_baseline",
"views",
"use_advanced_features",
"user_id",
]
check_visible_modules(actual, expected)
def test_churn_predictor(self):
# Arrange
time = [1453845953 + 20000 * x for x in range(500)]
user = [1,2,3,4,5] * 20 + [1,2,3,4] * 25 + [1,2,3] * 100
actions = turicreate.SFrame({
'user_id': user,
'timestamp': time,
'action': [1,2,3,4,5] * 100,
})
def _unix_timestamp_to_datetime(x):
import datetime
return datetime.datetime.fromtimestamp(x)
actions['timestamp'] = actions['timestamp'].apply(
_unix_timestamp_to_datetime)
actions = turicreate.TimeSeries(actions, 'timestamp')
# Act
m = turicreate.churn_predictor.create(actions)
actual = [x for x in dir(m) if not x.startswith('_')]
# Assert.
expected = ['categorical_features',
'evaluate',
'extract_features',
'get_feature_importance',
'churn_period',
'grace_period',
'features',
'get',
'is_data_aggregated',
'_list_fields',
'_list_fields',
'lookback_periods',
'model_options',
'name',
'num_features',
'num_observations',
'num_users',
'numerical_features',
'predict',
'explain',
'processed_training_data',
'save',
'show',
'summary',
'time_boundaries',
'time_period',
'trained_model',
'trained_explanation_model',
'get_churn_report',
'get_activity_baseline',
'views',
'use_advanced_features',
'user_id']
check_visible_modules(actual, expected)
def get_venue_authors_links_timeseries(self):
a = self.authors_affilations_sframe["Paper ID", "Author ID"]
a = self._all_papers_sf.join(a, on="Paper ID")
a = a['datetime', 'Author ID', 'Paper publish year', 'Paper ID']
links_sf = a.join(a, on="Paper ID")
p = self.papers_sframe["Paper ID", "Paper publish year"]
links_sf.rename({'Author ID': 'src_id', 'Author ID.1': 'dst_id'})
links_sf = links_sf["src_id", "dst_id", "datetime"]
links_sf = links_sf[links_sf["src_id"] != links_sf[
"dst_id"]] # because this is a direct network we keep for each link both (u,v) and (v,u)
return tc.TimeSeries(links_sf, index="datetime")
def get_venue_authors_timeseries(self):
p = self._all_papers_sf["Paper ID", "Paper publish year"]
a = self.authors_affilations_sframe["Paper ID", "Author ID"]
sf = p.join(a, on="Paper ID")["Author ID", "Paper publish year"]
sf = sf.groupby(
"Author ID", {
"mindate": agg.MIN("Paper publish year"),
"maxdate": agg.MAX("Paper publish year")
})
sf.rename({"Author ID": "v_id"})
sf["mindate"] = sf["mindate"].apply(
lambda y: datetime(year=y, month=1, day=1))
sf["maxdate"] = sf["maxdate"].apply(
lambda y: datetime(year=y, month=1, day=1))
if sf.num_rows() == 0:
return None
return tc.TimeSeries(sf, index="mindate")