def test_normalize_ww_init():
es = EntitySet()
df = pd.DataFrame({
'id': [1, 2, 3, 4],
'col': ['a', 'b', 'c', 'd'],
'df2_id': [1, 1, 2, 2],
'df2_col': [True, False, True, True]
})
df.ww.init(index='id', name='test_name')
es.add_dataframe(dataframe=df)
assert es['test_name'].ww.name == 'test_name'
assert es['test_name'].ww.schema.name == 'test_name'
es.normalize_dataframe('test_name',
'new_df',
'df2_id',
additional_columns=['df2_col'])
assert es['test_name'].ww.name == 'test_name'
assert es['test_name'].ww.schema.name == 'test_name'
assert es['new_df'].ww.name == 'new_df'
assert es['new_df'].ww.schema.name == 'new_df'
def test_normalize_dataframe():
df = pd.DataFrame({
'id':
range(4),
'full_name':
['Mr. John Doe', 'Doe, Mrs. Jane', 'James Brown', 'Ms. Paige Turner'],
'email': [
'*****@*****.**', np.nan, '*****@*****.**',
'*****@*****.**'
],
'phone_number':
['5555555555', '555-555-5555', '1-(555)-555-5555', '555-555-5555'],
'age':
pd.Series([33, None, 33, 57], dtype='Int64'),
'signup_date': [pd.to_datetime('2020-09-01')] * 4,
'is_registered':
pd.Series([True, False, True, None], dtype='boolean'),
})
df.ww.init(name='first_table', index='id', time_index='signup_date')
es = EntitySet('es')
es.add_dataframe(df)
es.normalize_dataframe('first_table',
'second_table',
'age',
additional_columns=['phone_number', 'full_name'],
make_time_index=True)
assert len(es.dataframe_dict) == 2
assert 'foreign_key' in es['first_table'].ww.semantic_tags['age']
def test_normalize_ww_init():
es = EntitySet()
df = pd.DataFrame({
"id": [1, 2, 3, 4],
"col": ["a", "b", "c", "d"],
"df2_id": [1, 1, 2, 2],
"df2_col": [True, False, True, True],
})
df.ww.init(index="id", name="test_name")
es.add_dataframe(dataframe=df)
assert es["test_name"].ww.name == "test_name"
assert es["test_name"].ww.schema.name == "test_name"
es.normalize_dataframe("test_name",
"new_df",
"df2_id",
additional_columns=["df2_col"])
assert es["test_name"].ww.name == "test_name"
assert es["test_name"].ww.schema.name == "test_name"
assert es["new_df"].ww.name == "new_df"
assert es["new_df"].ww.schema.name == "new_df"
def test_normalize_dataframe():
df = pd.DataFrame({
"id":
range(4),
"full_name": [
"Mr. John Doe",
"Doe, Mrs. Jane",
"James Brown",
"Ms. Paige Turner",
],
"email": [
"*****@*****.**",
np.nan,
"*****@*****.**",
"*****@*****.**",
],
"phone_number": [
"5555555555",
"555-555-5555",
"1-(555)-555-5555",
"555-555-5555",
],
"age":
pd.Series([33, None, 33, 57], dtype="Int64"),
"signup_date": [pd.to_datetime("2020-09-01")] * 4,
"is_registered":
pd.Series([True, False, True, None], dtype="boolean"),
})
df.ww.init(name="first_table", index="id", time_index="signup_date")
es = EntitySet("es")
es.add_dataframe(df)
es.normalize_dataframe(
"first_table",
"second_table",
"age",
additional_columns=["phone_number", "full_name"],
make_time_index=True,
)
assert len(es.dataframe_dict) == 2
assert "foreign_key" in es["first_table"].ww.semantic_tags["age"]