def test_add_missing_row_index_date_error():
"""
It should add missing row compare to a date_range
"""
input_df = pd.read_csv(
os.path.join(fixtures_base_dir, 'add_missing_row.csv'))
with pytest.raises(ParamsValueError) as e_info:
add_missing_row(input_df,
id_cols=['City', 'Country', 'Region'],
reference_col='Year',
complete_index={'type': 'my_date'})
assert "Unknown complete index type: my_date" == str(e_info.value)
def test_add_missing_row():
"""
It should add missing row compare to a reference column
"""
input_df = pd.read_csv(
os.path.join(fixtures_base_dir, 'add_missing_row.csv'))
new_df = add_missing_row(input_df,
id_cols=['City', 'Country', 'Region'],
reference_col='Year')
assert len(new_df) == 12
input_df = input_df.drop(['Country', 'Region'], axis=1).drop_duplicates()
new_df = add_missing_row(input_df, id_cols=['City'], reference_col='Year')
assert len(new_df) == 12
def test_add_missing_row_use_index():
"""
It should add missing row using the index provided
"""
input_df = pd.read_csv(
os.path.join(fixtures_base_dir, 'add_missing_row.csv'))
new_df = add_missing_row(input_df,
id_cols=['City', 'Country', 'Region'],
reference_col='Year',
complete_index=('2009', '2010', '2011', '2012'))
assert new_df.shape[0] == 16
def test_add_missing_row_cols_to_keep():
"""
It should add missing row compare to a reference column and keep an other column
"""
input_df = pd.read_csv(
os.path.join(fixtures_base_dir, 'add_missing_row_2.csv'))
new_df = add_missing_row(input_df,
id_cols=['group'],
reference_col='date',
cols_to_keep=['month'])
mask = (new_df['group'] == 'B') & (new_df['date'] == 20161001)
assert len(new_df.loc[mask, 'month']) == 1
assert new_df.loc[mask, 'month'].iloc[0] == 'octobre'
assert math.isnan(new_df.loc[mask, 'value'].iloc[0])
def test_add_missing_row_between_and_before():
"""
It should add missing row compare to a reference column that are
smaller than max of each group
"""
input_df = pd.read_csv(
os.path.join(fixtures_base_dir, 'add_missing_row.csv'))
expected = [2010, 2011]
new_df = add_missing_row(input_df,
id_cols=['City', 'Country', 'Region'],
reference_col='Year',
method='between_and_before')
result = new_df.loc[new_df['City'] == 'Nantes', 'Year'].unique().tolist()
result.sort()
assert result == expected
def test_add_missing_row_index_date_range():
"""
It should add missing row compare to a date_range
"""
input_df = pd.read_csv(
os.path.join(fixtures_base_dir, 'add_missing_row.csv'))
input_df['Year'] = input_df['Year'].astype(str)
complete_index = {
'type': 'date',
'format': '%Y',
'start': '2010',
'end': '2013',
'freq': {
'years': 1
}
}
new_df = add_missing_row(input_df,
id_cols=['City', 'Country', 'Region'],
reference_col='Year',
complete_index=complete_index)
assert len(new_df) == 16
