def test_operations_evaluated_for_each_reference(self,
mock_fetch_data: Mock,
*mocks):
eoe = ElectionOverElection(mock_dataset.fields.timestamp)
mock_operation = Mock(name='mock_operation ', spec=f.Operation)
mock_operation.alias, mock_operation.definition = 'mock_operation', mock_dataset.table.abc
mock_operation.metrics = []
mock_widget = f.Widget(mock_operation)
mock_widget.transform = Mock()
mock_df = {}
mock_fetch_data.return_value = mock_df
# Need to keep widget the last call in the chain otherwise the object gets cloned and the assertion won't work
mock_dataset.query \
.dimension(mock_dataset.fields.timestamp) \
.reference(eoe) \
.widget(mock_widget) \
.fetch()
mock_operation.apply.assert_has_calls([
call(mock_df, None),
call(mock_df, eoe),
])
def test_apply_to_timeseries_with_uni_dim_and_ref(self):
rolling_mean = RollingMean(mock_dataset.fields.wins, 3)
result = rolling_mean.apply(dimx2_date_str_ref_df, ElectionOverElection(mock_dataset.fields.timestamp))
expected = pd.Series([nan, nan, nan, nan, 4 / 3, 0., 4 / 3, 2 / 3, 4 / 3, 4 / 3, 2 / 3],
name='$wins_eoe',
index=dimx2_date_str_ref_df.index)
pandas.testing.assert_series_equal(expected, result)
def test_apply_to_timeseries_with_uni_dim_and_ref(self):
cumprod = CumProd(mock_dataset.fields.wins)
result = cumprod.apply(
dimx2_date_str_ref_df,
ElectionOverElection(mock_dataset.fields.timestamp))
expected = pd.Series([2.0] + [0.0] * 10,
name='$wins_eoe',
index=dimx2_date_str_ref_df.index)
pandas.testing.assert_series_equal(expected, result)
def test_time_series_ref(self):
dimensions = [mock_dataset.fields.timestamp, mock_dataset.fields.political_party]
references = [ElectionOverElection(mock_dataset.fields.timestamp)]
result = Pandas(mock_dataset.fields.votes).transform(dimx2_date_str_ref_df, dimensions, references)
expected = dimx2_date_str_ref_df.copy()[[f('votes'), f('votes_eoe')]]
expected.index.names = ['Timestamp', 'Party']
expected.columns = ['Votes', 'Votes EoE']
expected.columns.name = 'Metrics'
expected = expected.applymap(format_float)
pandas.testing.assert_frame_equal(expected, result)
def test_time_series_multi_ref(self):
query_dimensions = [mock_dataset.fields.timestamp, mock_dataset.fields.political_party]
query_references = [ElectionOverElection(mock_dataset.fields.timestamp)]
result = CSV(mock_dataset.fields.votes, mock_dataset.fields.wins) \
.transform(dimx2_date_str_ref_df, mock_dataset, query_dimensions, query_references)
expected = dimx2_date_str_ref_df.copy()[[f('votes'), f('votes_eoe'), f('wins'), f('wins_eoe')]]
expected.index.names = ['Timestamp', 'Party']
expected.columns = ['Votes', 'Votes EoE', 'Wins', 'Wins EoE']
expected = expected.applymap(_format_float)
self.assertEqual(expected.to_csv(**csv_options), result)
def test_apply_to_timeseries_with_uni_dim_and_ref(self):
cummean = CumMean(mock_dataset.fields.votes)
result = cummean.apply(
dimx2_date_str_ref_df,
ElectionOverElection(mock_dataset.fields.timestamp))
expected = pd.Series([
7579518.0, 1076384.0, 7072032.5, 4685666.5, 7503711.0,
6316507.333333333, 8136969.0, 7688157.0, 8407797.0, 8635351.2,
8364511.166666667
],
name='$votes_eoe',
index=dimx2_date_str_ref_df.index)
pandas.testing.assert_series_equal(expected, result)
def test_hidden_ref_dimx2_date_str(self):
dimensions = [
mock_dataset.fields.timestamp, mock_dataset.fields.political_party
]
references = [ElectionOverElection(mock_dataset.fields.timestamp)]
result = CSV(mock_dataset.fields.votes,
hide=['votes_eoe']).transform(dimx2_date_str_ref_df,
dimensions, references)
expected = dimx2_date_str_ref_df.copy()[[f('votes')]]
expected.index.names = ['Timestamp', 'Party']
expected.columns = ['Votes']
expected.columns.name = 'Metrics'
expected = expected.applymap(format_float_raw)
self.assertEqual(expected.to_csv(**csv_options), result)