From b987fa54c76f7f146feed655f135ed7ead5559bc Mon Sep 17 00:00:00 2001 From: edwinnglabs Date: Wed, 4 Jun 2025 22:19:05 -0700 Subject: [PATCH] fix formatting with black==25.1.0 --- orbit/template/dlt.py | 7 +++---- orbit/template/ktrlite.py | 12 ++++++------ orbit/template/lgt.py | 6 +++--- tests/orbit/diagnostics/test_backtest.py | 2 +- tests/orbit/models/test_ktr.py | 12 ++++++------ tests/orbit/models/test_ktrlite.py | 16 ++++++++-------- 6 files changed, 27 insertions(+), 28 deletions(-) diff --git a/orbit/template/dlt.py b/orbit/template/dlt.py index 40c4be88..ad288d57 100644 --- a/orbit/template/dlt.py +++ b/orbit/template/dlt.py @@ -751,10 +751,9 @@ def predict( global_trend_level + global_trend_slope * idx * self._time_delta ) elif self.global_trend_option == GlobalTrendOption.loglinear.name: - full_global_trend[ - :, idx - ] = global_trend_level + global_trend_slope * np.log( - 1 + idx * self._time_delta + full_global_trend[:, idx] = ( + global_trend_level + + global_trend_slope * np.log(1 + idx * self._time_delta) ) elif self.global_trend_option == GlobalTrendOption.logistic.name: full_global_trend[:, idx] = self.global_floor + ( diff --git a/orbit/template/ktrlite.py b/orbit/template/ktrlite.py index e1d13c70..f1f6cd96 100644 --- a/orbit/template/ktrlite.py +++ b/orbit/template/ktrlite.py @@ -189,9 +189,9 @@ def set_init_values(self): init_values = None if len(self._seasonality) > 1 and self.num_of_regressors > 0: init_values = dict() - init_values[ - RegressionSamplingParameters.COEFFICIENTS_KNOT.value - ] = np.zeros((self.num_of_regressors, self.num_knots_coefficients)) + init_values[RegressionSamplingParameters.COEFFICIENTS_KNOT.value] = ( + np.zeros((self.num_of_regressors, self.num_knots_coefficients)) + ) self._init_values = init_values def _set_default_args(self): @@ -496,9 +496,9 @@ def predict( seas_regression = np.sum( seas_coef * seasonal_regressor_matrix.transpose(1, 0), axis=-2 ) - seas_decomp[ - "seasonality_{}".format(self._seasonality[idx]) - ] = seas_regression + seas_decomp["seasonality_{}".format(self._seasonality[idx])] = ( + seas_regression + ) pos += len(cols) total_seas_regression += seas_regression if include_error: diff --git a/orbit/template/lgt.py b/orbit/template/lgt.py index d28f3b25..ad4b28c9 100644 --- a/orbit/template/lgt.py +++ b/orbit/template/lgt.py @@ -231,9 +231,9 @@ def set_init_values(self): -1.0, 1.0, ) - init_values[ - LatentSamplingParameters.INITIAL_SEASONALITY.value - ] = init_sea + init_values[LatentSamplingParameters.INITIAL_SEASONALITY.value] = ( + init_sea + ) if self.num_of_positive_regressors > 0: x = np.clip( np.random.normal( diff --git a/tests/orbit/diagnostics/test_backtest.py b/tests/orbit/diagnostics/test_backtest.py index bfbcf469..d3377b93 100644 --- a/tests/orbit/diagnostics/test_backtest.py +++ b/tests/orbit/diagnostics/test_backtest.py @@ -102,7 +102,7 @@ def test_backtester_test_metrics(iclaims_training_data, metrics): "missing_flag", [False, True], ids=["full-values", "missing-values"] ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_backtester_ktr_and_missing_val(make_daily_data, missing_flag): train_df, test_df, _ = make_daily_data diff --git a/tests/orbit/models/test_ktr.py b/tests/orbit/models/test_ktr.py index 97b3d4e4..e67f3ae7 100644 --- a/tests/orbit/models/test_ktr.py +++ b/tests/orbit/models/test_ktr.py @@ -8,7 +8,7 @@ SMAPE_TOLERANCE = 0.2 -@pytest.mark.parametrize("make_daily_data", [({"seasonality": None})], indirect=True) +@pytest.mark.parametrize("make_daily_data", [{"seasonality": None}], indirect=True) def test_ktr_basic(make_daily_data): train_df, _, _ = make_daily_data @@ -101,7 +101,7 @@ def test_ktr_seasonality(make_daily_data, seasonality, seas_segments): @pytest.mark.parametrize("regressor_col", [None, ["a", "b", "c"]]) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "dual", "with_coef": True})], indirect=True + "make_daily_data", [{"seasonality": "dual", "with_coef": True}], indirect=True ) def test_ktr_regression(make_daily_data, regressor_col): train_df, test_df, coef = make_daily_data @@ -134,7 +134,7 @@ def test_ktr_regression(make_daily_data, regressor_col): [pd.date_range(start="2016-03-01", end="2019-01-01", freq="3M")], ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "dual", "with_coef": True})], indirect=True + "make_daily_data", [{"seasonality": "dual", "with_coef": True}], indirect=True ) def test_ktrx_coef_knot_dates(make_daily_data, regression_knot_dates): train_df, test_df, coef = make_daily_data @@ -167,7 +167,7 @@ def test_ktrx_coef_knot_dates(make_daily_data, regression_knot_dates): @pytest.mark.parametrize("regression_knot_distance", [90, 120]) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "dual", "with_coef": True})], indirect=True + "make_daily_data", [{"seasonality": "dual", "with_coef": True}], indirect=True ) def test_ktrx_coef_knot_distance(make_daily_data, regression_knot_distance): train_df, test_df, coef = make_daily_data @@ -203,7 +203,7 @@ def test_ktrx_coef_knot_distance(make_daily_data, regression_knot_distance): ids=["positive_only", "negative_only", "regular_only", "mixed_signs"], ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "dual", "with_coef": True})], indirect=True + "make_daily_data", [{"seasonality": "dual", "with_coef": True}], indirect=True ) def test_ktrx_regressor_sign(make_daily_data, regressor_signs): train_df, test_df, coef = make_daily_data @@ -256,7 +256,7 @@ def test_ktrx_regressor_sign(make_daily_data, regressor_signs): ], ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "dual", "with_coef": True})], indirect=True + "make_daily_data", [{"seasonality": "dual", "with_coef": True}], indirect=True ) def test_ktrx_prior_ingestion(make_daily_data, coef_prior_list): train_df, test_df, coef = make_daily_data diff --git a/tests/orbit/models/test_ktrlite.py b/tests/orbit/models/test_ktrlite.py index 3d60223e..3e289d57 100644 --- a/tests/orbit/models/test_ktrlite.py +++ b/tests/orbit/models/test_ktrlite.py @@ -13,7 +13,7 @@ "seasonality_fs_order", [None, [5]], ids=["default_order", "manual_order"] ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_ktrlite_single_seas(make_daily_data, seasonality_fs_order): train_df, _, _ = make_daily_data @@ -45,7 +45,7 @@ def test_ktrlite_single_seas(make_daily_data, seasonality_fs_order): "seasonality_fs_order", [None, [2, 5]], ids=["default_order", "manual_order"] ) @pytest.mark.parametrize( - "make_daily_data", [({"with_dual_sea": True, "with_coef": False})], indirect=True + "make_daily_data", [{"with_dual_sea": True, "with_coef": False}], indirect=True ) def test_ktrlite_dual_seas(make_daily_data, seasonality_fs_order): train_df, _, _ = make_daily_data @@ -74,7 +74,7 @@ def test_ktrlite_dual_seas(make_daily_data, seasonality_fs_order): @pytest.mark.parametrize( - "make_daily_data", [({"with_dual_sea": True, "with_coef": False})], indirect=True + "make_daily_data", [{"with_dual_sea": True, "with_coef": False}], indirect=True ) @pytest.mark.parametrize("level_segments", [20, 10, 2]) def test_ktrlite_level_segments(make_daily_data, level_segments): @@ -112,7 +112,7 @@ def test_ktrlite_level_segments(make_daily_data, level_segments): ], ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_ktrlite_level_knot_dates(make_daily_data, level_knot_dates): train_df, test_df, coef = make_daily_data @@ -141,7 +141,7 @@ def test_ktrlite_level_knot_dates(make_daily_data, level_knot_dates): @pytest.mark.parametrize("level_knot_distance", [90, 120]) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_ktrlite_level_knot_distance(make_daily_data, level_knot_distance): train_df, test_df, coef = make_daily_data @@ -175,7 +175,7 @@ def test_ktrlite_level_knot_distance(make_daily_data, level_knot_distance): ], ) @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_ktrlite_seas_segments(make_daily_data, seas_segments): train_df, test_df, coef = make_daily_data @@ -204,7 +204,7 @@ def test_ktrlite_seas_segments(make_daily_data, seas_segments): @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_ktrlite_predict_decompose(make_daily_data): train_df, test_df, coef = make_daily_data @@ -245,7 +245,7 @@ def test_ktrlite_predict_decompose(make_daily_data): @pytest.mark.parametrize( - "make_daily_data", [({"seasonality": "single", "with_coef": False})], indirect=True + "make_daily_data", [{"seasonality": "single", "with_coef": False}], indirect=True ) def test_ktrlite_predict_decompose_point_estimate(make_daily_data): train_df, test_df, coef = make_daily_data