autogluon.timeseries 0.8.3b20231025__py3-none-any.whl → 0.8.3b20231027__py3-none-any.whl

autogluon/timeseries/metrics/point.py

@@ -64,7 +64,8 @@ class WAPE(TimeSeriesScorer):
 
 
 class sMAPE(TimeSeriesScorer):
-    "Symmetric mean absolute percentage error."
+    """Symmetric mean absolute percentage error."""
+
     optimized_by_median = True
     equivalent_tabular_regression_metric = "symmetric_mean_absolute_percentage_error"
 
@@ -76,7 +77,8 @@ class sMAPE(TimeSeriesScorer):
 
 
 class MAPE(TimeSeriesScorer):
-    "Mean Absolute Percentage Error."
+    """Mean Absolute Percentage Error."""
+
     optimized_by_median = True
     equivalent_tabular_regression_metric = "mean_absolute_percentage_error"
 

autogluon/timeseries/models/autogluon_tabular/mlforecast.py

@@ -174,7 +174,8 @@ class AbstractMLForecastModel(AbstractTimeSeriesModel):
             data = data.query("item_id in @items_to_keep")
 
         mlforecast_df = self._to_mlforecast_df(data, data.static_features)
-        df = self._mlf.preprocess(mlforecast_df, dropna=False)
+        # Unless we set static_features=[], MLForecast interprets all known covariates as static features
+        df = self._mlf.preprocess(mlforecast_df, dropna=False, static_features=[])
         # df.query results in 2x memory saving compared to df.dropna(subset="y")
         df = df.query("y.notnull()")
 
@@ -366,7 +367,7 @@ class DirectTabularModel(AbstractMLForecastModel):
         data_future[self.target] = float("inf")
         data_extended = pd.concat([data, data_future])
         mlforecast_df = self._to_mlforecast_df(data_extended, data.static_features)
-        df = self._mlf.preprocess(mlforecast_df, dropna=False)
+        df = self._mlf.preprocess(mlforecast_df, dropna=False, static_features=[])
         df = df.groupby(MLF_ITEMID, sort=False).tail(self.prediction_length)
         df = df.replace(float("inf"), float("nan"))
 
@@ -376,7 +377,7 @@ class DirectTabularModel(AbstractMLForecastModel):
 
         if hasattr(self._mlf.ts, "target_transforms"):
             # Ensure that transforms are fitted only on past data
-            self._mlf.preprocess(self._to_mlforecast_df(data, None))
+            self._mlf.preprocess(self._to_mlforecast_df(data, None), static_features=[])
             for tfm in self._mlf.ts.target_transforms[::-1]:
                 predictions = tfm.inverse_transform(predictions)
         predictions = predictions.rename(columns={MLF_ITEMID: ITEMID, MLF_TIMESTAMP: TIMESTAMP}).set_index(
@@ -405,12 +406,9 @@ class DirectTabularModel(AbstractMLForecastModel):
                 "eval_metric": "pinball_loss",
             }
         else:
-            tabular_metric = self.eval_metric.equivalent_tabular_regression_metric
-            if tabular_metric is None:
-                tabular_metric = "mean_absolute_error"
             return {
                 "problem_type": ag.constants.REGRESSION,
-                "eval_metric": tabular_metric,
+                "eval_metric": self.eval_metric.equivalent_tabular_regression_metric or "mean_absolute_error",
             }
 
 
@@ -470,15 +468,18 @@ class RecursiveTabularModel(AbstractMLForecastModel):
         from scipy.stats import norm
 
         new_df = self._to_mlforecast_df(data, data.static_features)
-        if known_covariates is not None:
-            dynamic_dfs = [self._to_mlforecast_df(known_covariates, data.static_features, include_target=False)]
-        else:
-            dynamic_dfs = None
+        if known_covariates is None:
+            future_index = get_forecast_horizon_index_ts_dataframe(data, self.prediction_length)
+            known_covariates = pd.DataFrame(columns=[self.target], index=future_index, dtype="float32")
+        X_df = self._to_mlforecast_df(known_covariates, data.static_features, include_target=False)
+        # If both covariates & static features are missing, set X_df = None to avoid exception from MLForecast
+        if len(X_df.columns.difference([MLF_ITEMID, MLF_TIMESTAMP])) == 0:
+            X_df = None
         with warning_filter():
             raw_predictions = self._mlf.predict(
                 h=self.prediction_length,
                 new_df=new_df,
-                dynamic_dfs=dynamic_dfs,
+                X_df=X_df,
             )
         predictions = raw_predictions.rename(columns={MLF_ITEMID: ITEMID, MLF_TIMESTAMP: TIMESTAMP})
 
@@ -495,10 +496,7 @@ class RecursiveTabularModel(AbstractMLForecastModel):
         return TimeSeriesDataFrame(predictions).reindex(data.item_ids, level=ITEMID)
 
     def _get_extra_tabular_init_kwargs(self) -> dict:
-        tabular_metric = self.eval_metric.equivalent_tabular_regression_metric
-        if tabular_metric is None:
-            tabular_metric = "mean_absolute_error"
         return {
             "problem_type": ag.constants.REGRESSION,
-            "eval_metric": tabular_metric,
+            "eval_metric": self.eval_metric.equivalent_tabular_regression_metric or "mean_absolute_error",
         }

autogluon/timeseries/version.py

@@ -1,3 +1,3 @@
 """This is the autogluon version file."""
-__version__ = '0.8.3b20231025'
+__version__ = '0.8.3b20231027'
 __lite__ = False

{autogluon.timeseries-0.8.3b20231025.dist-info → autogluon.timeseries-0.8.3b20231027.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: autogluon.timeseries
-Version: 0.8.3b20231025
+Version: 0.8.3b20231027
 Summary: AutoML for Image, Text, and Tabular Data
 Home-page: https://github.com/autogluon/autogluon
 Author: AutoGluon Community
@@ -45,12 +45,13 @@ Requires-Dist: statsmodels <0.15,>=0.13.0
 Requires-Dist: gluonts <0.14,>=0.13.1
 Requires-Dist: networkx <4,>=3.0
 Requires-Dist: statsforecast <1.5,>=1.4.0
-Requires-Dist: mlforecast <0.9.4,>=0.9.3
+Requires-Dist: mlforecast <0.10.1,>=0.10.0
+Requires-Dist: utilsforecast <0.0.11,>=0.0.10
 Requires-Dist: tqdm <5,>=4.38
 Requires-Dist: ujson <6,>=5
-Requires-Dist: autogluon.core[raytune] ==0.8.3b20231025
-Requires-Dist: autogluon.common ==0.8.3b20231025
-Requires-Dist: autogluon.tabular[catboost,lightgbm,xgboost] ==0.8.3b20231025
+Requires-Dist: autogluon.core[raytune] ==0.8.3b20231027
+Requires-Dist: autogluon.common ==0.8.3b20231027
+Requires-Dist: autogluon.tabular[catboost,lightgbm,xgboost] ==0.8.3b20231027
 Provides-Extra: all
 Provides-Extra: tests
 Requires-Dist: pytest ; extra == 'tests'

{autogluon.timeseries-0.8.3b20231025.dist-info → autogluon.timeseries-0.8.3b20231027.dist-info}/RECORD

@@ -1,17 +1,17 @@
-autogluon.timeseries-0.8.3b20231025-py3.8-nspkg.pth,sha256=cQGwpuGPqg1GXscIwt-7PmME1OnSpD-7ixkikJ31WAY,554
+autogluon.timeseries-0.8.3b20231027-py3.8-nspkg.pth,sha256=cQGwpuGPqg1GXscIwt-7PmME1OnSpD-7ixkikJ31WAY,554
 autogluon/timeseries/__init__.py,sha256=oGfAdHmGz9zGFH53Q4zDL42CavzjqFaWTgkx_vg17QM,370
 autogluon/timeseries/evaluator.py,sha256=CPqmtcPGVpD1ORydjAl65XS2Oq1Ua8q0IT1zX1cXv2Y,2206
 autogluon/timeseries/learner.py,sha256=LSoKWgWhyiwt4l-3IUSLk4EI7f-xiEQxH6kOBJ8Snmc,9361
 autogluon/timeseries/predictor.py,sha256=RQ2qGmWvNhjRugfafvfRxOK6LKgALVMAjM2QvY0YwmE,55808
 autogluon/timeseries/splitter.py,sha256=il3Z-75dNS2VeuejtzSP_V-bJYOwLWKNrbIOrz1dHnE,11589
-autogluon/timeseries/version.py,sha256=sV8LFJ8JVp9Hl4KfYLJNFcct1pIuG1I8_vKVTOLnP3c,90
+autogluon/timeseries/version.py,sha256=jbOZdK5SimX4-T8ajoRhDzQ2_mgCI5J7T7DtVbG7Obk,90
 autogluon/timeseries/configs/__init__.py,sha256=BTtHIPCYeGjqgOcvqb8qPD4VNX-ICKOg6wnkew1cPOE,98
 autogluon/timeseries/configs/presets_configs.py,sha256=mX0V5zajWWArVforLvbyr6W-JMsQBp2AkBqlWVP2Zuw,640
 autogluon/timeseries/dataset/__init__.py,sha256=UvnhAN5tjgxXTHoZMQDy64YMDj4Xxa68yY7NP4vAw0o,81
 autogluon/timeseries/dataset/ts_dataframe.py,sha256=bobu5rOplC1z0z8IU0RjPOloLGN9kbQx6Ncq5G4bEBk,41063
 autogluon/timeseries/metrics/__init__.py,sha256=a1yWUaYNnCboPDMtshw4l7LCDKw9OaloY6mZR_Drppo,1670
 autogluon/timeseries/metrics/abstract.py,sha256=BLH6VRcgsSHwSLiB1fAPzh_Q_N_3a19wob78w29JPcE,8055
-autogluon/timeseries/metrics/point.py,sha256=W-7mLM6xmHcJwvpSJvVmw-Hzc4b7TC_BSFArDU-25-E,5783
+autogluon/timeseries/metrics/point.py,sha256=xLserP_Zy1g_3vg8Rbm8PhDfF-wPS-C3fN67ka9-Zls,5793
 autogluon/timeseries/metrics/quantile.py,sha256=yfffrGJdmMFffeAjBng52OatFbYm5dq0CA89ZWIT8iA,847
 autogluon/timeseries/metrics/utils.py,sha256=eJ63TCR-UwbeJ1c2Qm7B2q-8B3sFthPgiooEccrf2Kc,912
 autogluon/timeseries/models/__init__.py,sha256=qjZKN0__rewaBgshr66lexmNskNQ0mXilgz8xF2c-pE,864
@@ -20,7 +20,7 @@ autogluon/timeseries/models/abstract/__init__.py,sha256=wvDsQAZIV0N3AwBeMaGItoQ8
 autogluon/timeseries/models/abstract/abstract_timeseries_model.py,sha256=hwxx1tW3Zuc0shpDMMH5dBwKyU8cyfqF8eg7w6ffdjo,19701
 autogluon/timeseries/models/abstract/model_trial.py,sha256=_5Nrk4CrG3u35tTd3elekfdnQI2Pn3P9AGS5CE6nuyg,3749
 autogluon/timeseries/models/autogluon_tabular/__init__.py,sha256=r9i6jWcyeLHYClkcMSKRVsfrkBUMxpDrTATNTBc_qgQ,136
-autogluon/timeseries/models/autogluon_tabular/mlforecast.py,sha256=dgs9qZ8mAfUSii025IiAYRuRgcmVyvwOg4e7xlmBR9A,23304
+autogluon/timeseries/models/autogluon_tabular/mlforecast.py,sha256=sWYkCjTapfCAvM75ui4mFxh-P6-G1ewmHLX6JKEZEIk,23591
 autogluon/timeseries/models/autogluon_tabular/utils.py,sha256=4-gTrBtizxeMVQlsuscugPqw9unaXWXhS1TVVssfzYY,2125
 autogluon/timeseries/models/ensemble/__init__.py,sha256=kFr11Gmt7lQJu9Rr8HuIPphQN5l1TsoorfbJm_O3a_s,128
 autogluon/timeseries/models/ensemble/abstract_timeseries_ensemble.py,sha256=b6o_kJsyk5_uCDcgSKMpqugViMJ0t3AhIYdysG3U-Uc,3393
@@ -48,11 +48,11 @@ autogluon/timeseries/utils/datetime/base.py,sha256=MsqIHY14m3QMjSwwtE7Uo1oNwepWU
 autogluon/timeseries/utils/datetime/lags.py,sha256=kcU4liKbHj7KP2ajNU-KLZ8OYSU35EgT4kJjZNSw0Zg,5875
 autogluon/timeseries/utils/datetime/seasonality.py,sha256=kgK_ukw2wCviEB7CZXRVC5HZpBJZu9IsRrvCJ9E_rOE,755
 autogluon/timeseries/utils/datetime/time_features.py,sha256=pROkYyxETQ8rHKfPGhf2paB73C7rWJ2Ui0cCswLqbBg,2562
-autogluon.timeseries-0.8.3b20231025.dist-info/LICENSE,sha256=CeipvOyAZxBGUsFoaFqwkx54aPnIKEtm9a5u2uXxEws,10142
-autogluon.timeseries-0.8.3b20231025.dist-info/METADATA,sha256=4DW3fOHGzPxH0Mr9iOkC4eQChGY2Rlg8r99naVnwYWg,13246
-autogluon.timeseries-0.8.3b20231025.dist-info/NOTICE,sha256=7nPQuj8Kp-uXsU0S5so3-2dNU5EctS5hDXvvzzehd7E,114
-autogluon.timeseries-0.8.3b20231025.dist-info/WHEEL,sha256=yQN5g4mg4AybRjkgi-9yy4iQEFibGQmlz78Pik5Or-A,92
-autogluon.timeseries-0.8.3b20231025.dist-info/namespace_packages.txt,sha256=giERA4R78OkJf2ijn5slgjURlhRPzfLr7waIcGkzYAo,10
-autogluon.timeseries-0.8.3b20231025.dist-info/top_level.txt,sha256=giERA4R78OkJf2ijn5slgjURlhRPzfLr7waIcGkzYAo,10
-autogluon.timeseries-0.8.3b20231025.dist-info/zip-safe,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
-autogluon.timeseries-0.8.3b20231025.dist-info/RECORD,,
+autogluon.timeseries-0.8.3b20231027.dist-info/LICENSE,sha256=CeipvOyAZxBGUsFoaFqwkx54aPnIKEtm9a5u2uXxEws,10142
+autogluon.timeseries-0.8.3b20231027.dist-info/METADATA,sha256=CTkjrgdgUUPxvJu93h2_Bj-PGBYyzr9cCFGTmqRMzAM,13294
+autogluon.timeseries-0.8.3b20231027.dist-info/NOTICE,sha256=7nPQuj8Kp-uXsU0S5so3-2dNU5EctS5hDXvvzzehd7E,114
+autogluon.timeseries-0.8.3b20231027.dist-info/WHEEL,sha256=yQN5g4mg4AybRjkgi-9yy4iQEFibGQmlz78Pik5Or-A,92
+autogluon.timeseries-0.8.3b20231027.dist-info/namespace_packages.txt,sha256=giERA4R78OkJf2ijn5slgjURlhRPzfLr7waIcGkzYAo,10
+autogluon.timeseries-0.8.3b20231027.dist-info/top_level.txt,sha256=giERA4R78OkJf2ijn5slgjURlhRPzfLr7waIcGkzYAo,10
+autogluon.timeseries-0.8.3b20231027.dist-info/zip-safe,sha256=AbpHGcgLb-kRsJGnwFEktk7uzpZOCcBY74-YBdrKVGs,1
+autogluon.timeseries-0.8.3b20231027.dist-info/RECORD,,