cleanup

atom/basemodel.py

@@ -3014,17 +3014,25 @@ def _prediction(
             called.
 
         """
-        Xt, yt = self.transform(X, y, verbose=verbose)
+        if y is not None or X is not None:
+            if isinstance(out := self.transform(X, y, verbose=verbose), tuple):
+                Xt, yt = out
+            elif X is not None:
+                Xt, yt = out, y
+            else:
+                Xt, yt = X, out
+        else:
+            Xt, yt = X, y
 
         if method != "score":
             fh = kwargs.get("fh")
             if fh is not None and not isinstance(fh, ForecastingHorizon):
                 kwargs["fh"] = self.branch._get_rows(fh).index
 
             if "y" in sign(func := getattr(self.estimator, method)):
-                return self.memory.cache(func)(fh=fh, y=yt, X=Xt, **kwargs)
+                return self.memory.cache(func)(y=yt, X=Xt, **kwargs)
             else:
-                return self.memory.cache(func)(fh=fh, X=Xt, **kwargs)
+                return self.memory.cache(func)(X=Xt, **kwargs)
         else:
             if metric is None:
                 scorer = self._metric[0]

atom/baserunner.py

@@ -40,7 +40,8 @@
     Bool, DataFrame, FloatZeroToOneExc, HarmonicsSelector, Int, IntLargerOne,
     MetricConstructor, Model, ModelSelector, ModelsSelector, Pandas,
     RowSelector, Scalar, Seasonality, Segment, Sequence, Series,
-    TargetSelector, YSelector, dataframe_t, int_t, segment_t, sequence_t,
+    TargetSelector, YSelector, bool_t, dataframe_t, int_t, segment_t,
+    sequence_t,
 )
 from atom.utils.utils import (
     ClassMap, DataContainer, Goal, SeasonalPeriod, Task, bk, check_is_fitted,
@@ -888,9 +889,18 @@ def _delete_models(self, models: str | Model | Sequence[str | Model]):
             self._metric = ClassMap()
 
     @crash
-    def available_models(self) -> pd.DataFrame:
+    def available_models(self, **kwargs) -> pd.DataFrame:
         """Give an overview of the available predefined models.
 
+        Parameters
+        ----------
+        **kwargs
+            Filter the returned models providing any of the column as
+            keyword arguments, where the value is the desired filter,
+            e.g., `accepts_sparse=True`, to get all models that accept
+            sparse input or `supports_engines="cuml"` to get all models
+            that support the [cuML][] engine.
+
         Returns
         -------
         pd.DataFrame
@@ -902,8 +912,8 @@ def available_models(self) -> pd.DataFrame:
             - **estimator:** Name of the model's underlying estimator.
             - **module:** The estimator's module.
             - **handles_missing:** Whether the model can handle missing
-              (`NaN`) values without preprocessing. If False, consider using
-              the [Imputer][] class before training the models.
+              values without preprocessing. If False, consider using the
+              [Imputer][] class before training the models.
             - **needs_scaling:** Whether the model requires feature scaling.
               If True, [automated feature scaling][] is applied.
             - **accepts_sparse:** Whether the model accepts [sparse input][sparse-datasets].
@@ -922,7 +932,16 @@ def available_models(self) -> pd.DataFrame:
         for model in MODELS:
             m = model(goal=self._goal, branches=self._branches)
             if self._goal.name in m._estimators:
-                rows.append(m.get_tags())
+                tags = m.get_tags()
+
+                for key, value in kwargs.items():
+                    k = tags.get(key)
+                    if isinstance(value, bool_t) and value is not bool(k):
+                        break
+                    elif isinstance(value, str) and not re.search(value, k, re.I):
+                        break
+                else:
+                    rows.append(tags)
 
         return pd.DataFrame(rows)
 

atom/models/ts.py

@@ -21,7 +21,7 @@
 
 
 class ARIMA(ForecastModel):
-    """Autoregressive Integrated Moving Average Model.
+    """Autoregressive Integrated Moving Average.
 
     Seasonal ARIMA models and exogenous input is supported, hence this
     estimator is capable of fitting SARIMA, ARIMAX, and SARIMAX.
@@ -178,7 +178,7 @@ def _get_distributions(self) -> dict[str, BaseDistribution]:
 
 
 class AutoARIMA(ForecastModel):
-    """Automatic Autoregressive Integrated Moving Average Model.
+    """Automatic Autoregressive Integrated Moving Average.
 
     [ARIMA][] implementation that includes automated fitting of
     (S)ARIMA(X) hyperparameters (p, d, q, P, D, Q). The AutoARIMA
@@ -649,7 +649,7 @@ def _get_distributions(self) -> dict[str, BaseDistribution]:
 
 
 class MSTL(ForecastModel):
-    """Multiple Seasonal-Trend decomposition using LOESS model.
+    """Multiple Seasonal-Trend decomposition using LOESS.
 
     The MSTL decomposes the time series in multiple seasonalities using
     LOESS. Then forecasts the trend using a custom non-seasonal model
@@ -956,7 +956,7 @@ def _get_distributions() -> dict[str, BaseDistribution]:
 
 
 class SARIMAX(ForecastModel):
-    """Seasonal Autoregressive Integrated Moving Average with eXogenous factors.
+    """Seasonal Autoregressive Integrated Moving Average.
 
     SARIMAX stands for Seasonal Autoregressive Integrated Moving Average
     with eXogenous factors. It extends [ARIMA][] by incorporating seasonal
@@ -1106,7 +1106,7 @@ def _get_distributions(self) -> dict[str, BaseDistribution]:
 
 
 class STL(ForecastModel):
-    """Seasonal-Trend decomposition using Loess.
+    """Seasonal-Trend decomposition using LOESS.
 
     STL is a technique commonly used for decomposing time series data
     into components like trend, seasonality, and residuals.
@@ -1381,7 +1381,7 @@ def _get_distributions() -> dict[str, BaseDistribution]:
 
 
 class VARMAX(ForecastModel):
-    """Vector Autoregressive Moving-Average with exogenous variables.
+    """Vector Autoregressive Moving-Average.
 
     VARMAX is an extension of the [VAR][] model that incorporates not
     only lagged values of the endogenous variables, but also includes

atom/pipeline.py

@@ -477,6 +477,9 @@ def transform(
             Transformed target column. Only returned if provided.
 
         """
+        if X is None and y is None:
+            raise ValueError("X and y cannot be both None.")
+
         for _, _, transformer in self._iter(**kwargs):
             with adjust_verbosity(transformer, self.verbose):
                 X, y = self._mem_transform(transformer, X, y)
@@ -520,6 +523,9 @@ def inverse_transform(
             Transformed target column. Only returned if provided.
 
         """
+        if X is None and y is None:
+            raise ValueError("X and y cannot be both None.")
+
         for _, _, transformer in reversed(list(self._iter())):
             with adjust_verbosity(transformer, self.verbose):
                 X, y = self._mem_transform(transformer, X, y, method="inverse_transform")

atom/plots/predictionplot.py

@@ -967,10 +967,11 @@ def plot_feature_importance(
     def plot_forecast(
         self,
         models: ModelsSelector = None,
-        fh: RowSelector | ForecastingHorizon = "test",
+        fh: RowSelector | ForecastingHorizon = "dataset",
         X: XSelector | None = None,
         target: TargetSelector = 0,
         *,
+        plot_insample: Bool = False,
         plot_interval: Bool = True,
         title: str | dict[str, Any] | None = None,
         legend: Legend | dict[str, Any] | None = "upper left",
@@ -988,7 +989,7 @@ def plot_forecast(
         models: int, str, Model, segment, sequence or None, default=None
             Models to plot. If None, all models are selected.
 
-        fh: hashable, segment, sequence, dataframe or [ForecastingHorizon][], default="test"
+        fh: hashable, segment, sequence, dataframe or [ForecastingHorizon][], default="dataset"
             The [forecasting horizon][row-and-column-selection] for
             which to plot the predictions.
 
@@ -999,6 +1000,10 @@ def plot_forecast(
         target: int or str, default=0
             Target column to look at. Only for [multivariate][] tasks.
 
+        plot_insample: bool, default=False
+            Whether to draw in-sample predictions (predictions on the training
+            set). Models that do not support this feature are silently skipped.
+
         plot_interval: bool, default=True
             Whether to plot prediction intervals together with the exact
             predicted values. Models wihtout a `predict_interval` method
@@ -1040,7 +1045,7 @@ def plot_forecast(
         --------
         atom.plots:DataPlot.plot_distribution
         atom.plots:DataPlot.plot_series
-        atom.plots:PredictionPlot.plot_roc
+        atom.plots:PredictionPlot.plot_errors
 
         Examples
         --------
@@ -1070,7 +1075,7 @@ def plot_forecast(
             fh = self.branch._get_rows(fh).index
 
         if X is None:
-            X = self.branch.X.loc[fh]
+            X = self.branch._all.loc[fh]
         else:
             X = self.transform(X)
 
@@ -1083,9 +1088,12 @@ def plot_forecast(
             if self.task.is_multioutput:
                 y_pred = y_pred[target_c]
 
+            if not plot_insample:
+                y_pred.loc[m.branch.train.index] = np.NaN
+
             fig.add_trace(
                 self._draw_line(
-                    x=self._get_plot_index(y_pred),
+                    x=(x := self._get_plot_index(y_pred)),
                     y=y_pred,
                     mode="lines+markers",
                     parent=m.name,
@@ -1098,7 +1106,7 @@ def plot_forecast(
             if plot_interval:
                 try:
                     y_pred = m.predict_interval(fh=fh, X=X)
-                except NotImplementedError:
+                except (AttributeError, NotImplementedError):
                     continue  # Fails for some models like ES
 
                 if self.task.is_multioutput:
@@ -1107,10 +1115,13 @@ def plot_forecast(
                 else:
                     y = y_pred  # Univariate
 
+                if not plot_insample:
+                    y_pred.loc[m.branch.train.index] = np.NaN
+
                 fig.add_traces(
                     [
                         go.Scatter(
-                            x=self._get_plot_index(y_pred),
+                            x=x,
                             y=y.iloc[:, 1],
                             mode="lines",
                             line={"width": 1, "color": BasePlot._fig.get_elem(m.name)},
@@ -1121,7 +1132,7 @@ def plot_forecast(
                             yaxis=yaxis,
                         ),
                         go.Scatter(
-                            x=self._get_plot_index(y_pred),
+                            x=x,
                             y=y.iloc[:, 0],
                             mode="lines",
                             line={"width": 1, "color": BasePlot._fig.get_elem(m.name)},
@@ -1139,12 +1150,11 @@ def plot_forecast(
         # Draw original time series
         fig.add_trace(
             go.Scatter(
-                x=y_pred.index,
-                y=self.branch.dataset.loc[y_pred.index, target_c],
+                x=x,
+                y=self.branch._all.loc[y_pred.index, target_c],
                 mode="lines+markers",
                 line={"width": 1, "color": "black", "dash": "dash"},
                 opacity=0.6,
-                layer="below",
                 showlegend=False,
                 xaxis=xaxis,
                 yaxis=yaxis,

docs_sources/about.md

@@ -215,6 +215,13 @@ core project contributors with a set of developer tools free of charge.
       </a>
     </div>
   </div>
+  <div class="column">
+    <div class="logo">
+      <a href="../API/models/prophet" draggable="false">
+        <img src="../img/logos/prophet.png" alt="prophet" draggable="false">
+      </a>
+    </div>
+  </div>
   <div class="column">
     <div class="logo">
       <a href="../user_guide/accelerating/#gpu-acceleration" draggable="false">
@@ -229,15 +236,22 @@ core project contributors with a set of developer tools free of charge.
       </a>
     </div>
   </div>
+</div>
+<div class="row">
   <div class="column">
     <div class="logo">
       <a href="../user_guide/models" draggable="false">
         <img src="../img/logos/sklearn.png" alt="scikit-learn" draggable="false">
       </a>
     </div>
   </div>
-</div>
-<div class="row">
+  <div class="column">
+    <div class="logo">
+      <a href="../user_guide/time-series" draggable="false">
+        <img src="../img/logos/sktime.png" alt="sktime" draggable="false">
+      </a>
+    </div>
+  </div>
   <div class="column">
     <div class="logo">
       <a href="../API/models/xgb" draggable="false">

docs_sources/api/ATOM/atomforecaster.md

@@ -145,8 +145,8 @@ of utility methods to handle the data and manage the pipeline.
         - eda
         - evaluate
         - export_pipeline
-        - get_class_weight
         - get_sample_weight
+        - get_seasonal_period
         - inverse_transform
         - load
         - merge

docs_sources/api/ATOM/atomregressor.md

@@ -145,7 +145,6 @@ of utility methods to handle the data and manage the pipeline.
         - eda
         - evaluate
         - export_pipeline
-        - get_class_weight
         - get_sample_weight
         - inverse_transform
         - load

docs_sources/license.md

@@ -1,7 +1,7 @@
 # MIT License
 -------------
 
-Copyright © 2023 Mavs
+Copyright © 2019-2024 Mavs
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

docs_sources/user_guide/data_management.md

@@ -396,7 +396,12 @@ The check is performed in the order described hereunder:
 
 Additionally, the forecast horizon (parameter `fh`) in [forecasting tasks][time-series]
 can be selected much in the same way as `rows`, where the horizon is inferred
-as the index of the row selection.
+as the index of the row selection. Note that, contrary to sktime's API but for
+consistency with the rest of ATOM's API, atom's fh starts with the training set,
+i.e., selecting `#!python atom.nf.predict(fh=range(5))` forecasts the first 5
+rows of the training set, not the test set. To get the same result as sktime, use
+`#!python atom.nf.predict(fh=range(len(atom.test), len(atom.test) + 5))` or
+`#!python atom.nf.predict(fh=atom.test.index[:5])` instead.
 
 
 !!! info

docs_sources/user_guide/models.md

@@ -40,7 +40,7 @@ per task, but can include:
 - **fullname:** Name of the model's class.
 - **estimator:** Name of the model's underlying estimator.
 - **module:** The estimator's module.
-- **handles_missing:** Whether the model can handle missing (`NaN`) values
+- **handles_missing:** Whether the model can handle missing values
   without preprocessing. If False, consider using the [Imputer][] class
   before training the models.
 - **needs_scaling:** Whether the model requires feature scaling. If True,
@@ -54,6 +54,11 @@ per task, but can include:
 - **validation:** Whether the model has [in-training validation][].
 - **supports_engines:** [Engines][estimator-acceleration] supported by the model.
 
+To filter for specific tags, specify the column name with the desired value
+in the arguments of `available_models`, e.g., `#!python atom.available_models(accepts_sparse=True)`
+to get all models that accept sparse input or `#!python atom.available_models(supports_engines="cuml")`
+to get all models that support the [cuML][] engine.
+
 
 <br>