Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

support prediction intervals in auto #370

Merged
merged 1 commit into from
Jul 3, 2024
Merged

support prediction intervals in auto #370

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 12 additions & 2 deletions mlforecast/auto.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,7 @@
LocalStandardScaler,
GlobalSklearnTransformer,
)
from .utils import PredictionIntervals

# %% ../nbs/auto.ipynb 4
def lightgbm_space(trial: optuna.Trial):
Expand Down Expand Up @@ -442,6 +443,7 @@ def fit(
study_kwargs: Optional[Dict[str, Any]] = None,
optimize_kwargs: Optional[Dict[str, Any]] = None,
fitted: bool = False,
prediction_intervals: Optional[PredictionIntervals] = None,
) -> "AutoMLForecast":
"""Carry out the optimization process.
Each model is optimized independently and the best one is trained on all data
Expand Down Expand Up @@ -475,6 +477,8 @@ def fit(
Keyword arguments to be passed to the optuna.Study.optimize method.
fitted : bool (default=False)
Whether to compute the fitted values when retraining the best model.
prediction_intervals :
Configuration to calibrate prediction intervals when retraining the best model.

Returns
-------
Expand Down Expand Up @@ -538,6 +542,7 @@ def config_fn(trial: optuna.Trial) -> Dict[str, Any]:
self.results_[name] = study
best_config = study.best_trial.user_attrs["config"]
best_config["mlf_fit_params"].pop("fitted", None)
best_config["mlf_fit_params"].pop("prediction_intervals", None)
best_model = clone(auto_model.model)
best_model.set_params(**best_config["model_params"])
self.models_[name] = MLForecast(
Expand All @@ -548,6 +553,7 @@ def config_fn(trial: optuna.Trial) -> Dict[str, Any]:
self.models_[name].fit(
df,
fitted=fitted,
prediction_intervals=prediction_intervals,
**best_config["mlf_fit_params"],
)
return self
Expand All @@ -556,6 +562,7 @@ def predict(
self,
h: int,
X_df: Optional[DataFrame] = None,
level: Optional[List[Union[int, float]]] = None,
) -> DataFrame:
""" "Compute forecasts

Expand All @@ -565,6 +572,8 @@ def predict(
Number of periods to predict.
X_df : pandas or polars DataFrame, optional (default=None)
Dataframe with the future exogenous features. Should have the id column and the time column.
level : list of ints or floats, optional (default=None)
Confidence levels between 0 and 100 for prediction intervals.

Returns
-------
Expand All @@ -573,11 +582,12 @@ def predict(
"""
all_preds = None
for name, model in self.models_.items():
preds = model.predict(h=h, X_df=X_df)
preds = model.predict(h=h, X_df=X_df, level=level)
if all_preds is None:
all_preds = preds
else:
all_preds = ufp.assign_columns(all_preds, name, preds[name])
model_cols = [c for c in preds.columns if c not in all_preds.columns]
all_preds = ufp.horizontal_concat([all_preds, preds[model_cols]])
return all_preds

def save(self, path: Union[str, Path]) -> None:
Expand Down
Loading
Loading