eugene.evaluate.train_val_predictions

eugene.evaluate.train_val_predictions(model, train_dataloader, val_dataloader, train_var='train_val', gpus=None, out_dir=None, name=None, version='')

Predictions from a model and train/val dataloaders.

Same as predictions function, but takes separate train and val dataloaders.

Parameters:

• model (LightningModule) – Model to predict with.

• train_dataloader (DataLoader) – Dataloader to predict with.

• val_dataloader (DataLoader) – Dataloader to predict with.

• train_var (str, optional) – Key in sdata to use as the train/val variable. If None, uses “train_val”.

• gpus (int, optional) – Number of GPUs to use. If None, uses settings.gpus.

• out_dir (os.PathLike, optional) – Directory to write predictions to. If None, does not write predictions to disk.

• name (str, optional) – Name of the model. If None, uses model.model_name.

• version (str, optional) – Version of the model. If None, uses “”.

Returns:

preds – Predictions from the model and dataloader in a pandas dataframe.

Return type:

pd.DataFrame