parsnip.ParsnipModel.predict_dataset_augmented

ParsnipModel.predict_dataset_augmented(dataset, augments=10)

Generate predictions for a dataset with augmentation

This will first generate predictions for the dataset without augmentation, and will then generate predictions for the dataset with augmentation the given number of times. This returns a dataframe in the same format as predict_dataset, but with the following additional columns: - original_object_id: the original object_id for each augmentation. - augmented: True for augmented light curves, False for original ones.

Parameters:
  • dataset (Dataset) – Dataset to generate predictions for.

  • augments (int, optional) – Number of times to augment the dataset, by default 10

Returns:

predictions – astropy Table with one row for each light curve and columns with each of the predicted values.

Return type:

Table