hyrax.datasets.collate_utils#
Functions#
|
Collate the given field in the samples as if it were a light curve |
Module Contents#
- collate_as_1d_light_curve(samples: list[dict], field: str) dict[source]#
Collate the given field in the samples as if it were a light curve
- Parameters:
samples – List of dicts; each dict is expected to have the key passed in for the field argument
field – The field to collate
- Returns:
Contains three keys: <field>, <field>_lengths, and <field>_mask field - float32 array (batch, max_len) containing the padded light curves <field>_lengths - int64 array (batch) of true light curve lengths <field>_mask - int64 array (batch, max_len) of masks denoting light-curve data vs. padding
- Return type:
dict