FeatureChunk#

class rojak.atmosphere.feature_composite.extension.FeatureChunk(centres, labels)[source]#

The centres of a feature found in one time chunk, as returned by a FeatureTabulator.

Parameters:

  • centres (pd.DataFrame)

  • labels (np.ndarray | None)

centres#

One row per centre, with the columns named by Column: CENTRE_ID (unique across all chunks, e.g. encoded from the time and grid indices), FEATURE_ID (equal to CENTRE_ID for a POINT feature; shared by the centres along one instance of a LINE feature), TIME_INDEX (along the whole time axis, see ChunkInfo.first_time_index), CENTRE_TIME, CENTRE_LAT and CENTRE_LON. A feature found on the grid also fills CENTRE_LAT_INDEX and CENTRE_LON_INDEX, and CENTRE_VALUE where the feature has a value at its centre.

Type:

pandas.DataFrame

labels#

Integer array with the shape (time, lat, lon) of the chunk’s blocks, holding the CENTRE_ID of the centre that owns each grid cell, or 0 where no centre does. None if the feature does not label regions, see FeatureSource.has_labels.

Type:

numpy.ndarray | None

__init__(centres, labels)[source]#
Parameters:

  • centres (pd.DataFrame)

  • labels (np.ndarray | None)

Return type:

None

Methods

__init__(centres, labels)

Attributes

centres

labels