raptorstats.stats module

class raptorstats.stats.Stats(stats=None, categorical=False)

Bases: object

Configuration for statistics computation.

Stores the statistics to be computed and has methods for computing them from arrays or partial results.

clean_results(results)

empty()

Return an empty stats dict with the requested keys.

from_array(data: MaskedArray) → dict

Compute the configured statistics on data.

Parameters:

data (np.ndarray or np.ma.MaskedArray) – Pixel values for one feature. Masked or NaN values are treated as nodata.

Returns:

Keys are the statistic names requested at construction time (plus every {value: count} entry if categorical=True). Values are floats or ints; empty inputs yield None.

Return type:

dict

from_partials(partials: list[dict]) → dict

Combine already-computed chunk statistics into one feature-level dict.

Each element of partials must have been produced by from_array() with the same Stats configuration.

Returns:

Feature-level statistics – exactly the same schema you get from calling from_array() on the full dataset at once.

Return type:

dict