improver.percentile module#

Module containing percentiling classes.

class PercentileConverter(collapse_coord, percentiles=None, retained_coordinates=None, fast_percentile_method=True)[source]#

Plugin for converting from a set of values to a PDF.

Generate percentiles together with min, max, mean, stdev.

__init__(collapse_coord, percentiles=None, retained_coordinates=None, fast_percentile_method=True)[source]#

Create a PDF plugin with a given source plugin.

Parameters:

collapse_coord (Union[str, List[str]]) – The name of the coordinate(s) to collapse over. This coordinate(s) will no longer be present on the output cube, as it will have been replaced by the percentile coordinate.
percentiles (Optional[List[float]]) – Percentile values at which to calculate; if not provided uses DEFAULT_PERCENTILES. (optional)
retained_coordinates (Union[str, List[str], None]) – Optional list of collapsed coordinates that should be retained in their new scalar form. The default behaviour is to remove the scalar coordinates that result from coordinate collapse.
fast_percentile_method (bool) – If True use the numpy percentile method within Iris, which is much faster than scipy, but cannot handle masked data.

Raises:

TypeError – If collapse_coord is not a string.

process(cube)[source]#

Create a cube containing the percentiles as a new dimension.

What’s generated by default is:

15 percentiles - (0%, 5%, 10%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 90%, 95%, 100%)

Parameters:: cube (Cube) – Given the collapse coordinate, convert the set of values along that coordinate into a PDF and extract percentiles.
Return type:: Cube
Returns:: A single merged cube of all the cubes produced by each percentile collapse.

improver.percentile module