Source code for featuretools.primitives.standard.aggregation.std

import numpy as np
from woodwork.column_schema import ColumnSchema

from featuretools.primitives.base.aggregation_primitive_base import AggregationPrimitive
from featuretools.utils.gen_utils import Library


[docs]class Std(AggregationPrimitive): """Computes the dispersion relative to the mean value, ignoring `NaN`. Examples: >>> std = Std() >>> round(std([1, 2, 3, 4, 5, None]), 3) 1.414 """ name = "std" input_types = [ColumnSchema(semantic_tags={"numeric"})] return_type = ColumnSchema(semantic_tags={"numeric"}) stack_on_self = False compatibility = [Library.PANDAS, Library.DASK, Library.SPARK] description_template = "the standard deviation of {}" def get_function(self, agg_type=Library.PANDAS): if agg_type in [Library.DASK, Library.SPARK]: return "std" return np.std