arch.univariate.distribution.Distribution.starting_values¶

abstract Distribution.starting_values(std_resid: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]][source]¶

Construct starting values for use in optimization.

Parameters:¶

std_resid: ndarray[Any, dtype[float64]]¶

Estimated standardized residuals to use in computing starting values for the shape parameter

Returns:¶

sv – The estimated shape parameters for the distribution

Return type:¶

numpy.ndarray

Notes

Size of sv depends on the distribution