You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Introduce a new function (e.g. l_moments or just moments) that accepts max_order: int, / and kwargs first_order: int = 1 and ratio: bool = True, always returns an xp.Array (np.ndarray by default) of shape (max_order - first_order + 1,) and real floating dtype (dtype: type[xp.floating] = xp.float64), where ratio=True is like l_stats and ratio=False like l_moment.
Deprecate l_moment, l_ratio and l_stats on lmo<1, and remove them in the v1.0 release.
The current l_{loc,variation,scale,skew,kurt[osis]} should be modified to autonomically calculate these statistics, preferably as an array-api compatible gufunc-like (potentially delegating to a numba gufunc).
The same can be done with theoretical ones in lmo.theoretical, and similarly in lmo.diagnostic.
The text was updated successfully, but these errors were encountered:
Introduce a new function (e.g.
l_moments
or justmoments
) that acceptsmax_order: int, /
and kwargsfirst_order: int = 1
andratio: bool = True
, always returns anxp.Array
(np.ndarray
by default) of shape(max_order - first_order + 1,)
andreal floating
dtype (dtype: type[xp.floating] = xp.float64
), whereratio=True
is likel_stats
andratio=False
likel_moment
.Deprecate
l_moment
,l_ratio
andl_stats
onlmo<1
, and remove them in the v1.0 release.The current
l_{loc,variation,scale,skew,kurt[osis]}
should be modified to autonomically calculate these statistics, preferably as an array-api compatible gufunc-like (potentially delegating to a numba gufunc).The same can be done with theoretical ones in
lmo.theoretical
, and similarly inlmo.diagnostic
.The text was updated successfully, but these errors were encountered: