Skip to content

Commit 8efaff4

Browse files
vlad-perevezentsevvlad-perevezentsev
committed
Require positive scale parameters and use the exported method tokens
1 parent 0f8bad3 commit 8efaff4

1 file changed

Lines changed: 21 additions & 26 deletions

File tree

‎mkl_random/mklrand.pyx‎

Lines changed: 21 additions & 26 deletions
Original file line numberDiff line numberDiff line change
@@ -3236,9 +3236,8 @@ cdef class _MKLRandomState:
32363236
single value is returned.
32373237
method : {'ICDF', 'BoxMuller', 'BoxMuller2'}, optional
32383238
Sampling method used by Intel MKL. Can also be specified using
3239-
the tokens ``mkl_random.mklrand.ICDF``,
3240-
``mkl_random.mklrand.BOXMULLER``,
3241-
``mkl_random.mklrand.BOXMULLER2``.
3239+
the tokens ``mkl_random.ICDF``, ``mkl_random.BOXMULLER``,
3240+
``mkl_random.BOXMULLER2``.
32423241
32433242
Returns
32443243
-------
@@ -3305,16 +3304,15 @@ cdef class _MKLRandomState:
33053304
Mean ("centre") of the distribution.
33063305
scale : float or array_like of floats
33073306
Standard deviation (spread or "width") of the distribution.
3308-
Must be non-negative.
3307+
Must be positive.
33093308
size : int or tuple of ints, optional
33103309
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
33113310
``m * n * k`` samples are drawn. Default is None, in which case a
33123311
single value is returned.
33133312
method : {'ICDF', 'BoxMuller', 'BoxMuller2'}, optional
33143313
Sampling method used by Intel MKL. Can also be specified using
3315-
the tokens ``mkl_random.mklrand.ICDF``,
3316-
``mkl_random.mklrand.BOXMULLER``,
3317-
``mkl_random.mklrand.BOXMULLER2``.
3314+
the tokens ``mkl_random.ICDF``, ``mkl_random.BOXMULLER``,
3315+
``mkl_random.BOXMULLER2``.
33183316
33193317
Returns
33203318
-------
@@ -3479,9 +3477,9 @@ cdef class _MKLRandomState:
34793477
Parameters
34803478
----------
34813479
a : float or array_like of floats
3482-
Alpha, non-negative.
3480+
Alpha, must be positive.
34833481
b : float or array_like of floats
3484-
Beta, non-negative.
3482+
Beta, must be positive.
34853483
size : int or tuple of ints, optional
34863484
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
34873485
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -3550,7 +3548,7 @@ cdef class _MKLRandomState:
35503548
----------
35513549
scale : float or array_like of floats
35523550
The scale parameter, :math:`\\beta = 1/\\lambda`.
3553-
Must be non-negative.
3551+
Must be positive.
35543552
size : int or tuple of ints, optional
35553553
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
35563554
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -3787,9 +3785,9 @@ cdef class _MKLRandomState:
37873785
----------
37883786
shape : float or array_like of floats
37893787
The shape of the gamma distribution.
3790-
Must be non-negative.
3788+
Must be positive.
37913789
scale : float or array_like of floats, optional
3792-
The scale of the gamma distribution. Must be non-negative.
3790+
The scale of the gamma distribution. Must be positive.
37933791
Default is equal to 1.
37943792
size : int or tuple of ints, optional
37953793
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
@@ -4738,7 +4736,7 @@ cdef class _MKLRandomState:
47384736
Parameters
47394737
----------
47404738
a : float or array_like of floats
4741-
Shape parameter of the distribution. Must be nonnegative.
4739+
Shape parameter of the distribution. Must be positive.
47424740
size : int or tuple of ints, optional
47434741
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
47444742
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -4967,7 +4965,7 @@ cdef class _MKLRandomState:
49674965
Default is 0.
49684966
scale : float or array_like of floats, optional
49694967
:math:`\\lambda`, the exponential decay.
4970-
Default is 1. Must be non-negative.
4968+
Default is 1. Must be positive.
49714969
size : int or tuple of ints, optional
49724970
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
49734971
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -5074,7 +5072,7 @@ cdef class _MKLRandomState:
50745072
The location of the mode of the distribution. Default is 0.
50755073
scale : float or array_like of floats, optional
50765074
The scale parameter of the distribution. Default is 1.
5077-
Must be non-negative.
5075+
Must be positive.
50785076
size : int or tuple of ints, optional
50795077
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
50805078
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -5214,7 +5212,7 @@ cdef class _MKLRandomState:
52145212
loc : float or array_like of floats, optional
52155213
Parameter of the distribution. Default is 0.
52165214
scale : float or array_like of floats, optional
5217-
Parameter of the distribution. Must be non-negative.
5215+
Parameter of the distribution. Must be positive.
52185216
Default is 1.
52195217
size : int or tuple of ints, optional
52205218
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
@@ -5324,15 +5322,14 @@ cdef class _MKLRandomState:
53245322
Mean value of the underlying normal distribution. Default is 0.
53255323
sigma : float or array_like of floats, optional
53265324
Standard deviation of the underlying normal distribution.
5327-
Must be non-negative. Default is 1.
5325+
Must be positive. Default is 1.
53285326
size : int or tuple of ints, optional
53295327
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
53305328
``m * n * k`` samples are drawn. Default is None, in which case a
53315329
single value is returned.
53325330
method : {'ICDF', 'BoxMuller'}, optional
53335331
Sampling method used by Intel MKL. Can also be specified using
5334-
the tokens ``mkl_random.mklrand.ICDF``,
5335-
``mkl_random.mklrand.BOXMULLER``.
5332+
the tokens ``mkl_random.ICDF``, ``mkl_random.BOXMULLER``.
53365333
53375334
Returns
53385335
-------
@@ -5491,7 +5488,7 @@ cdef class _MKLRandomState:
54915488
Parameters
54925489
----------
54935490
scale : float or array_like of floats, optional
5494-
Scale, also equals the mode. Should be >= 0.
5491+
Scale, also equals the mode. Must be positive.
54955492
size : int or tuple of ints, optional
54965493
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
54975494
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -5583,7 +5580,7 @@ cdef class _MKLRandomState:
55835580
mean : float or array_like of floats
55845581
Distribution mean, should be > 0.
55855582
scale : float or array_like of floats
5586-
Scale parameter, should be >= 0.
5583+
Scale parameter, must be positive.
55875584
size : int or tuple of ints, optional
55885585
Output shape. If the given shape is, e.g., ``(m, n, k)``, then
55895586
``m * n * k`` samples are drawn. Default is None, in which case a
@@ -6038,8 +6035,7 @@ cdef class _MKLRandomState:
60386035
single value is returned.
60396036
method : {'POISNORM', 'PTPE'}, optional
60406037
Sampling method used by Intel MKL. Can also be specified using
6041-
the tokens ``mkl_random.mklrand.POISNORM``,
6042-
``mkl_random.mklrand.PTPE``.
6038+
the tokens ``mkl_random.POISNORM``, ``mkl_random.PTPE``.
60436039
60446040
Returns
60456041
-------
@@ -7422,9 +7418,8 @@ cdef class MKLRandomState(_MKLRandomState):
74227418
If no shape is specified, a single (`N`-D) sample is returned.
74237419
method : {'ICDF', 'BoxMuller', 'BoxMuller2'}, optional
74247420
Sampling method used by Intel MKL. Can also be specified using
7425-
the tokens ``mkl_random.mklrand.ICDF``,
7426-
``mkl_random.mklrand.BOXMULLER``,
7427-
``mkl_random.mklrand.BOXMULLER2``.
7421+
the tokens ``mkl_random.ICDF``, ``mkl_random.BOXMULLER``,
7422+
``mkl_random.BOXMULLER2``.
74287423
74297424
Returns
74307425
-------

0 commit comments

Comments
 (0)