unspecify methods

source/elements/oneMKL/source/domains/rng/device_api/device-distributions-template-parameter-method.rst

@@ -48,80 +48,5 @@ Distributions Template Parameter Method
         * Triangle
         * Left exponential tail
         * Right exponential tail
-   * -  ``beta_method::cja``
-        ``beta_method::cja_accurate``
-     -  ``beta``
-     -  Acceptance/rejection method depending on the parameters of the distribution, ``p`` and ``q``:
-
-        * If ``max(p,q) < 1``, beta distributed variates are obtained by means of
-          composition of two algorithms (below constants K and C are such that
-          ``K~=0.852``, ``C~=-0.956`` ):
-
-          * if ``q+K*p^2+C<=0``, method of Johnk is used (see References 1. for details);
-          * if ``q+K*p^2+C>0``, switching method of Atkinson is used (see
-            References 2. for details). In the method of Atkinson interval ``(0,1)``
-            is divided into two parts, ``(0,t)`` and ``(t,1)``, where
-            ``t=sqrt(p*(1-p))/(sqrt(p*(1-p))+sqrt(q*(1-q)))``. Then the
-            acceptance/rejection technique is applied on each of these
-            intervals.
-        * If ``p < 1`` and ``q > 1``, switching algorithm of Atkinson with
-          ``t=(1-p)/(q+1-p)`` is used;
-        * If ``p < 1`` and ``q > 1``, switching algorithm of Atkinson with
-          interchanging parameters, ``p`` and ``q`` and ``t=(1-p)/(q+1-p)`` is used.
-          After generation ``x`` using the algorithm random number ``1 - x`` is returned
-          as required result.
-        * If ``min(p,q) > 1``, algorithm of Cheng is used. This method based on
-          acceptance/rejection technique operates by obtaining beta variates
-          of the second kind and making their transformation to beta variates
-          of the first kind (see References 3. for details);
-        * If ``p = 1`` and ``q`` is arbitrary positive number inverse cumulative
-          distribution function (ICDF) method is used;
-        * If ``q = 1`` and ``p`` is arbitrary positive number inverse cumulative
-          distribution function (ICDF) method is used;
-        * If ``p = 1`` and ``q = 1`` beta distribution reduces to uniform distribution
-          over ``(0,1)`` interval.
-
-   * -  ``gamma_method::marsaglia``
-        ``gamma_method::marsaglia_accurate``
-     -  ``gamma``
-     -  Acceptance/rejection method depending on the :math:`\alpha` parameter of the distribution:
-
-        * If :math:`\alpha > 1` method of Marsaglia and Tsang is used.
-          This algorithm is based on the acceptance/rejection
-          method using squeeze technique (see References 4. for details). Gaussian random
-          numbers are generated by means of the Box-Muller 2 method. Due to
-          nature of the Box-Muller algorithm (pair of uniformly distributed
-          random number is transformed into pair of standard normal numbers) to
-          calculate ``n`` gamma variates one always generates even number of normal
-          and uniformly distributed random numbers. If the last pair
-          of gaussian and uniformly distributed random numbers is not used it is
-          saved for later usage.
-        * If :math:`\alpha = 1` gamma distribution reduces to exponential distribution
-          with parameters ``(0,1)``.
-        * If :math:`\alpha \geq 0.6` and :math:`\alpha < 1` modified method of Vaduva is used.
-          Gamma distributed random number is generated using rejection from
-          the Weibull distribution (see References 5. and 6. for details).
-        * If :math:`\alpha < 0.6` gamma distributed random number is got by transformation
-          of random variate with exponential power distribution (EPD). EPD random
-          number is obtained using acceptance/rejection method. The algorithm for
-          generation of gamma variate with parameter :math:`\alpha < 1` is a specific case of
-          acceptance/rejection method used for generation of EPD random number.
-          (see References 6. for details).
-
-References
-==========
-
-   #. M.D. Johnk, Erzeugung von Betaverteilten und Gammaverteilten Zufallszahlen, Metrika, 8, 5-15, 1964
-   #. A.C. Atkinson, A family of switching algorithms for the computer
-      generation of beta random variables. Biometrika, 66, 1, 141-145, 1979
-   #. Cheng R.C.H. Generating Beta variates with Nonintegral Shape
-      Parameters, Communications of the ACM, 21, 4, 317-322, 1978.
-   #. G. Marsaglia, and W.W. Tsang, A simple method for generating gamma
-      variables. ACM Transactions on Mathematical Software, vol. 26, No. 3, 363-372, September 2000.
-   #. I. Vaduva, On computer generation of gamma random variables by
-      rejection and composition procedures. Mathematische Operationsforschung
-      und Statistik, Series Statistics, vol. 8, 545-576, 1977.
-   #. L. Devroye, Non-Uniform Random Variate Generation, Springer-Verlag, New York, 1986.
-
 
 `NOTE:` Methods provided for exposition purposes.

source/elements/oneMKL/source/domains/rng/device_api/device-rng-beta.rst

@@ -72,14 +72,8 @@ class beta
 
     .. container:: section
 
-        typename Method = oneapi::mkl::rng::beta_method::by_default
-            Generation method. The specific values are as follows:
-
-                * ``oneapi::mkl::rng::device::beta_method::by_default``
-                * ``oneapi::mkl::rng::device::beta_method::cja``
-                * ``oneapi::mkl::rng::device::beta_method::cja_accurate``
-
-            See description of the methods in :ref:`Distributions methods template parameter<onemkl_device_rng_distributions_method>`.
+        typename Method
+            Generation method. The type is unspecified.
 
 
 .. container:: section

source/elements/oneMKL/source/domains/rng/device_api/device-rng-gamma.rst

@@ -69,14 +69,8 @@ class gamma
 
     .. container:: section
 
-        typename Method = oneapi::mkl::rng::gamma_method::by_default
-            Generation method. The specific values are as follows:
-
-                * ``oneapi::mkl::rng::device::gamma_method::by_default``
-                * ``oneapi::mkl::rng::device::gamma_method::marsaglia``
-                * ``oneapi::mkl::rng::device::gamma_method::marsaglia_accurate``
-
-            See description of the methods in :ref:`Distributions methods template parameter<onemkl_device_rng_distributions_method>`.
+        typename Method
+            Generation method. The type is unspecified.
 
 
 .. container:: section