DŽgM*vddlZddlZddlmZddlmZddlmZmZddl m Z dgZ dZ dZ d ZGd deZy) N) constraints) Distribution)_standard_normal lazy_property)_sizeMultivariateNormalcjtj||jdjdS)a Performs a batched matrix-vector product, with compatible but different batch shapes. This function takes as input `bmat`, containing :math:`n \times n` matrices, and `bvec`, containing length :math:`n` vectors. Both `bmat` and `bvec` may have any number of leading dimensions, which correspond to a batch shape. They are not necessarily assumed to have the same batch shape, just ones which can be broadcasted. )torchmatmul unsqueezesqueeze)bmatbvecs o/home/mcse/projects/flask_80/flask-venv/lib/python3.12/site-packages/torch/distributions/multivariate_normal.py _batch_mvrs) <<dnnR0 1 9 9" ==c$|jd}|jdd}t|}|jdz }||z }||z}|d|zz}|jd|} t |jdd|j|dD]\} } | | | z| fz } | |fz } |j | }t t|t t||dzt t|dz|dz|gz} |j| }|j d||} |j d| jd|}|jddd}tjj| |djdjd}|j}|j |jdd}t t|}t|D]}|||z||zgz }|j|}|j |S) aK Computes the squared Mahalanobis distance :math:`\mathbf{x}^\top\mathbf{M}^{-1}\mathbf{x}` for a factored :math:`\mathbf{M} = \mathbf{L}\mathbf{L}^\top`. Accepts batches for both bL and bx. They are not necessarily assumed to have the same batch shape, but `bL` one should be able to broadcasted to `bx` one. r NrFupper)sizeshapelendimzipreshapelistrangepermuter linalgsolve_triangularpowsumt)bLbxnbx_batch_shape bx_batch_dims bL_batch_dimsouter_batch_dimsold_batch_dimsnew_batch_dims bx_new_shapesLsx permute_dimsflat_Lflat_x flat_x_swapM_swapM permuted_Mpermute_inv_dimsi reshaped_Ms r_batch_mahalanobisr>s0  AXXcr]N'MFFHqLM$}4% 5N%M(99N88--.LbhhsmRXX.>r%BC'Br2& 'QDL L !B U# $% u%~q9 : ; u%)>1= > ?    L !B ZZAq !F ZZFKKNA .F..Aq)K %%fk%GKKANRRSUV   A288CR=)JE"234 = !G-1>A3EFFG##$45J   n --rcxtjjtj|d}tjtj|ddd}tj |j d|j|j}tjj||d}|S)N)rr rr dtypedeviceFr) r r#choleskyflip transposeeyerrArBr$)PLfL_invIdLs r_precision_to_scale_trilrLNs   uzz!X6 7B OOEJJr84b" =E 1772;aggahh ?B %%eRu%=A HrcdeZdZdZej ej ej ejdZej Z dZ dfd Z dfd Z e dZe dZe dZed Zed Zed Zej,fd ed ej0fdZdZdZxZS)ra Creates a multivariate normal (also called Gaussian) distribution parameterized by a mean vector and a covariance matrix. The multivariate normal distribution can be parameterized either in terms of a positive definite covariance matrix :math:`\mathbf{\Sigma}` or a positive definite precision matrix :math:`\mathbf{\Sigma}^{-1}` or a lower-triangular matrix :math:`\mathbf{L}` with positive-valued diagonal entries, such that :math:`\mathbf{\Sigma} = \mathbf{L}\mathbf{L}^\top`. This triangular matrix can be obtained via e.g. Cholesky decomposition of the covariance. Example: >>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_LAPACK) >>> # xdoctest: +IGNORE_WANT("non-deterministic") >>> m = MultivariateNormal(torch.zeros(2), torch.eye(2)) >>> m.sample() # normally distributed with mean=`[0,0]` and covariance_matrix=`I` tensor([-0.2102, -0.5429]) Args: loc (Tensor): mean of the distribution covariance_matrix (Tensor): positive-definite covariance matrix precision_matrix (Tensor): positive-definite precision matrix scale_tril (Tensor): lower-triangular factor of covariance, with positive-valued diagonal Note: Only one of :attr:`covariance_matrix` or :attr:`precision_matrix` or :attr:`scale_tril` can be specified. Using :attr:`scale_tril` will be more efficient: all computations internally are based on :attr:`scale_tril`. If :attr:`covariance_matrix` or :attr:`precision_matrix` is passed instead, it is only used to compute the corresponding lower triangular matrices using a Cholesky decomposition. )loccovariance_matrixprecision_matrix scale_trilTc|jdkr td|du|duz|duzdk7r td|h|jdkr tdtj|jdd|jdd}|j |dz|_n|h|jdkr td tj|jdd|jdd}|j |dz|_ng|jdkr td tj|jdd|jdd}|j |dz|_|j |d z|_ |jjdd}t|-||| |||_ y|%tjj||_ yt||_ y) Nrz%loc must be at least one-dimensional.zTExactly one of covariance_matrix or precision_matrix or scale_tril may be specified.rzZscale_tril matrix must be at least two-dimensional, with optional leading batch dimensionsrr )r r zZcovariance_matrix must be at least two-dimensional, with optional leading batch dimensionszYprecision_matrix must be at least two-dimensional, with optional leading batch dimensions)r  validate_args)r ValueErrorr broadcast_shapesrexpandrQrOrPrNsuper__init___unbroadcasted_scale_trilr#rCrL) selfrNrOrPrQrT batch_shape event_shape __class__s rrYzMultivariateNormal.__init__s 779q=DE E T )j.D E D (  f   !~~!# = 001A1A#21F RUSUWK(// h0FGDO  * $$&* = 00!'',ciinK&7%=%=kH>T%UD "##%) = 00 &&s+SYYs^K%5$;$;K(CN  .#'#8#8#?#? #JC   #/ )) 0 "00 rc|jj|j|jz|jzSN)rZrW _batch_shape _event_shaper[s rrQzMultivariateNormal.scale_trils:--44    1 1 1D4E4E E  rctj|j|jjj |j |j z|j zSri)r r rZmTrWrjrkrls rrOz$MultivariateNormal.covariance_matrixsQ||  * *D,J,J,M,M &""T%6%669J9JJ K Lrctj|jj|j|j z|j zSri)r cholesky_inverserZrWrjrkrls rrPz#MultivariateNormal.precision_matrixsE%%d&D&DELL    1 1 1D4E4E E  rc|jSrirNrls rmeanzMultivariateNormal.mean xxrc|jSrirrrls rmodezMultivariateNormal.modertrc|jjdjdj|j|j zS)Nrr )rZr%r&rWrjrkrls rvariancezMultivariateNormal.variancesA  * * . .q 1 SW VD%%(9(99 : r sample_shapereturnc|j|}t||jj|jj}|jt |j |zS)Nr@)_extended_shaperrNrArBrrZ)r[ryrepss rrsamplezMultivariateNormal.rsamplesL$$\2uDHHNN488??Sxx)D$B$BCHHHrcx|jr|j|||jz }t|j|}|jj ddj jd}d|jdtj dtjzz|zz|z S)Nrr dim1dim2grr) rc_validate_samplerNr>rZdiagonallogr&rkmathpi)r[valuediffr9 half_log_dets rlog_probzMultivariateNormal.log_probs     ! !% (txx t==t D  * * 3 3" 3 E I I K O OPR S t((+dhhq477{.CCaGH<WWrc^|jjddjjd}d|jdzdt jdt j zzz|z}t|jdk(r|S|j|jS)Nrr rg?rg?r) rZrrr&rkrrrrjrW)r[rHs rentropyzMultivariateNormal.entropys  * * 3 3" 3 E I I K O OPR S  $##A& &#TWW0E*E F U t  !Q &H88D--. .r)NNNNri)__name__ __module__ __qualname____doc__r real_vectorpositive_definitelower_choleskyarg_constraintssupport has_rsamplerYrWrrQrOrPpropertyrsrvrxr rarTensorr~rr __classcell__)r^s@rrrWs"H&&(::'99!00 O %%GK  7Xr&  LL     -7EJJLIEIU\\I X/r)rr torch.distributionsr torch.distributions.distributionrtorch.distributions.utilsrr torch.typesr__all__rr>rLrrrrs? +9E  >/.d r/r/r