a Sicã@s6dZddlZddlmZddgZd dd„Zdd„ZdS) z1 Determines if a contraction can use BLAS or not éNé)ÚhelpersÚcan_blasÚ tensor_blasc CsÎt|ƒdkrdS|\}}t||ƒD]Z}| |¡| |¡}}|dksZ|dksZ||dkr`dS||dt||vƒkr$dSq$|dur¼|D].}|d| |¡|d| |¡krŒdSqŒt|ƒdkrÌdSdd„|Dƒ} | d|} | d|} t|ƒ} |d|dkrd S| d| dkr&d S|| d…|d| …krFd S|d| …|| d…krfd S|| d…|| d…krˆd S|d| …|d| …kr¦d St| ƒdksÂt| ƒdkrÆd Sd SdS)aŠ Checks if we can use a BLAS call. Parameters ---------- inputs : list of str Specifies the subscripts for summation. result : str Resulting summation. idx_removed : set Indices that are removed in the summation shapes : sequence of tuple[int], optional If given, check also that none of the indices are broadcast dimensions. Returns ------- type : str or bool The type of BLAS call to be used or False if none. Notes ----- We assume several operations are not efficient such as a transposed DDOT, therefore 'ijk,jki->' should prefer einsum. These return the blas type appended with "/EINSUM" to differentiate when they can still be done with tensordot if required, e.g. when a backend has no einsum. Examples -------- >>> can_blas(['ij', 'jk'], 'ik', set('j')) 'GEMM' >>> can_blas(['ijj', 'jk'], 'ik', set('j')) False >>> can_blas(['ab', 'cd'], 'abcd', set()) 'OUTER/EINSUM' >>> # looks like GEMM but actually 'j' is broadcast: >>> can_blas(['ij', 'jk'], 'ik', set('j'), shapes=[(4, 1), (5, 6)]) False éFrNrz OUTER/EINSUMcSsg|] }t|ƒ‘qS©)Úset©Ú.0ÚxrrúK/var/www/html/django/DPS/env/lib/python3.9/site-packages/opt_einsum/blas.pyÚ Uózcan_blas..ÚDOTz DOT/EINSUMZGEMMz GEMV/EINSUMZTDOT)ÚlenrÚcountÚintÚfind) ÚinputsÚresultÚ idx_removedÚshapesÚ input_leftÚ input_rightÚcÚnlÚnrÚsetsÚ keep_leftÚ keep_rightÚrsrrr r sD+ $   csft|ƒ}t|ƒ|}t|ƒ|}i‰t||jƒD]\}} | ˆ|<q0t||jƒD]\}} | ˆ|<qNt|ƒ} t |ˆ¡} t |ˆ¡} t |ˆ¡} ||}|D]} | | d¡}q˜||krÊt |  ¡|  ¡¡}n8|| d…|d| …krt |  | | ¡|  | | ¡¡}nþ|d| …|| d…krBt |  | | ¡j |  | | ¡j ¡}nÀ|| d…|| d…kr€t |  | | ¡|  | | ¡j ¡}n‚|d| …|d| …krºt |  | | ¡j |  | | ¡¡}nHd\}}|D]&} ||  | ¡f7}||  | ¡f7}qÆtj ||||fd}t‡fdd„|Dƒƒ}|j|krDt|ƒdkr:||_n t |¡}||krbt |d||¡}|S) a Computes the dot product between two tensors, attempts to use np.dot and then tensordot if that fails. Parameters ---------- view_left : array_like The left hand view input_left : str Indices of the left view view_right : array_like The right hand view input_right : str Indices of the right view index_result : str The resulting indices idx_removed : set Indices removed in the contraction Returns ------- type : array The resulting BLAS operation. Notes ----- Interior function for tensor BLAS. This function will attempt to use `np.dot` by the iterating through the four possible transpose cases. If this fails all inner and matrix-vector operations will be handed off to einsum while all matrix-matrix operations will first copy the data, perform the DGEMM, and then copy the data to the required order. Examples -------- >>> a = np.random.rand(4, 4) >>> b = np.random.rand(4, 4) >>> tmp = tensor_blas(a, 'ij', b, 'jk', 'ik', set('j')) >>> np.allclose(tmp, np.dot(a, b)) ÚN)rr)Úaxesc3s|]}ˆ|VqdS)Nrr ©Údimension_dictrr Ú érztensor_blas..rz->)rÚzipÚshaperrZcompute_size_by_dictÚreplaceÚnpÚdotÚravelÚreshapeÚTrÚ tensordotÚtupleÚsqueezeÚeinsum)Z view_leftrZ view_rightrZ index_resultrrrÚiÚsr Zdim_leftZ dim_rightZ dim_removedÚ tensor_resultÚnew_viewÚleft_posÚ right_posÚ tensor_shaperr#r r{sL-       "     )N)Ú__doc__Únumpyr)r!rÚ__all__rrrrrr Ús   o