tso tawm Python tsev qiv ntawv rau scientific xam , tsom mus rau kev ua hauj lwm nrog ntau qhov sib txawv arrays thiab matrices, thiab tseem muab kev sau loj ntawm kev ua haujlwm nrog kev siv ntau yam algorithms ntsig txog kev siv matrices. NumPy yog ib lub tsev qiv ntawv nrov tshaj plaws siv rau kev suav lej. Txoj haujlwm code yog sau hauv Python siv optimizations hauv C thiab raws li BSD daim ntawv tso cai.
NumPy 1.19 tsis txhawb nqa Python 3.5 ntxiv lawm thiab tshem tawm cov lej los txhawb Python 2 (cov txheej numpy.compat tau tso rau hauv qhov chaw tam sim no). Cov kev txhawb nqa yog Python 3.6, 3.7 thiab 3.8. Module kev txhim kho txuas ntxiv mus rau kev ua haujlwm nrog cov qauv random. Txhim kho kev txhawb nqa rau NumPy log pob ntawm Aarch64 architecture thiab thaum siv Python kev siv . functionality numpy.frompyfunc, np.str_, numpy.copy, numpy.linalg.multi_dot, numpy.count_nonzero thiab numpy.array_equal. Txhim kho kev tshawb pom ntawm CPU muaj peev xwm xws li kev txhawb nqa AVX. Ntxiv kev siv uas ua haujlwm 5-7 zaug sai dua raws li AVX512, siv rau cov ntaub ntawv tawm tswv yim np.float64.
Tau qhov twg los: opennet.ru