Eminyakeni yamuva nje, ukusetshenziswa kwePython emikhakheni ehlukahlukene kuye kwanda kakhulu, ikakhulukazi emkhakheni wokukhohlisa idatha kanye nekhompyutha yesayensi. Omunye wemitapo yolwazi esetshenziswa kakhulu kule misebenzi iNumPy. I-NumPy iwumtapo wolwazi onamandla futhi oguquguqukayo osetshenziswa kakhulu ekusebenzeni ngamaqoqo amakhulu, ama-multidimensional kanye namatrices, phakathi kweminye imisebenzi yezibalo. Umsebenzi owodwa ovamile ekusebenzeni nalezi zakhiwo zedatha isidingo sokugoqa noma ukwehlisa ubukhulu bokugcina belungu elifanayo. Kulesi sihloko, sizohlola lesi sihloko ngokuningiliziwe, siqale ngesingeniso senkinga, silandelwe yisixazululo, kanye nencazelo yesinyathelo ngesinyathelo sekhodi. Okokugcina, sizocubungula ezinye izihloko ezihlobene namalabhulali okungenzeka kube nentshisekelo kuwo.
Isidingo soku bhidliza ubukhulu bokugcina yamalungu afanayo angavela ezimeni ezihlukahlukene, njengalapho ubale umphumela ovela kulungu elifanayo elinezinhlangothi eziningi futhi ufuna ukuthola ukumelwa okulula, okuncishisiwe kwedatha. Lokhu kusebenza kuhilela ukuguqula amalungu afanayo okuqala abe okukodwa okunobukhulu obumbalwa ngokususa, noma ukugoqa, ubukhulu bokugcina ngokuhambisana ne-eksisi yayo.
Isixazululo: Ukusebenzisa i-np.squeeze
Enye yezindlela zokubhekana nale nkinga ukusebenzisa i- numpy.cindezela umsebenzi. Lo msebenzi ususa okufakiwe okunohlangothi olulodwa kumumo wamalungu afanayo okokufaka.
import numpy as np
arr = np.random.rand(2, 3, 1)
print("Original array shape:", arr.shape)
collapsed_arr = np.squeeze(arr, axis=-1)
print("Collapsed array shape:", collapsed_arr.shape)
Isinyathelo ngesinyathelo Incazelo
Manje ake sihlukanise ikhodi futhi siqonde ukuthi isebenza kanjani.
1. Okokuqala, singenisa umtapo wolwazi we-NumPy njenge-np:
import numpy as np
2. Okulandelayo, sidala uhlu olungahleliwe lwe-dimensional engu-3 enomumo (2, 3, 1):
arr = np.random.rand(2, 3, 1)
print("Original array shape:", arr.shape)
3. Manje, sisebenzisa i np.cindezela umsebenzi wokugoqa ubukhulu bokugcina bamalungu afanayo ngokucacisa i i-axis ipharamitha njenge -1:
collapsed_arr = np.squeeze(arr, axis=-1)
print("Collapsed array shape:", collapsed_arr.shape)
4. Njengomphumela, sithola amalungu afanayo amasha anomumo othi (2, 3), okubonisa ukuthi ubukhulu bokugcina bugoqwe ngempumelelo.
Esinye Isixazululo: Hlela kabusha
Enye indlela yokugoqa ubukhulu bokugcina ukusebenzisa i- numpy.reshape sebenza ngamapharamitha afanele ukuze uthole umphumela oyifunayo.
collapsed_arr_reshape = arr.reshape(2, 3)
print("Collapsed array shape using reshape:", collapsed_arr_reshape.shape)
Kulesi simo, simise kabusha ngokusobala amalungu afanayo okuqala ukuze abe nomumo wokuthi (2, 3), sigoqe ngempumelelo ubukhulu bokugcina.
Imitapo yolwazi kanye Nemisebenzi Ehlobene
Ngaphandle kweNumPy, kuneminye imitapo yolwazi eminingana ku-Python ecosystem enikezela ngamathuluzi okusebenza ngama-arrays kanye namatrices. Omunye umtapo onjalo ngu I-SciPy, eyakhelwe phezu kwe-NumPy futhi inikeze ukusebenza okwengeziwe kwekhompyutha yesayensi. Emkhakheni wokufunda ngomshini, umtapo wolwazi I-TensorFlow iphinda isebenze ngama-tensor (okungukuthi, ama-multi-dimensional array) futhi inikeza isethi yayo yemisebenzi yokukhohlisa yematrix. Ngaphezu kwalokho, i- AmaPandas umtapo wolwazi ungasetshenziswa ukukhohlisa IdathaFrames, isakhiwo sedatha esiphezulu esingacatshangwa njengamathebula aqukethe amalungu afanayo. Ngaphezu kwalokho, i- numpy.newaxis ukusebenza kukuvumela ukuthi wengeze i-eksisi entsha kulungu elifanayo, okungaba usizo uma udinga ukunweba ubukhulu belungu elifanayo ukuze lifane nomumo odingekayo ukuze usebenze.
Sengiphetha, ikhono lokukhohlisa nokusebenza ngama-arrays ngempumelelo liyikhono elibalulekile emhlabeni wezinhlelo nesayensi yedatha. I-NumPy iwumtapo wolwazi onamandla ngokwedlulele ohlinzeka ngokusebenza okubanzi, futhi amasu okuqonda anjengokubhidliza ubukhulu bokugcina azoba yinzuzo ezimeni ezihlukahlukene lapho usebenza namasethi edatha amakhulu nayinkimbinkimbi.
