标签: nd4j
像这样的NDArray相当于numpy转置的out = np.transpose(in, (0,2,1,3))是什么? (in的排名为4)
NDArray
out = np.transpose(in, (0,2,1,3))
in
答案 0 :(得分:0)
显然,INDArray.permute(dimension rages)确实可以满足我的需求。
ND4J的outNDArray = inNDArray.permute(0,2,1,3)等同于NumPy的out = np.transpose(in, (0,2,1,3))
outNDArray = inNDArray.permute(0,2,1,3)