pytorch tensor api

pytorch

a=torch.rand(3,4)
b=torch.rand(3,4)
a.ceil(input,out=None)#进一法
a.clamp(min,max)#裁剪
a.expand(*size)#复制
a.gather(dim,index)#取出一个和index一个shape的tensor，在dim维取
a.index_add_(dim,index,tensor)#在dim维特定index上加上tensor
a.index_copy_(dim,index,tensor)
a.index_fill_(dim,index,val)
a.index_select(dim,index)
torch.lerp(start,end,weight)#outi=starti++weight(endi-starti)
torch.logsumexp(input,dim,keepdim=False,out=None)
a.map_(tensor,callable)#a和tensor element-wise 执行callable
a.masked_scatter(mask,b)#在mask为１的位置讲ｂ copy给a
a.masked_select(mask)#返回mask为１的所有数值
a.permute(*dims)#permute the dimensions of this tensor