Pytorch中torch.cat()函數舉例解析

一. torch.cat（）函數解析

1. 函數說明

1.1 官網：torch.cat()，函數定義及引數說明如下圖所示：

1.2 函數功能

函數將兩個張量（tensor）按指定維度拼接在一起，注意：除拼接維數dim數值可不同外其餘維數數值需相同，方能對齊，如下面例子所示。torch.cat()函數不會新增維度，而torch.stack()函數會新增一個維度，相同的是兩個都是對張量進行拼接

2. 程式碼舉例

2.1 輸入兩個二維張量（dim=0）：dim=0對行進行拼接

a = torch.randn(2,3)
b =  torch.randn(3,3)
c = torch.cat((a,b),dim=0)
a,b,c

輸出結果如下：
(tensor([[-0.90, -0.37,  1.96],
         [-2.65, -0.60,  0.05]]),
 tensor([[ 1.30,  0.24,  0.27],
         [-1.99, -1.09,  1.67],
         [-1.62,  1.54, -0.14]]),
 tensor([[-0.90, -0.37,  1.96],
         [-2.65, -0.60,  0.05],
         [ 1.30,  0.24,  0.27],
         [-1.99, -1.09,  1.67],
         [-1.62,  1.54, -0.14]]))

2.2 輸入兩個二維張量（dim=1）: dim=1對列進行拼接

a = torch.randn(2,3)
b =  torch.randn(2,4)
c = torch.cat((a,b),dim=1)
a,b,c

輸出結果如下：
(tensor([[-0.55, -0.84, -1.60],
         [ 0.39, -0.96,  1.02]]),
 tensor([[-0.83, -0.09,  0.05,  0.17],
         [ 0.28, -0.74, -0.27, -0.85]]),
 tensor([[-0.55, -0.84, -1.60, -0.83, -0.09,  0.05,  0.17],
         [ 0.39, -0.96,  1.02,  0.28, -0.74, -0.27, -0.85]]))

2.3 輸入兩個三維張量：dim=0 對通道進行拼接

a = torch.randn(2,3,4)
b =  torch.randn(1,3,4)
c = torch.cat((a,b),dim=0)
a,b,c

輸出結果如下：
(tensor([[[ 0.51, -0.72, -0.02,  0.76],
          [ 0.72,  1.01,  0.39, -0.13],
          [ 0.37, -0.63, -2.69,  0.74]],
 
         [[ 0.72, -0.31, -0.27,  0.10],
          [ 1.66, -0.06,  1.91, -0.66],
          [ 0.34, -0.23, -0.18, -1.22]]]),
 tensor([[[ 0.94,  0.77, -0.41, -1.20],
          [-0.23, -1.03, -0.25,  1.67],
          [-1.00, -0.68, -0.35, -0.50]]]),
 tensor([[[ 0.51, -0.72, -0.02,  0.76],
          [ 0.72,  1.01,  0.39, -0.13],
          [ 0.37, -0.63, -2.69,  0.74]],
 
         [[ 0.72, -0.31, -0.27,  0.10],
          [ 1.66, -0.06,  1.91, -0.66],
          [ 0.34, -0.23, -0.18, -1.22]],
 
         [[ 0.94,  0.77, -0.41, -1.20],
          [-0.23, -1.03, -0.25,  1.67],
          [-1.00, -0.68, -0.35, -0.50]]]))

2.4 輸入兩個三維張量：dim=1對行進行拼接

a = torch.randn(2,3,4)
b =  torch.randn(2,4,4)
c = torch.cat((a,b),dim=1)
a,b,c

輸出結果如下：
(tensor([[[-0.86,  0.00, -1.26,  1.20],
          [-0.46, -1.08, -0.82,  2.03],
          [-0.89,  0.43,  1.92,  0.49]],
 
         [[ 0.24, -0.02,  0.32,  0.97],
          [ 0.33, -1.34,  0.76, -1.55],
          [ 0.38,  1.45,  0.27, -0.64]]]),
 tensor([[[ 0.82,  0.85, -0.30, -0.58],
          [-0.09,  0.40,  0.02,  0.75],
          [-0.70,  0.67, -0.88, -0.50],
          [-0.62, -1.65, -1.10, -1.39]],
 
         [[-0.85, -1.61, -0.35, -0.56],
          [ 0.00,  1.40,  0.41,  0.39],
          [-0.01,  0.04,  0.80,  0.41],
          [-1.21, -0.64,  1.14,  1.64]]]),
 tensor([[[-0.86,  0.00, -1.26,  1.20],
          [-0.46, -1.08, -0.82,  2.03],
          [-0.89,  0.43,  1.92,  0.49],
          [ 0.82,  0.85, -0.30, -0.58],
          [-0.09,  0.40,  0.02,  0.75],
          [-0.70,  0.67, -0.88, -0.50],
          [-0.62, -1.65, -1.10, -1.39]],
 
         [[ 0.24, -0.02,  0.32,  0.97],
          [ 0.33, -1.34,  0.76, -1.55],
          [ 0.38,  1.45,  0.27, -0.64],
          [-0.85, -1.61, -0.35, -0.56],
          [ 0.00,  1.40,  0.41,  0.39],
          [-0.01,  0.04,  0.80,  0.41],
          [-1.21, -0.64,  1.14,  1.64]]]))

2.5 輸入兩個三維張量：dim=2對列進行拼接

a = torch.randn(2,3,4)
b =  torch.randn(2,3,5)
c = torch.cat((a,b),dim=2)
a,b,c

輸出結果如下：
(tensor([[[ 0.13, -0.02,  0.13, -0.25],
          [ 1.42, -0.22, -0.87,  0.27],
          [-0.07,  1.04, -0.06,  0.91]],
 
         [[ 0.88, -1.46,  0.04,  0.35],
          [ 1.36,  0.64,  0.75,  0.39],
          [ 0.36,  1.13,  0.83,  0.56]]]),
 tensor([[[-0.47, -2.30, -0.49, -1.02,  1.74],
          [ 0.71,  0.89,  0.80, -0.05, -1.35],
          [-0.40,  0.26, -0.78, -1.50, -0.92]],
 
         [[-0.77, -0.01,  1.23,  0.70, -0.66],
          [ 0.28, -0.18, -0.91,  2.23,  1.14],
          [-1.93, -0.17,  0.15,  0.40,  0.32]]]),
 tensor([[[ 0.13, -0.02,  0.13, -0.25, -0.47, -2.30, -0.49, -1.02,  1.74],
          [ 1.42, -0.22, -0.87,  0.27,  0.71,  0.89,  0.80, -0.05, -1.35],
          [-0.07,  1.04, -0.06,  0.91, -0.40,  0.26, -0.78, -1.50, -0.92]],
 
         [[ 0.88, -1.46,  0.04,  0.35, -0.77, -0.01,  1.23,  0.70, -0.66],
          [ 1.36,  0.64,  0.75,  0.39,  0.28, -0.18, -0.91,  2.23,  1.14],
          [ 0.36,  1.13,  0.83,  0.56, -1.93, -0.17,  0.15,  0.40,  0.32]]]))

總結

到此這篇關於Pytorch中torch.cat()函數舉例解析的文章就介紹到這了,更多相關Pytorch中torch.cat()函數內容請搜尋it145.com以前的文章或繼續瀏覽下面的相關文章希望大家以後多多支援it145.com！