# PyTorch一小时掌握之基本操作篇

## 创建数据

### torch.empty()

```torch.empty(*size, *, out=None, dtype=None, layout=torch.strided, device=None, requires_grad=False, pin_memory=False, memory_format=torch.contiguous_format) → Tensor
```

• size: 生成矩阵的形状, 必选
• dtype: 数据类型, 默认为 None

```# 创建一个形状为[2, 2]的矩阵
a = torch.empty(2, 2)
print(a)

# 创建一个形状为[3, 3]的矩阵
b = torch.empty(3, 3)
print(b)
```

tensor([[0., 0.],
[0., 0.]])
tensor([[0., 0., 0.],
[0., 0., 0.],
[0., 0., 0.]])

### torch.zeros()

```torch.zeros(*size, *, out=None, dtype=None, layout=torch.strided, device=None, requires_grad=False) → Tensor
```

• size: 生成矩阵的形状, 必选
• dtype: 数据类型, 默认为 None

```# 创建一个形状为[2, 2]的全零数组
a = torch.zeros([2, 2], dtype=torch.float32)
print(a)

# 创建一个形状为[3, 3]的全零数组
b = torch.zeros([3, 3], dtype=torch.float32)
print(b)
```

tensor([[0., 0.],
[0., 0.]])
tensor([[0., 0., 0.],
[0., 0., 0.],
[0., 0., 0.]])

### torch.ones()

```torch.ones(*size, *, out=None, dtype=None, layout=torch.strided, device=None, requires_grad=False) → Tensor
```

• size: 生成矩阵的形状, 必选
• dtype: 数据类型, 默认为 None

```# 创建一个形状为[2, 2]的全一数组
a = torch.ones([2, 2], dtype=torch.float32)
print(a)

# 创建一个形状为[3, 3]的全一数组
b = torch.ones([3, 3], dtype=torch.float32)
print(b)
```

tensor([[1., 1.],
[1., 1.]])
tensor([[1., 1., 1.],
[1., 1., 1.],
[1., 1., 1.]])

### torch.tensor()

```torch.tensor(data, *, dtype=None, device=None, requires_grad=False, pin_memory=False) → Tensor
```

• data: 数据 (数组, 元组, ndarray, scalar)
• dtype: 数据类型, 默认为 None

```# 通过数据创建张量
array = np.arange(1, 10).reshape(3, 3)
print(array)
print(type(array))

tensor = torch.tensor(array)
print(tensor)
print(type(tensor))
```

[[1 2 3]
[4 5 6]
[7 8 9]]

tensor([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]], dtype=torch.int32)

### torch.rand()

```torch.rand(*size, *, out=None, dtype=None, layout=torch.strided, device=None, requires_grad=False) → Tensor
```

• size: 生成矩阵的形状, 必选
• dtype: 数据类型, 默认为 None

```# 创建形状为[2, 2]的随机数矩阵
rand = torch.rand(2, 2)
print(rand)
```

tensor([[0.6209, 0.3424],
[0.3506, 0.7986]])

## 数学运算

```torch.add(input, other, *, out=None) → Tensor
```

```# 张量相加
input1 = torch.tensor([[1, 2], [3, 4]])
print(input1)

input2 = torch.tensor([[4, 3], [2, 1]])
print(input2)

print(output)
```

tensor([[1, 2],
[3, 4]])
tensor([[4, 3],
[2, 1]])
tensor([[5, 5],
[5, 5]])

### torch.sub()

```# 张量相减
input1 = torch.tensor([[1, 2], [3, 4]])
print(input1)

input2 = torch.tensor([[4, 3], [2, 1]])
print(input2)

output = torch.sub(input1, input2)
print(output)
```

tensor([[1, 2],
[3, 4]])
tensor([[4, 3],
[2, 1]])
tensor([[-3, -1],
[ 1, 3]])

### torch.matmul()

```# 张量矩阵相乘
input1 = torch.tensor([[1, 1, 1]])
print(input1)

input2 = torch.tensor([[3], [3], [3]])
print(input2)

output = torch.matmul(input1, input2)
print(output)
```

tensor([[1, 1, 1]])
tensor([[3],
[3],
[3]])
tensor([[9]])

## 索引操作

```# 简单的索引操作
ones = torch.ones([3, 3])
print(ones[: 2])
print(ones[:, : 2])
```

tensor([[1., 1., 1.],
[1., 1., 1.]])
tensor([[1., 1.],
[1., 1.],
[1., 1.]])