Skip to content

_tensor_sqrt

tensor_sqrt

Computes element-wise square root of tensor elements.

This method can be used with Numpy data: 

n = np.array([[1, 4, 6], [4, 9, 16]])
b = fe.backend.tensor_sqrt(n)  # [[1.0, 2.0, 2.44948974], [2.0, 3.0, 4.0]]

This method can be used with TensorFlow tensors: 

t = tf.constant([[1, 4, 6], [4, 9, 16]], dtype=tf.float32)
b = fe.backend.tensor_sqrt(t)  # [[1.0, 2.0, 2.4494898], [2.0, 3.0, 4.0]]

This method can be used with PyTorch tensors: 

p = torch.tensor([[1, 4, 6], [4, 9, 16]], dtype=torch.float32)
b = fe.backend.tensor_sqrt(p)  # [[1.0, 2.0, 2.4495], [2.0, 3.0, 4.0]]

Parameters:

Name Type Description Default
tensor Tensor

The input tensor.

 required

Returns:

Type Description
Tensor

The tensor that contains square root of input values.

Raises:

Type Description
ValueError

If tensor is an unacceptable data type.
Source code in fastestimator/fastestimator/backend/_tensor_sqrt.py 
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
def tensor_sqrt(tensor: Tensor) -> Tensor:
    """Computes element-wise square root of tensor elements.

    This method can be used with Numpy data:
    ```python
    n = np.array([[1, 4, 6], [4, 9, 16]])
    b = fe.backend.tensor_sqrt(n)  # [[1.0, 2.0, 2.44948974], [2.0, 3.0, 4.0]]
    ```

    This method can be used with TensorFlow tensors:
    ```python
    t = tf.constant([[1, 4, 6], [4, 9, 16]], dtype=tf.float32)
    b = fe.backend.tensor_sqrt(t)  # [[1.0, 2.0, 2.4494898], [2.0, 3.0, 4.0]]
    ```

    This method can be used with PyTorch tensors:
    ```python
    p = torch.tensor([[1, 4, 6], [4, 9, 16]], dtype=torch.float32)
    b = fe.backend.tensor_sqrt(p)  # [[1.0, 2.0, 2.4495], [2.0, 3.0, 4.0]]
    ```

    Args:
        tensor: The input tensor.

    Returns:
        The `tensor` that contains square root of input values.

    Raises:
        ValueError: If `tensor` is an unacceptable data type.
    """
    if tf.is_tensor(tensor):
        return tf.sqrt(tensor)
    elif isinstance(tensor, torch.Tensor):
        return torch.sqrt(tensor)
    elif isinstance(tensor, np.ndarray):
        return np.sqrt(tensor)
    else:
        raise ValueError("Unrecognized tensor type {}".format(type(tensor)))