repeat

tripy.repeat(input: Tensor, repeats: int | DimensionSize, dim: int) → Tensor[source]

Repeats each element of a tensor after itself along the specified dimension.

Parameters:

  • input (Tensor) – The input tensor.

  • repeats (int | DimensionSize) – The number of times to repeat each element.

  • dim (int) – The dimension along which to repeat values.

Returns:

The new tensor.

Return type:

Tensor

Example: 1D tensor
1D tensor
1inp = tp.arange(4, dtype=tp.int32)
2out0 = tp.repeat(inp, 2, dim=0)

>>> inp
tensor([0, 1, 2, 3], dtype=int32, loc=gpu:0, shape=(4,))
>>> out0
tensor([0, 0, 1, 1, 2, 2, 3, 3], dtype=int32, loc=gpu:0, shape=(8,))
Example: 2D tensor
2D tensor
1inp = tp.reshape(tp.arange(4, dtype=tp.int32), (2, 2))
2out0 = tp.repeat(inp, 2, dim=0)
3out1 = tp.repeat(inp, 2, dim=1)

>>> inp
tensor(
    [[0, 1],
     [2, 3]], 
    dtype=int32, loc=gpu:0, shape=(2, 2))
>>> out0
tensor(
    [[0, 1],
     [0, 1],
     [2, 3],
     [2, 3]], 
    dtype=int32, loc=gpu:0, shape=(4, 2))
>>> out1
tensor(
    [[0, 0, 1, 1],
     [2, 2, 3, 3]], 
    dtype=int32, loc=gpu:0, shape=(2, 4))