Skip to main content

Tensor

Tensor is a multi-dimensional array with a uniform type. In Hedgehog Lab, tensor is only used to store data. There are not any operations on tensor supported in Hedgehog Lab.

Create a tensor​

You can create a tensor by using the Tensor(n_dim_array) constructor or tensor wrapper function. Both of them take an n-dimensional array as input.

Member variables and methods​

val: any[]: The value of the tensor. It is a n-dimensional array.

ndim: number: The number of dimensions of the tensor.

shape: number[]: The shape of the tensor.

dimensions(): number[]: Return the shape of the tensor.

toString(): string: Return the string representation of the tensor, for example, a three-dimensional tensor will be serialized as a string in this format:

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

toStringDenseMode: Return the string representation of the tensor in dense mode, for example, a three-dimensional tensor will be serialized as a string in this format:

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

clone(): Tensor: Return a copy of the tensor.

copy( A: Tensor ): Tensor: Copy the value of tensor A to the tensor.

zerosAsShape( shape: number[] ): Tensor: Return a tensor with the same shape as the input shape, and all elements are zeros.

zeros( d1: number, d2: number, ...dn: number ): Tensor: Return a tensor with the input shape, and all elements are zeros.