# tinker.types.TensorData

## *class* [**tinker.types.TensorData**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L25)()

**Fields:**

- [**dtype**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L26) (*[TensorDtype](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_dtype.py#L5)*)

- [**shape**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L28) (*Optional\[List[int]\]*) – Optional.

  The shape of the tensor (see PyTorch tensor.shape). The shape of a one-dimensional list of length N is `(N,)`. Can usually be inferred if not provided, and is generally inferred as a 1D tensor.

- [**sparse_crow_indices**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L36) (*Optional\[List[int]\]*) – Optional CSR compressed row pointers. When set, this tensor is sparse CSR:

  - data contains only the non-zero values (flattened)
  - sparse_crow_indices contains the row pointers (length = nrows + 1)
  - sparse_col_indices contains the column indices (length = nnz)
  - shape is required and specifies the dense shape

- [**sparse_col_indices**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L44) (*Optional\[List[int]\]*) – Optional CSR column indices. Must be set together with sparse_crow_indices.

### *property* [**data**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L82)

Flattened tensor data as array of numbers.

**Returns:** *Union\[List[int], List[float]\]*

### [**from_torch_sparse**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L104)(*tensor*)

Create a sparse CSR TensorData from a dense 2-D torch tensor.

Automatically detects sparsity and encodes as CSR when it saves space. Falls back to dense if the tensor is 1-D or mostly non-zero.

**Parameters:**

- [**tensor**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L104) (*torch.Tensor*)

**Returns:** *[TensorData](https://tinker-docs.thinkingmachines.ai/tinker/api-reference/types/tensordata/index.md)*

### [**to_numpy**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L134)()

Convert TensorData to numpy array.

**Returns:** *npt.NDArray[Any]*

### [**to_torch**](https://github.com/thinking-machines-lab/tinker/blob/main/src/tinker/types/tensor_data.py#L140)()

Convert TensorData to torch tensor.

**Returns:** *torch.Tensor*

## Referenced by

- [tinker.types.ForwardBackwardOutput](https://tinker-docs.thinkingmachines.ai/tinker/api-reference/types/forwardbackwardoutput/index.md)