For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Mojo function

idx2crd

def idx2crd[idx_t: IntTuple, shape_t: IntTuple, stride_t: IntTuple](idx: RuntimeTuple[idx_t, element_type=idx.element_type], shape: RuntimeTuple[shape_t, element_type=shape.element_type], stride: RuntimeTuple[stride_t, element_type=stride.element_type]) -> RuntimeTuple[idx2crd(idx_t, shape_t, stride_t), element_type=shape.element_type]

Converts a linear index to multi-dimensional coordinates. This function transforms a flat index into coordinate values based on the provided shape and stride information. This is essential for mapping linear memory accesses to multi-dimensional tensor elements.

Constraints:

The index must be a scalar value (not a tuple).

Parameters:

• ​idx_t (IntTuple): IntTuple type of the index.
• ​shape_t (IntTuple): IntTuple type of the shape.
• ​stride_t (IntTuple): IntTuple type of the stride.

Args:

• ​idx (RuntimeTuple[idx_t, element_type=idx.element_type]): The linear index to convert.
• ​shape (RuntimeTuple[shape_t, element_type=shape.element_type]): The shape of the multi-dimensional array.
• ​stride (RuntimeTuple[stride_t, element_type=stride.element_type]): The stride values for each dimension.

Returns:

RuntimeTuple[idx2crd(idx_t, shape_t, stride_t), element_type=shape.element_type]: A RuntimeTuple containing the multi-dimensional coordinates.

def idx2crd[idx_t: IntTuple, shape_t: IntTuple](idx: RuntimeTuple[idx_t, element_type=idx.element_type], shape: RuntimeTuple[shape_t, element_type=shape.element_type]) -> RuntimeTuple[idx2crd(idx_t, shape_t, prefix_product(shape_t)), element_type=shape.element_type]

Converts a linear index to multi-dimensional coordinates using shape-derived strides. This is a convenience overload of idx2crd that automatically calculates the stride values from the shape using prefix_product. This is the common case for row-major storage order tensors.

Parameters:

• ​idx_t (IntTuple): IntTuple type of the index.
• ​shape_t (IntTuple): IntTuple type of the shape.

Args:

• ​idx (RuntimeTuple[idx_t, element_type=idx.element_type]): The linear index to convert.
• ​shape (RuntimeTuple[shape_t, element_type=shape.element_type]): The shape of the multi-dimensional array.

Returns:

RuntimeTuple[idx2crd(idx_t, shape_t, prefix_product(shape_t)), element_type=shape.element_type]: A RuntimeTuple containing the multi-dimensional coordinates calculated using automatically derived strides from the shape.