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

compact_order

def compact_order(shape: IntTuple, order: IntTuple) -> IntTuple

Create a compact stride based on shape and order.

This function generates a stride tuple where lower order numbers imply faster varying strides. The resulting shape and stride form a bijective layout.

Performance:

• Always inlined for optimal performance in tight loops.
• Flattens inputs and re-nests results for consistent behavior.

Example:

from layout import IntTuple
from layout.int_tuple import compact_order

# Create a compact layout with dimensions (2,3,4,5) and ordering (1,4,3,5)
var x = compact_order(IntTuple(2,3,4,5), IntTuple(1,4,3,5))  # returns (1,8,2,24)

# Create a compact layout with nested dimensions and corresponding ordering
var y = compact_order(IntTuple(2,IntTuple(3,4),5), IntTuple(1,IntTuple(2,3),4))  # returns (1,(2,6),24)

Args:

• ​shape (IntTuple): The shape tuple defining dimensions.
• ​order (IntTuple): The order tuple defining the relative ordering of dimensions.

Returns:

IntTuple: A stride tuple that creates a compact memory layout according to the specified order.