Skip to contents

Subset MLX array

Source: R/subset.R
mlx_subset.Rd

MLX subsetting mirrors base R for the common cases while avoiding a few of the language's historical footguns:

Usage

# S3 method for class 'mlx'
x[..., drop = FALSE]

# S3 method for class 'mlx'
x[...] <- value

Arguments

x

An mlx array, or an R array/matrix/vector that will be converted via as_mlx().

...

Indices for each dimension. Provide one per axis; omitted indices select the full extent. Logical indices recycle to the dimension length.

drop

Should dimensions be dropped? (default: FALSE)

value

Replacement values, recycled to match the selection.

Value

The subsetted MLX object.

Details

  • Numeric indices: positive (1-based) and purely negative vectors are supported. Negative indices drop the listed elements, just as in base R. Mixing signs is an error and 0 is not allowed.

  • Logical indices: recycled to the target dimension length. Logical masks may be mixed with numeric indices across dimensions.

  • Matrices/arrays: numeric matrices (or higher dimensional arrays) select individual elements, one coordinate per row. The trailing dimension must match the array rank and entries must be positive; negative matrices are rejected to avoid ambiguous complements.

  • mlx indices: mlx vectors, logical masks, and matrices behave the same as their R equivalents. One-dimensional MLX arrays are treated as vectors rather than 1-column matrices.

  • drop: dimensions are preserved by default (drop = FALSE), matching the package's preference for explicit shapes.

  • Unsupported: character indices and named lookups are not implemented.

See also

mlx.core.take

Examples

x <- mlx_matrix(1:9, 3, 3)
x[1, ]
#> mlx array [1 x 3]
#>   dtype: float32
#>   device: gpu
#>   values:
#>      [,1] [,2] [,3]
#> [1,]    1    4    7