ggml_task: Construct a single-cell compute task

In ggmlR: 'GGML' Tensor Operations for Machine Learning

Construct a single-cell compute task

Description

A ggml_task is the contract object passed from the extraction layer to the dispatch layer. It bundles the operation name, the dense feature matrix (features in rows, cells in columns — the single-cell convention), the operation parameters and the requested device. It performs no computation.

Usage

ggml_task(op, matrix, params = list(), device = c("auto", "vulkan", "cpu"))

Arguments

op	Operation name; must be registered in ggml_ops_registry (e.g. "embed").
matrix	A numeric matrix (dense) or dgCMatrix (sparse). Rows are features (genes), columns are cells.
params	Named list of operation parameters (e.g. n_components).
device	"vulkan", "cpu" or "auto" (default).

Value

An object of class ggml_task.

See Also

ggml_run, ggml_ops_registry

ggmlR documentation built on July 14, 2026, 1:08 a.m.