Control parameters
Usage
mlxs_glm_control(
epsilon = 1e-08,
epsilon_f64 = 1e-06,
maxit = 25,
trace = FALSE,
rank_tol = NULL
)Arguments
- epsilon
Convergence tolerance parameter, interpreted as in
stats::glm.control(). Iterations converge whenabs(deviance - deviance_old)/(abs(deviance) + 0.1) < epsilon.- epsilon_f64
Move operations to float64 on the cpu when convergence is this close (using the same expression as above). Doing this allows more precision but slows computation.
- maxit
Maximum number of IWLS iterations.
- trace
Logical: trace each iteration?
- rank_tol
Optional relative tolerance used to detect rank-deficient systems.
NULLuses the package default, which varies by dtype and is 1e-6 for float32 matrices. Set toFALSEto skip rank checks entirely. Note that higher numbers indicate lower tolerance.