powered by
Provides control parameters for the Quantile Penalized Generalized Estimating Equations (QPGEE) fitting procedure. Similar in spirit to `geese.control()` in the geepack package.
qpgeeControl( epsilon = 1e-04, decay = 1, maxit = 100, trace = FALSE, standardize = FALSE, shrinkCutoff = 1e-04 )
A list with the components epsilon, maxit,
epsilon
maxit
trace, and shrinkCutoff.
trace
shrinkCutoff
Convergence tolerance for the parameter estimates. Iteration stops when the maximum change in coefficients is below this value.
Decay rate of learning step.
Maximum number of iterations.
Logical indicating if output should be produced for each iteration. (You can decide how much information to show inside the C++/R loop.)
Logical indicating whether to scale X.
Threshold below which coefficients are shrunk to zero (removal of “small” coefficients).
ctrl <- qpgeeControl(epsilon = 1e-5, maxit = 200, trace = TRUE)
Run the code above in your browser using DataLab