mouthwash_coordinate: Coordinate ascent for optimizing t likelihood with uniform...
In dcgerard/vicar: Various Ideas for Confounder Adjustment in Regression

View source: R/mouthwash_uniform_coordinate.R

Set degrees_freedom = Inf if you want to use a normal likelihood.

mouthwash_coordinate(
  pi_init,
  z_init,
  xi_init,
  betahat_ols,
  S_diag,
  alpha_tilde,
  a_seq,
  b_seq,
  lambda_seq,
  degrees_freedom,
  scale_var = TRUE,
  tol = 10^-6,
  maxit = 100,
  plot_update = FALSE,
  var_inflate_pen = 0
)

`pi_init`	The initial values of the mixing proportions. A numeric vector of positive values that sum to one.
`z_init`	The initial values of z2. A numeric vector.
`xi_init`	The initial (or known) value of the variance inflation term.
`betahat_ols`	A vector of numerics. The OLS estimates of the coefficients of interest.
`S_diag`	A vector of positive numerics. The estimated standard errors.
`alpha_tilde`	A matrix. The number of rows should be equal the length of betahat_ols. The number of columns should equal the number of hidden confounders.
`a_seq`	The grid of lower bounds for the uniforms if `mixing_dist` is one of the uniforms.
`b_seq`	The grid of upper bounds for the uniforms if `mixing_dist` is one of the uniforms.
`lambda_seq`	A numeric vector with elements all greater than or equal to 1. These are the tuning parameters for the mixing proportions.
`degrees_freedom`	The degrees of freedom of the t-distribution if `likelihood = "t"`.
`scale_var`	A logical. Should we estimate a variance inflation parameter (`TRUE`) or not (`FALSE`)?
`tol`	The tolerance for the stopping condition.
`maxit`	The maximum number of iterations to run in the optimization scheme.
`plot_update`	A logical. Should I plot the the path of the log-likelihood (`TRUE`) or not (`FALSE`)?
`var_inflate_pen`	The penalty to apply on the variance inflation parameter. Defaults to 0, but should be something non-zero when `alpha = 1` and `scale_var = TRUE`.