controlOut
constructs a list with all necessary control parameters
for outcome model.
Usage
controlOut(
epsilon = 1e-04,
maxit = 100,
trace = FALSE,
k = 1,
penalty = c("SCAD", "lasso", "MCP"),
a_SCAD = 3.7,
a_MCP = 3,
lambda_min = 0.001,
nlambda = 100,
nfolds = 10,
treetype = "kd",
searchtype = "standard",
predictive_match = 1:2,
pmm_weights = c("none", "prop_dist"),
pmm_k_choice = c("none", "min_var"),
pmm_reg_engine = c("glm", "loess")
)
Arguments
- epsilon
Tolerance for fitting algorithms. Default is
1e-6
.- maxit
Maximum number of iterations.
- trace
logical value. If
TRUE
trace steps of the fitting algorithms. Default isFALSE
.- k
The k parameter in the
RANN::nn2()
function. Default is 5.- penalty
penalty algorithm for variable selection. Default is
SCAD
- a_SCAD
The tuning parameter of the SCAD penalty for outcome model. Default is 3.7.
- a_MCP
The tuning parameter of the MCP penalty for outcome model. Default is 3.
- lambda_min
The smallest value for lambda, as a fraction of lambda.max. Default is .001.
- nlambda
The number of lambda values. Default is 100.
- nfolds
The number of folds during cross-validation for variables selection model.
- treetype
Type of tree for nearest neighbour imputation passed to
RANN::nn2()
function.- searchtype
Type of search for nearest neighbour imputation passed to
RANN::nn2()
function.- predictive_match
(Only for predictive mean matching) Indicates how to select 'closest' unit from nonprobability sample for each unit in probability sample. Either
1
(default) or2
where1
is matching by minimizing distance between \(\hat{y}_{i}\) for \(i \in S_{A}\) and \(y_{j}\) for \(j \in S_{B}\) and2
is matching by minimizing distance between \(\hat{y}_{i}\) for \(i \in S_{A}\) and \(\hat{y}_{i}\) for \(i \in S_{A}\).- pmm_weights
(Only for predictive mean matching) Indicate how to weight
k
nearest neighbours in \(S_{B}\) to create imputed value for units in \(S_{A}\). The default value"none"
indicates that mean ofk
nearest \(y\)'s from \(S_{B}\) should be used whereas"prop_dist"
results in weighted mean of thesek
values where weights are inversely proportional to distance between matched values.- pmm_k_choice
Character value indicating how
k
hyper-parameter should be chosen, by default"none"
meaningk
provided incontrol_outcome
argument will be used. For now the only other option"min_var"
means thatk
will be chosen by minimizing estimated variance of estimator for mean. Parameterk
provided in this control list will be chosen as starting point.- pmm_reg_engine
TODO
See also
nonprob()
-- for fitting procedure with non-probability samples.