SCM() is the main function that calculates the estimated ATE by SC method and Bootstrap CI. It calls subject_SC() and lambdacv().
Usage
SCM(
data,
outcome_col_name,
trial_status_col_name,
treatment_col_name,
covariates_col_name,
T_cross,
Bootstrap = TRUE,
R = 100,
bootstrap_CI_type = "bca",
alpha = 0.05,
lambda.min = 0,
lambda.max = 0.1,
nlambda = 2,
parallel = "no",
ncpus = 1,
quiet = TRUE
)Arguments
- data
A data frame containing all subject-level data.
- outcome_col_name
Character vector of outcome column names.
- trial_status_col_name
Name of the trial status column.
- treatment_col_name
Name of the treatment column.
- covariates_col_name
Character vector of covariate column names.
- T_cross
Integer crossover time point.
- Bootstrap
Logical. Whether to use bootstrap inference.
- R
Number of bootstrap replicates.
- bootstrap_CI_type
Type of bootstrap CI (e.g.
"bca","perc").- alpha
Significance level.
- lambda.min
Numeric. Minimum penalty parameter.
- lambda.max
Numeric. Maximum penalty parameter.
- nlambda
Integer. Number of lambda values for cross-validation.
- parallel
Character. Parallelization type for
boot.- ncpus
Integer. Number of CPUs for parallel bootstrap.
- quiet
Logical. If
TRUE, suppress printed output.