miss_gen_an: generation and analysis of incomplete data
In yuliasidi/nibinom:

Description Usage Arguments Value See Also Examples

View source: R/miss_gen_an.R

imposes missing values, and performs analysis of incomplete data

miss_gen_an(dt, m2, b_trt = 0, b_y = 0, b_x1 = 0, b_x2 = 0,
  b_ty = 0, do = 0.1, seed, dt_out = FALSE, sing_anal = TRUE,
  mice_anal = FALSE, bw_anal = FALSE, mu_C = 1, sd_C = 0,
  mu_T = 1, sd_T = 0, ci_method = wald_ci, seed_mice,
  alpha = 0.025, n_mi, m_mi, method = c("wald", "wn", "fm"),
  pro_wnmi = FALSE)

`dt`	dataframe, on which missing values are imposed
`m2`	numeric, margin
`b_trt`	numeric, group missingness parameter, Default: 0
`b_y`	numeric, outcome missingness parameter, Default: 0
`b_x1`	numeric, covariate x1 missingness parameter, Default: 0
`b_x2`	numeric, covariate x2 missingness parameter, Default: 0
`b_ty`	numeric, covariatet treatment by group interaction, Default: 0
`do`	numeric, target drop-out rate, Default: 0.1
`seed`	numeric, seed specification before deleting observations
`dt_out`	logic, determines if the incomplete dataset needs to be returned (if returned, no analysis will be done), Default: FALSE
`sing_anal`	logic, determines if single value analysis needs to be performed (complete case analysis, best and worst case scenario imputation), Default: TRUE
`mice_anal`	logic, determines if multiple imputation via MICE needs to be performed (nested imputation), Default: FALSE
`bw_anal`	logic, determines if the following single value imputation needs to be performed: best case scenario for group 'c' and worst case scenario for 't', Default: FALSE
`mu_C`	numeric, mean parameter for k multiplier normal distribution for group 'c', Default: 1
`sd_C`	numeric, stadnard deviation parameter for k multiplier normal distribution for group 'c', Default: 0
`mu_T`	Pnumeric, mean parameter for k multiplier normal distribution for group 't',Default: 1
`sd_T`	numeric, stadnard deviation parameter for k multiplier normal distribution for group 't', Default: 0
`ci_method`	ci function name, Default: wald_ci
`seed_mice`	numeric, Default: seed_mice
`alpha`	numeric, one-sided alpha, Default: 0.025
`n_mi`	numeric, number of multiple imputed patient level data
`m_mi`	numeric, number of multiple imputed models
`method`	character, method to construct confidence interval, Default: c('wald','wn','fm')
`pro_wnmi`	logic, specifies whether proper wn cobminaiton rules should be used, if TRUE, then bin2mi package needs to be used, Default: FALSE

list

spread glm set_names

 t <- sim_cont(
 p_C = 0.6, p_T = 0.5, n_arm = 200,
 mu1 = 4, mu2 = 100, sigma1 = 1,
 sigma2 = 20, r12 = -0.3,
 b1 = 0.1, b2 = -0.01)
 miss_gen_an(t, m2 = 0.1 , b_x1 = 0.5, do = 0.1, seed = 1555, seed_mice = 1555, alpha = 0.025)