Package 'simcovariates'

Functions to define individual covariate properties

Description

These functions serve as constructors to create objects that define the properties of different covariate types (continuous, binary, categorical). These definition objects are then used within the gen_data function's covariate_defns argument to specify the characteristics of the simulated covariates.

Usage

continuous(mean, sd)

binary(prob)

categorical(levels, probs = NULL)

Details

Constructors for Covariate Definitions

---

Generate simulated datasets of IPD covariates and outcome for a trial

Description

Generate simulated datasets of IPD covariates and outcome for a trial

Usage

gen_data(
  N,
  b_0,
  b_trt,
  covariate_defns,
  b_prognostic = NULL,
  b_effect_modifier = NULL,
  cor_matrix = NULL,
  trt_assignment = list(prob_trt1 = 0.5),
  error_params = list(sd = 1),
  family = gaussian(link = "identity")
)

Arguments

cor_matrix	Applies to all latent covariates
family

Value

Data frame of simulated IPD

---

Generate simulated datasets of IPD covariates and outcome for a trial

Description

Generate simulated datasets of IPD covariates and outcome for a trial

Usage

gen_data_orig(
  N,
  b_trt,
  meanX = NULL,
  sdX = NULL,
  b_X = NULL,
  meanX_EM = NULL,
  sdX_EM = NULL,
  b_EM = NULL,
  prob_X_bin = NULL,
  b_X_bin = NULL,
  prob_EM_bin = NULL,
  b_EM_bin = NULL,
  b_0,
  corX,
  allocation,
  sigma = 1,
  family = binomial("logit")
)

Arguments

N	Total number of patients
b_trt	b coefficient for active treatment vs. common comparator
meanX	Mean vector of each normally-distributed covariate X
sdX	Standard deviation vector of each covariate X
b_X	b coefficients for each prognostic variable X
meanX_EM	Mean vector of each normally-distributed EM covariate X
sdX_EM	Standard deviation vector of each EM covariate X
b_EM	b coefficients effect modifiers
prob_X_bin	Probability for each binary distributed covariate X
b_X_bin	b coefficients for each prognostic variable X
prob_EM_bin	Probability for each binary distributed EM covariate X
b_EM_bin	b coefficients effect modifiers
b_0	Intercept coefficient
corX	Covariate correlation coefficient of X
allocation	Allocation to active treatment as proportion of total; 0 to 1
sigma	Standard deviation of outcome; optional
family	Family object

Value

Data frame of X, trt and y

Examples

## Not run: 
x <- gen_data(
 N = 100,
 b_trt = log(0.17),
 b_X = -log(0.5),
 b_EM = -log(0.67),
 b_0 = -0.62,
 meanX = c(0.6, 0.6),
 sdX = c(0.4, 0.4),
 meanX_EM = c(0.6, 0.6),
 sdX_EM = c(0.4, 0.4),
 corX = 0.2,
 allocation = 2/3)

head(x)

## End(Not run)

---