simitation (version 0.0.7)

simstudy.lm: simstudy.lm

Description

simstudy.lm

Usage

simstudy.lm(
  the.steps,
  n,
  num.experiments,
  the.formula,
  conf.level = 0.95,
  the.quantiles = c(0.025, 0.1, 0.25, 0.5, 0.75, 0.9, 0.975),
  experiment.name = "experiment",
  step.split = "~",
  coef.name = "Coefficient",
  estimate.name = "Estimate",
  lm.p.name = "Pr(>|t|)",
  f.p.name = "f.pvalue",
  seed = 41,
  vstr = 3.6
)

Value

A list containing four elements: 'the.steps': The steps or processes defined for the simulation, detailing the structure and relationships among the variables. 'simdat': A data frame of the simulated data based on linear regression. 'statistics: A data frame containing regression statistics, coefficient estimates, and related results. 'sim.analysis': Summary statistics or analysis results from the simulated linear regression.

Arguments

the.steps

A character vector of variables to simulate. The variables are simulated in the order specified. Later variables can be generated to depend on earlier variables. The possible specifications include: Normal "X ~ N(100, 5)" with the mean and SD. Uniform "X ~ U(0, 100)" with the minimum and maximum. Poisson "X ~ Poisson(3)" with the mean. Binary "X ~ Binary(0.5)" with the probability of success. Binomial "X ~ Bin(10, 0.2)" with the number of trials and probability of success. Categorical "Diet ~ sample(('Light', 'Moderate', 'Heavy'), (0.2, 0.45, 0.35))" with the values in the first set of parentheses and their respective probabilities in the second. Regression "Healthy.Lifestyle ~ logistic(log(0.45) - 0.1 * (Age -45) + 0.05 * Female + 0.01 * Health.Percentile + 0.5 * Exercise.Sessions - 0.1 * (Diet == 'Moderate') - 0.4 * (Diet == 'Heavy'))" Linear Regression "Weight ~ lm(150 - 15 * Female + 0.5 * Age - 0.1 * Health.Percentile - 0.2 * Exercise.Sessions + 5 * (Diet == 'Moderate') + 15 * (Diet == 'Heavy') - 2 * Healthy.Lifestyle + N(0, 10))". Note that the error term may be specified symbolically with any of the above distributions.

n

A numeric value for the number of observations in each experiment.

num.experiments

A numeric value representing the number of simulated experiments.

the.formula

A formula object or character value specifying the formula for the regression model.

conf.level

A numeric value between 0 and 1 representing the confidence level (1 - significance level).

the.quantiles

A numeric vector of values between 0 and 1. Summary statistics to analyze the tests will return the specified quantiles.

experiment.name

A character value providing the name for the column identifying the experiment.

step.split

A character value that separates the name of the variable to be simulated (left side) from its distribution (right side). Using the.steps = "X ~ N(0,1)" with step.split = "~" will generate a variable named X from a standard Normal distribution.

coef.name

A character value specifying the column of the.coefs that contains the names of the input variables of the linear regression model.

estimate.name

A character value specifying the column of the.coefs that contains the estimated coefficients of the linear regression model.

lm.p.name

A character value specifying the column of the.coefs that contains the p-values for the tests of the estimated coefficients of the linear regression model.

f.p.name

A character value specifying the column of summary.stats that contains the p-value for the linear regression model's F test.

seed

A single numeric value, interpreted as an integer, or NULL. See help(set.seed).

vstr

A character string containing a version number, e.g., "1.6.2". The default RNG configuration of the current R version is used if vstr is greater than the current version. See help(set.seed).

Examples

step.age <- "Age ~ N(45, 10)"
step.female <- "Female ~ binary(0.53)"
step.health.percentile <- "Health.Percentile ~ U(0,100)"
step.exercise.sessions <- "Exercise.Sessions ~ Poisson(2)"
step.diet <- "Diet ~ sample(('Light', 'Moderate', 'Heavy'),
(0.2, 0.45, 0.35))"

step.healthy.lifestyle <- "Healthy.Lifestyle ~ logistic(log(0.45) - 0.1 *
(Age -45) + 0.05 * Female + 0.01 * Health.Percentile + 0.5 *
Exercise.Sessions - 0.1 * (Diet == 'Moderate') - 0.4 *
(Diet == 'Heavy'))"

step.weight <- "Weight ~ lm(150 - 15 * Female + 0.5 * Age - 0.1 *
Health.Percentile - 0.2 * Exercise.Sessions  + 5 * (Diet == 'Moderate') +
15 * (Diet == 'Heavy') - 2 * Healthy.Lifestyle + N(0, 10))"

the.steps <- c(step.age, step.female, step.health.percentile,
step.exercise.sessions, step.diet, step.healthy.lifestyle,
step.weight)

study.lm <- simstudy.lm(the.steps = the.steps, n = 100,
num.experiments = 2,
the.formula = Weight ~ Age + Female + Health.Percentile +
Exercise.Sessions + Healthy.Lifestyle,
conf.level = 0.95, the.quantiles = c(0.25, 0.75),
experiment.name = "sim", seed = 11)