generate_parameters: Pick parameter values for simulation, from a normal...

In audreyrenson/paralleltrends: Estimators for sustained treatment effects under parallel trends assumptions

Pick parameter values for simulation, from a normal distribution

Description

Pick parameter values for simulation, from a normal distribution

Usage

generate_parameters(
  Tt,
  mu_Beta_W1 = 0.2,
  mu_Beta_W2 = 0.2,
  mu_Beta_A = 0.2,
  mu_Beta_Y = 0.2,
  sd_Beta_W1 = 0.2,
  sd_Beta_W2 = 0.2,
  sd_Beta_A = 0.2,
  sd_Beta_Y = 0.2,
  range_ymeans = c(-5, 5),
  Beta_YU = 0.1
)

Arguments

Tt	Final period (t=0,1,...,Tt)
mu_Beta_W1	numeric. mean for parameters in covariate models
mu_Beta_W2	numeric. mean for parameters in covariate models
mu_Beta_A	numeric. mean for parameters in treatment models
mu_Beta_Y	numeric. mean for parameters in outcome models
sd_Beta_W1	numeric. standard deviation for parameters in covariate models
sd_Beta_W2	numeric. standard deviation for parameters in covariate models
sd_Beta_A	numeric. standard deviation for parameters in treatment models
sd_Beta_Y	numeric. standard deviation for parameters in outcome models
range_ymeans	numeric vector of length 2. minimum and maximum values for marginal means of Y
Beta_YU	numeric. constant coefficient for effect of unmeasured, time-invariant, binary covariate U in Y models

Value

list of matrices, one for each variable to be generated (U, W1, W2, A, Y). Each matrix has T+1 rows and a different number of columns corresponding to the coefficients of a (generalized) linear model. The columns names refer to the term in that model.

See Also

Other simulation functions: generate_data()

Examples

generate_parameters(Tt=2)

audreyrenson/paralleltrends documentation built on May 4, 2022, 2:53 a.m.