factorModelSim1: Simulate some data from a sparse 2 factor model

In dajmcdon/aimer: Amplified, Initially Marginal, Eigenvector Regression

Description

Simulate some data from a sparse 2 factor model

Usage

factorModelSim1(n, p, p1 = 2, lambdas = c(5, 2), beta1 = 1, sig0 = 0.1,
  sig1 = 1)

Arguments

n	number of observations
p	number of predictors
p1	number of predictors which load onto the factors (requires 2*p1 <= p)
lambdas	relative weights of the factors
beta1	coefficient of y on the first factor (the second is calculated automatically to make the marginal correlation between those predictors and the response 0)
sig0	Standard deviation of the factor
sig1	Standard deviation of the noise

Value

A list containing:

1. X — n x p matrix of predictors

2. Y — response vector of length n

3. theta — regression coefficients on the predictors

4. beta — regression coefficients on the factors (first is beta1)

5. U — p x 2 matrix of factors

6. Lambda — matrix of factor weights (as input)

7. SigXY — marginal correlation between columns of X and Y

Examples

dat = factorModelSim1(100,1000,5,c(5,1),2,.1,.1)

dajmcdon/aimer documentation built on May 6, 2019, 1:31 a.m.