uni_fun: Univariate regressions
In sahirbhatnagar/shim: High Dimensional Interactions with an Environment Variable

Description Usage Arguments Value Note Author(s) See Also Examples

Function used to create initial estimates in fitting algorithm of the strong heredity interaction model implemented in the shim function

1 2	uni_fun(x, y, type = c("ridge", "univariate"), variables, include.intercept = FALSE)

`x`	Design matrix of dimension `n x q`, where `n` is the number of subjects and q is the total number of variables; each row is an observation vector. This must include all main effects and interactions as well, with column names corresponding to the names of the main effects (e.g. `x1, x2, E`) and their interactions (e.g. `x1:E, x2:E`). All columns should be scaled to have mean 0 and variance 1; this is done internally by the `shim` function.
`y`	response variable (matrix form) of dimension `n x 1`
`type`	The procedure used to estimate the regression coefficients. If `"univariate"` then a series of univariate regressions is performed with the response variable `y`. If `"ridge"` then ridge regression is performed using the `cv.glmnet` function and the tuning parameter is chosen using 10 fold cross validation. The default is `"ridge"`.
`variables`	character vector of variable names for which you want the univariate regression estimate. Must be contained in the column names of `x`. This only applies if `type="univariate"`.
`include.intercept`	Should intercept be fitted (default is `FALSE`). Should be set to `TRUE` if `y` is not centered

Regression coefficients as a q x 1 data.frame

p is defined as the number of main effects. I have introduced q as being the total number of variables (e.g. the number of columns in the design matrix).

Sahir Bhatnagar

Maintainer: Sahir Bhatnagar sahir.bhatnagar@mail.mcgill.ca

shim, cv.glmnet

# number of observations
n <- 100

# number of predictors
p <- 5

# environment variable
e <- sample(c(0,1), n, replace = T)

# main effects
x <- cbind(matrix(rnorm(n*p), ncol = p), e)

# need to label columns
dimnames(x)[[2]] <- c(paste0("x",1:p), "e")

# design matrix without intercept
X <- model.matrix(~(x1+x2+x3+x4+x5)*e-1, data = as.data.frame(x))

# response
Y <- X %*% rbinom(ncol(X), 1, 0.2) + 3*rnorm(n)

uni_fun(X, Y)