R/rliue.r

In lrmest: Different Types of Estimators to Deal with Multicollinearity

rliu<-function (formula, r, R, delt, d, data = NULL, na.action, ...) 
{
    d <- as.matrix(d)
    d1 <- d[1L]
    rliues <- function(formula, r, R, delt, d1, data = NULL, 
        na.action, ...) {
        cal <- match.call(expand.dots = FALSE)
        mat <- match(c("formula", "data", "na.action"), names(cal))
        cal <- cal[c(1L, mat)]
        cal[[1L]] <- as.name("model.frame")
        cal <- eval(cal)
        y <- model.response(cal)
        md <- attr(cal, "terms")
        x <- model.matrix(md, cal, contrasts)
        s <- t(x) %*% x
        xin <- solve(s)
        r <- as.matrix(r)
        RC <- matrix(R, NCOL(x))
        RR <- t(RC)
        if (is.matrix(R)) 
            RR <- R
        else RR <- RR
        del <- as.matrix(delt)
        bb <- xin %*% t(x) %*% y
        rb <- bb + solve(s) %*% t(RR) %*% solve(RR %*% solve(s) %*% 
            t(RR)) %*% (r - RR %*% bb)
        I <- diag(NCOL(x))
        fd <- solve(s + I) %*% (s + d1 * I)
        brs <- fd %*% rb
        colnames(brs) <- c("Estimate")
        ev <- (t(y) %*% y - t(bb) %*% t(x) %*% y)/(NROW(x) - 
            NCOL(x))
        ev <- diag(ev)
        dbd <- ev * fd %*% (solve(s) - solve(s) %*% t(RR) %*% 
            solve(RR %*% solve(s) %*% t(RR)) %*% RR %*% solve(s)) %*% 
            t(fd)
        bibet <- (d1 - 1) * fd %*% solve(s + d1 * I) %*% bb + 
            fd %*% t(RR) %*% solve(RR %*% solve(s) %*% t(RR)) %*% 
                del
        bibets <- bibet %*% t(bibet)
        mse <- dbd + bibets
        mse1 <- sum(diag(mse))
        Standard_error <- sqrt(diag(abs(dbd)))
        rdel <- matrix(del, nrow(RR))
        lenr <- length(RR)
        dlpt <- diag(RR %*% xin %*% t(RR))
        if (lenr == ncol(RR)) 
            ilpt <- sqrt(solve(abs(dlpt)))
        else ilpt <- sqrt(solve(diag(abs(dlpt))))
        upt <- RR %*% brs
        tb <- t(upt)
        t_statistic <- ((tb - t(rdel)) %*% ilpt)/sqrt(ev)
        tst <- t(2L * pt(-abs(t_statistic), df <- (NROW(x) - 
            NCOL(x))))
        pvalue <- c(tst, rep(NA, (NCOL(x) - NROW(RR))))
        mse1 <- sum(diag(mse))
        mse1 <- round(mse1, digits <- 4L)
        names(mse1) <- c("MSE")
        t_statistic <- c(t_statistic, rep(NA, (NCOL(x) - NROW(RR))))
        ans1 <- cbind(brs, Standard_error, t_statistic, pvalue)
        ans <- round(ans1, digits <- 4L)
        anw <- list(`*****Restricted Liu Estimator*****` = ans, 
            `*****Mean square error value*****` = mse1)
        return(anw)
    }
    npt <- rliues(formula, r, R, delt, d1, data, na.action)
    plotrliu <- function(formula, r, R, delt, d, data = NULL, 
        na.action, ...) {
        i <- 0
        arr <- 0
        for (i in 1:NROW(d)) {
            if (d[i] < 0L) 
                d[i] <- 0L
            else d[i] <- d[i]
            if (d[i] > 1L) 
                d[i] <- 1L
            else d[i] <- d[i]
            rlim <- function(formula, r, R, delt, d, data, na.action, 
                ...) {
                cal <- match.call(expand.dots = FALSE)
                mat <- match(c("formula", "data", "na.action"), 
                  names(cal))
                cal <- cal[c(1L, mat)]
                cal[[1L]] <- as.name("model.frame")
                cal <- eval(cal)
                y <- model.response(cal)
                md <- attr(cal, "terms")
                x <- model.matrix(md, cal, contrasts)
                s <- t(x) %*% x
                xin <- solve(s)
                r <- as.matrix(r)
                RC <- matrix(R, NCOL(x))
                RR <- t(RC)
                if (is.matrix(R)) 
                  RR <- R
                else R <- RR
                del <- as.matrix(delt)
                bb <- xin %*% t(x) %*% y
                I <- diag(NCOL(x))
                fd <- solve(s + I) %*% (s + d * I)
                rb <- bb + solve(s) %*% t(RR) %*% solve(RR %*% 
                  solve(s) %*% t(RR)) %*% (r - RR %*% bb)
                brs <- fd %*% rb
                ev <- (t(y) %*% y - t(bb) %*% t(x) %*% y)/(NROW(x) - 
                  NCOL(x))
                ev <- diag(ev)
                dbd <- ev * fd %*% (solve(s) - solve(s) %*% t(RR) %*% 
                  solve(RR %*% solve(s) %*% t(RR)) %*% RR %*% 
                  solve(s)) %*% t(fd)
                bibet <- (d - 1) * fd %*% solve(s + d * I) %*% 
                  bb + fd %*% t(RR) %*% solve(RR %*% solve(s) %*% 
                  t(RR)) %*% del
                bibets <- bibet %*% t(bibet)
                mse1 <- dbd + bibets
                mse <- sum(diag(mse1))
                return(mse)
            }
            arr[i] <- rlim(formula, r, R, delt, d[i], data, na.action)
        }
        MSE <- arr
        parameter <- d
        pvl <- cbind(parameter, MSE)
        colnames(pvl) <- c("Parameter", "MSE")
        sval <- pvl
        return(sval)
    }
    prliu <- plotrliu(formula, r, R, delt, d, data, na.action)
    if (nrow(d) > 1L) 
        val <- prliu
    else val <- npt
    val
}