I've applied Wright's rank and sign test. I got R1,R2 but R3 resulted in NaN

[ShahadMH]

I've applied Wright's rank and sign test using Rstudio. I got R1,R2 but R3 resulted in NaN.. what is the reason of such a result? and how to get the sign result?
I am using the "vrtest" package.

Thank you.

[HanOostdijk]

Hello @ShahadMH

the documentation of the function does not mention R3 (however it mentions S1).
The example uses data(exrates) but executing that gives me an error:
probably the data belongs to a package that is not installed on my computer.
So no further help from me, I am afraid.

EDIT: forgot to do library(vrtest)
The data is in the vrtest package itself.

The documentation mentions S1 but the output indicates R3 ??

Wright {vrtest}	R Documentation
Wright's Rank and Sign Tests
Description
The function returns R1, R2 and S1 tests statistics detailed in Wright (2000)

> data(exrates)
> y <- exrates$ca                    
> nob <- length(y)
> r <- log(y[2:nob])-log(y[1:(nob-1)])  
> kvec <- c(2,5,10)
> Wright(r,kvec) 
$Stats
           R1       R2       R3
k=2  2.483482 2.511358 2.015755
k=5  2.847969 2.106418 3.431291
k=10 2.188621 1.105754 3.606674

[HanOostdijk]

Viewed the code, but it did not help me much.
Maybe depending on the function argument for Wright that you specify.
vrtest functions below:

> Wright 
function (y, kvec) 
{
    y <- as.matrix(y)
    n <- nrow(y)
    W_mat <- matrix(NA, nrow = length(kvec), ncol = 3)
    for (i in 1:length(kvec)) {
        k <- kvec[i]
        W <- Wright_stat(y, k)
        W_mat[i, ] <- cbind(W$WR1, W$WR2, W$WS1)
    }
    VR <- W_mat
    rownames(VR) <- paste("k=", kvec, sep = "")
    colnames(VR) <- c("R1", "R2", "R3")
    return(list(Stats = VR))
}
<bytecode: 0x0000024df212f038>
<environment: namespace:vrtest>
> vrtest:::Wright_stat 
function (y, k) 
{
    y <- as.matrix(y)
    n <- nrow(y)
    ranking <- as.matrix(rank(y))
    r1 <- (ranking - 0.5 * (n + 1))/sqrt((n - 1) * (n + 1)/12)
    r2 <- qnorm(ranking/(n + 1))
    s <- sign(y)
    s[s == 0] <- -1
    R1 <- stat(r1, k)
    R2 <- stat(r2, k)
    S1 <- stat(s, k)
    return(list(WR1 = R1, WR2 = R2, WS1 = S1))
}
<bytecode: 0x0000024df2127498>
<environment: namespace:vrtest>
> vrtest:::stat 
function (x, k) 
{
    x <- as.matrix(x)
    n <- nrow(x)
    index <- 1:k
    summ <- 0
    for (i in k:n) {
        summ <- summ + sum(x[index])^2
        index <- index + 1
    }
    vr1 <- sum(x^2)/n
    vr2 <- summ/(n * k)
    vr <- vr2/vr1
    tem1 <- 2 * (2 * k - 1) * (k - 1)
    tem2 <- 3 * k * n
    vrstat <- (vr - 1)/sqrt(tem1/tem2)
}
<bytecode: 0x0000024df211b770>
<environment: namespace:vrtest>
>