Ndtyjky

I have the following code in R but I have problems in the output, it should display something different. This is what shows

Summary(x, y)



The total square sum is: 17.5The error square sum is: 0 

[[1]]

NULL



[[2]]

[1] "n"



[[3]]

NULL

It was supposed to display

The total square sum is: number1



The error square sum is: number2

Could you please check it?

(This is somehow a sample, in reality I have to display more things the standar error is: number3, the variance is number4, etc..)

Summary <- function(x, y, print=TRUE) {

      p <- 2

      n <- length(x)



      x <- matrix(c(rep(1,n),x),n,p)

      bg <- solve(t(x)%*%x,t(x)%*%y)

      invx <- solve(t(x)%*%x)

      xty <- t(x)%*%y

      e <- y-x%*%bg

      SCT <- sum(y^2)-n*(mean(y)^2)

      SCE <- sum(e*e)



      result <- list(

        cat("The total square sum is:", SCT), 

        "n", 

        cat("The error square sum is:", SCE, "n"))

      return(result)      

}



x <- y <- 1:6

Summary(x, y)

list() creates a list from objects, but cat() does not return an object, it just prints to the console. That is why two of the list elements says NULL (they are empty, while one contains the character string "n" (an actual object).

Printing text with more involved formatting can be difficult and non-intuitive, but I find that much inspiration and help can be found in existing R code.

Take print.lm() for example, which is the function responsible for displaying the result from linear regressions using lm().

Run stats:::print.lm and you will see:

function (x, digits = max(3L, getOption("digits") - 3L), ...) 

{

    cat("nCall:n", paste(deparse(x$call), sep = "n", collapse = "n"), 

        "nn", sep = "")

    if (length(coef(x))) {

        cat("Coefficients:n")

        print.default(format(coef(x), digits = digits), print.gap = 2L, 

            quote = FALSE)

    }

    else cat("No coefficientsn")

    cat("n")

    invisible(x)

}

Looks a bit busy, but it's not too bad. You'll see that each call to cat() contains one or more character strings and delimiters (like n and t for new line and tabulate) laid out in order, with sep, the separator, specified at the end. Sometimes there is a paste() call inside cat(), in which case paste() merely 'prepares' some characters for cat() to print. We also note that there are several calls to cat() and print(), and mixing and matching is done, without issue. And at the very end is an example of invisible() as MrFlick mentioned in a comment. This command makes sure that the function won't print its argument (in this case x), but you can still assign it to a variable.

With these insights, can we improve on Summary()?

Summary  <-  function(x, y, print=TRUE) {

    p  <-  2

    n <- length(x)



    xm <- matrix(c(rep(1,n),x),n,p)

    bg <- solve(t(xm)%*%xm,t(xm)%*%y)

    invx <- solve(t(xm)%*%xm)

    xty <- t(xm)%*%y

    e <- y-xm%*%bg

    SCT <- sum(y^2)-n*(mean(y)^2)

    SCE <- sum(e*e)



    results <- list(TSS=SCT, ESS=SCE, p=p, x=x, y=y)



    if (SCE == 0) warning("Error square sum is zero", call.=FALSE)



    if (print) {

        cat("Results for the variables", "nt",

            deparse(match.call()$x), " and ", deparse(match.call()$y),

            "nn", sep="")

        cat("The total square sum is: ", SCT, "nn",

            "The error square sum is: ", SCE, "nn", sep="")

        invisible(results)

    } else {

        results

    }      

}

Looks a bit more involved. Lets test it.

Wendy <- Carlos <- 1:6



Summary(x=Wendy, y=Carlos)



Results for the variables

    Wendy and Carlos



The total square sum is: 17.5



The error square sum is: 0



Warning message:

Error square sum is zero