Adjusting Yaxis (ylim) limits on a plotMeans(DV, IV1, IV2, error.bars="se")

[unl-neuropsych]

Hello everyone,

I have tried to look for this everywhere and so far have no luck.  I have a plotMeans(DV, IV1, IV2, error.bars="se") graph that plots my data (DV-continuous, IVs are factors, IV1 - two levels, IV2-four levels).  I am trying to increase a scale of my y-axis (to be consistent with my other graphs), but unfortunately nothing works with "plotMeans" function, which is the only one I see so fat to plot what I need.  Could anyone provide any suggestions or advice?

Thank you.

Sergey (UNL, Behavioral Psychopharmacology Lab)

[Steven McKinney]

Hi Sergey

I've attached a script with a ylim argument added to plotMeans.

If you are using Rcmdr, you can load this script via the Rcmdr File menu item
File
  Open Script File

so save this plotMeans.R script somewhere,
and load it and run it in Rcmdr, then
do your plot command
 plotMeans(DV, IV1, IV2, error.bars="se", ylim = c(27, 99))
or whatever your desired ylim values are.

HTH
 

Steven McKinney, Ph.D.

Statistician
Molecular Oncology and Breast Cancer Program
British Columbia Cancer Research Centre

email: smckinney +at+ bccrc +dot+ ca

tel: [phone] x7561

BCCRC
Molecular Oncology
675 West 10th Ave, Floor 4
Vancouver B.C.
V5Z 1L3
Canada

________________________________________
From: r-help-bounces@... [r-help-bounces@...] On Behalf Of Sergios (Sergey) Charntikov [sergioschr@...]
Sent: November 5, 2009 7:51 PM
To: r-help@...
Subject: [R] Adjusting Yaxis (ylim) limits on a plotMeans(DV, IV1, IV2, error.bars="se")

Hello everyone,

I have tried to look for this everywhere and so far have no luck.  I have a
plotMeans(DV, IV1, IV2, error.bars="se") graph that plots my data
(DV-continuous, IVs are factors, IV1 - two levels, IV2-four levels).  I am
trying to increase a scale of my y-axis (to be consistent with my other
graphs), but unfortunately nothing works with "plotMeans" function, which is
the only one I see so fat to plot what I need.  Could anyone provide any
suggestions or advice?

Thank you.

Sergey (UNL, Behavioral Psychopharmacology Lab)

        [[alternative HTML version deleted]]

______________________________________________
R-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.


plotMeans <-
function (response, factor1, factor2, error.bars = c("se", "sd",
    "conf.int", "none"), level = 0.95, xlab = deparse(substitute(factor1)),
    ylab = paste("mean of", deparse(substitute(response))), legend.lab = deparse(substitute(factor2)),
    main = "Plot of Means", pch = 1:n.levs.2, lty = 1:n.levs.2,
    col = palette(), ylim = NULL)
{
    if (!is.numeric(response))
        stop(gettextRcmdr("Argument response must be numeric."))
    xlab
    ylab
    legend.lab
    error.bars <- match.arg(error.bars)
    if (missing(factor2)) {
        if (!is.factor(factor1))
            stop(gettextRcmdr("Argument factor1 must be a factor."))
        valid <- complete.cases(factor1, response)
        factor1 <- factor1[valid]
        response <- response[valid]
        means <- tapply(response, factor1, mean)
        sds <- tapply(response, factor1, sd)
        ns <- tapply(response, factor1, length)
        if (error.bars == "se")
            sds <- sds/sqrt(ns)
        if (error.bars == "conf.int")
            sds <- qt((1 - level)/2, df = ns - 1, lower.tail = FALSE) *
                sds/sqrt(ns)
        sds[is.na(sds)] <- 0
        yrange <- if (error.bars != "none")
            c(min(means - sds, na.rm = TRUE), max(means + sds,
                na.rm = TRUE))
        else range(means, na.rm = TRUE)
        levs <- levels(factor1)
        n.levs <- length(levs)
        if ( is.null(ylim) ) {
          plot(c(1, n.levs), yrange, type = "n", xlab = xlab, ylab = ylab,
               axes = FALSE, main = main)
        } else {
          plot(c(1, n.levs), yrange, type = "n", xlab = xlab, ylab = ylab,
               axes = FALSE, main = main, ylim = ylim)
        }        
        points(1:n.levs, means, type = "b", pch = 16, cex = 2)
        box()
        axis(2)
        axis(1, at = 1:n.levs, labels = levs)
        if (error.bars != "none")
            arrows(1:n.levs, means - sds, 1:n.levs, means + sds,
                angle = 90, lty = 2, code = 3, length = 0.125)
    }
    else {
        if (!(is.factor(factor1) | is.factor(factor2)))
            stop(gettextRcmdr("Arguments factor1 and factor2 must be factors."))
        valid <- complete.cases(factor1, factor2, response)
        factor1 <- factor1[valid]
        factor2 <- factor2[valid]
        response <- response[valid]
        means <- tapply(response, list(factor1, factor2), mean)
        sds <- tapply(response, list(factor1, factor2), sd)
        ns <- tapply(response, list(factor1, factor2), length)
        if (error.bars == "se")
            sds <- sds/sqrt(ns)
        if (error.bars == "conf.int")
            sds <- qt((1 - level)/2, df = ns - 1, lower.tail = FALSE) *
                sds/sqrt(ns)
        sds[is.na(sds)] <- 0
        yrange <- if (error.bars != "none")
            c(min(means - sds, na.rm = TRUE), max(means + sds,
                na.rm = TRUE))
        else range(means, na.rm = TRUE)
        levs.1 <- levels(factor1)
        levs.2 <- levels(factor2)
        n.levs.1 <- length(levs.1)
        n.levs.2 <- length(levs.2)
        if (length(pch) == 1)
            pch <- rep(pch, n.levs.2)
        if (length(col) == 1)
            col <- rep(col, n.levs.2)
        if (length(lty) == 1)
            lty <- rep(lty, n.levs.2)
        if (n.levs.2 > length(col))
            stop(sprintf(gettextRcmdr("Number of groups for factor2, %d, exceeds number of distinct colours, %d."),
                n.levs.2, length(col)))
        if ( is.null(ylim) ) {
          plot(c(1, n.levs.1 * 1.4), yrange, type = "n", xlab = xlab,
               ylab = ylab, axes = FALSE, main = main)
        } else {
          plot(c(1, n.levs.1 * 1.4), yrange, type = "n", xlab = xlab,
               ylab = ylab, axes = FALSE, main = main, ylim = ylim)
        }
        box()
        axis(2)
        axis(1, at = 1:n.levs.1, labels = levs.1)
        for (i in 1:n.levs.2) {
            points(1:n.levs.1, means[, i], type = "b", pch = pch[i],
                cex = 2, col = col[i], lty = lty[i])
            if (error.bars != "none")
                arrows(1:n.levs.1, means[, i] - sds[, i], 1:n.levs.1,
                  means[, i] + sds[, i], angle = 90, code = 3,
                  col = col[i], lty = lty[i], length = 0.125)
        }
        x.posn <- n.levs.1 * 1.1
        y.posn <- sum(c(0.1, 0.9) * par("usr")[c(3, 4)])
        text(x.posn, y.posn, legend.lab, adj = c(0, -0.5))
        legend(x.posn, y.posn, levs.2, pch = pch, col = col,
            lty = lty)
    }
    invisible(NULL)
}

______________________________________________
R-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[unl-neuropsych]

Thank you very much Steven,

This works like a charm.  Thank you, Thank you.


Sincerely,

Sergios Charntikov (Sergey), MA

Behavioral Neuropharmacology Lab
Department of Psychology
University of Nebraska-Lincoln
Lincoln, NE 68588-0308  USA




On Thu, Nov 5, 2009 at 11:27 PM, Steven McKinney <smckinney@...> wrote:

> Hi Sergey
>
> I've attached a script with a ylim argument added to plotMeans.
>
> If you are using Rcmdr, you can load this script via the Rcmdr File menu
> item
> File
>  Open Script File
>
> so save this plotMeans.R script somewhere,
> and load it and run it in Rcmdr, then
> do your plot command
>  plotMeans(DV, IV1, IV2, error.bars="se", ylim = c(27, 99))
> or whatever your desired ylim values are.
>
> HTH
>
>
> Steven McKinney, Ph.D.
>
> Statistician
> Molecular Oncology and Breast Cancer Program
> British Columbia Cancer Research Centre
>
> email: smckinney +at+ bccrc +dot+ ca
>
> tel: [phone] x7561
>
> BCCRC
> Molecular Oncology
> 675 West 10th Ave, Floor 4
> Vancouver B.C.
> V5Z 1L3
> Canada
>
> ________________________________________
> From: r-help-bounces@... [r-help-bounces@...] On
> Behalf Of Sergios (Sergey) Charntikov [sergioschr@...]
> Sent: November 5, 2009 7:51 PM
> To: r-help@...
> Subject: [R] Adjusting Yaxis (ylim) limits on a plotMeans(DV, IV1, IV2,
> error.bars="se")
>
> Hello everyone,
>
> I have tried to look for this everywhere and so far have no luck.  I have a
> plotMeans(DV, IV1, IV2, error.bars="se") graph that plots my data
> (DV-continuous, IVs are factors, IV1 - two levels, IV2-four levels).  I am
> trying to increase a scale of my y-axis (to be consistent with my other
> graphs), but unfortunately nothing works with "plotMeans" function, which
> is
> the only one I see so fat to plot what I need.  Could anyone provide any
> suggestions or advice?
>
> Thank you.
>
> Sergey (UNL, Behavioral Psychopharmacology Lab)
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@... mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

        [[alternative HTML version deleted]]

______________________________________________
R-help@... mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.