TatsukiRcode < Main < Vanderbilt Biostatistics Wiki

You are here: Vanderbilt Biostatistics Wiki>Main Web>TatsukiRcode (01 Aug 2022, TatsukiKoyama) (raw view)EditAttach

<!-- %TOC% -->

---+ R code for tplot, kmplot and other functions.
   * [[%ATTACHURL%/RFunctions1.R][RFunctions1.R]]: RFunctions1.R
---+ R code for other miscellaneous functions.
   * [[%ATTACHURL%/RFunctions2.R][RFunctions2.R]]: RFunctions2.R
---+ R code for simulating 3+3 designs.
   * [[%ATTACHURL%/Three3.R][Three3.R]]: Three3.R

---+ R code for inference from two-stage adaptive design
   * Koyama T and Chen H. ``Proper inference from Simon's two-stage designs'' <i>Statistics in Medicine<i> <b>27</b>(16), 2008. PMID: 17960777. PMCID: PMC6047527.
   * [[https://biostat.app.vumc.org/wiki/Main/TwoStageInference][The main page]]
   * [[%ATTACHURL%/twostage2020Web.R][twostage2020Web.R]]: R code: Inference from two-stage designs

---+ R code for confidence intervals on micro-averaged F1 and macro-averaged F1 scores
   * %PMIDC{35317080}%
   * [[%ATTACHURL%/f1Scores.R][f1Scores.R]]: R code: Confidence intervals for F1 scores

---+ R code for the Length of the Beatles' Songs plot
   * LengthOfTheBeatlesSongs

---------
[[#RcodeKmplot][kmplot()]]

[[#RcodeTplot][tplot()]]

[[#RcodeJmplot][jmplot()]]

[[#RcodeDsplot][dsplot()]]

[[#RcodeShowColors][showcolors()]]
---------

#RcodeKmplot
---++ kmplot()
   * Kaplan-Meier survival plot – with at-risk table
   * plot Kaplan-Meier curves with 'numbers at risk' at bottom margin.
      * can specify at which time n at-risk is displayed
      * can specify line type / line width / line color for each line
      * can change the order of appearance in the n at-risk table
      * n.at.risk numbers are right-justified!
      * simple=TRUE suppresses the 'at risk' table.

---++++ Example
<highlight>
require(survival)

kma <- survfit( Surv(time, status) ~ rx + adhere, data=colon )

kmplot(kma, mark='', simple=FALSE,
	xaxis.at=c(0,.5,1:9)*365, xaxis.lab=c(0,.5,1:9), # n.risk.at
	lty.surv=c(1,2), lwd.surv=1, col.surv=c(1,1,2,2,4,4), # survival.curves
	col.ci=0, # confidence intervals not plotted
	group.names=c('Obs ','Obs tumor adh','Lev','Lev tumor adh','Lev+5FU ','Lev+5FU tumor adh'),
	group.order=c(5,3,1,6,4,2), # order of appearance in the n.risk.at table and legend.
	extra.left.margin=6, label.n.at.risk=FALSE, draw.lines=TRUE, 
	cex.axis=.8, xlab='Years', ylab='Survival Probability', # labels
	grid=TRUE, lty.grid=1, lwd.grid=1, col.grid=grey(.9), 
	legend=TRUE, loc.legend='bottomleft',
	cex.lab=.8, xaxs='r', bty='L', las=1, tcl=-.2  # other parameters passed to plot()
) 
title(main='Chemotherapy for stage B/C colon cancer', adj=.1, font.main=1, line=0.5, cex.main=1)
</highlight>

     <img src="%ATTACHURLPATH%/kmplotExample.png" alt="kmplotExample.png" width='750' height='600' />



#RcodeTplot
---++ tplot()
tplot() is an alternative to boxplot().  The individual data can be shown (either in the foreground or background) with jittering if necessary. 

---++++ Example

This example shows what options are available with tplot().  This may not be a good way to show these data (unnecessary color varieties...)<br>col and pch may be specified at the individual level or group level. (In this example, <i>col</i> is specified at the individual level, and _pch_ is specified for groups.)<br>colors of boxes and box borders are specified by _boxcol_ and _boxborder_.<br><i>boxplot.pars</i> is passed to boxplot().

<highlight>
set.seed(100)
age <- rnorm(80, rep(c(26,36), c(70,10)), 4)
sex <- factor( sample(c('Female','Male'), 80, TRUE) )
group <- paste('Group ', sample(1:4, 40, prob=c(2,5,4,1), replace=TRUE), sep='')
d <- data.frame(age, sex, group)

tplot( age ~ group, data=d, 
    las=1, cex=1, cex.axis=1, bty='L', 
    show.n=TRUE, dist=.25, jit=.05, type=c('db','db','db','d'),
    group.pch=TRUE, pch=c(15,17,19,8), 
    col=c('darkred','darkblue')[c(sex)],
    boxcol=c('lightsteelblue1','lightyellow1',grey(.9), 0), boxborder=grey(.8),
    boxplot.pars=list(notch=TRUE, boxwex=.5)
    )
</highlight>

<img src="%ATTACHURLPATH%/tplotExample.png" alt="Example of tplot"  width="600" height="600"  />


#RcodeJmplot
---++ jmplot()
   * Joint distribution and marginal distributions on its margins. 
   * requires tplot() [See above.]  

---++++ Example
<highlight>
 set.seed(274)
 x <- rnorm(100, 20, 5)
 y <- rexp(100)
 levels <- as.factor(sample(c("Male","Female"), 100, TRUE))
 jmplot(x, y, levels, col=levels, las=1, type='db', main='Some scores', jit=.02)
</highlight>

     <img src="%ATTACHURLPATH%/jmplotExample1.png" alt="jmplotExample1.png" width='829' height='653' />


#RcodeDsplot
---++ dsplot()
   * Discrete Scatter Plot for Integer-valued variables
   * This creates a scatter plot (sort of) for discrete bivariate data.  An alternative to sunflower plots.

---++++ Example
<highlight>
set.seed(6)
x <- round(rnorm(400, 100, 4))
y <- round(rnorm(400, 200, 4))
sex <- sample(c('Female','Male'), 400, T)
dsplot(y ~ x, pch=19, col=1+(sex=='Female'), cex=.6, bkgr=T, 
   xlab='measurement 1', ylab='measurement 2', bty='L')
legend( 'bottomright', pch=19, col=1:2, legend=c('Male','Female'), cex=.8 )
</highlight>

<img src="%ATTACHURLPATH%/dspExample.png" alt="showColors.png"  width="670" height="670"  />


#RcodeShowColors
---++ Show colors
<p>In R, there are 657 named colors.  The following funciton, show.colors(), shows these colors and corresponding numbers.<br>Find the color you like in the plot.  You can find the name of the color by running >colors()[number].</p>
<highlight>
show.colors <- function(){
   par(mfrow=c(1,1))
   #   par(mai=c(.4,.4,.4,.4), oma=c(.2,0,0,.2))
   x <- 22 ; y <- 30
   plot(c(-1,x),c(-1,y), xlab='', ylab='', type='n', xaxt='n', yaxt='n', bty='n')
   for(i in 1:x){for(j in 1:y){
         k <- y*(i-1)+j ; co <- colors()[k]
         rect(i-1, j-1, i, j, col=co, border=grey(.5) )}}

   text(rep(-.5,y),(1:y)-.5, 1:y, cex=1.2-.016*y)
   text((1:x)-.5, rep(-.5,x), y*(0:(x-1)), cex=1.2-.022*x)
   title('col=colors()[n]')
}

> colors()[465]
[1] "mediumorchid3"
</highlight>

    <img src="%ATTACHURLPATH%/showColors.png" alt="showColors.png"  width="689" height="690"  />

<!--   * [[%ATTACHURL%/TatsukiRcodeFormatSum.R][R Code Format Summary]]
   * [[%ATTACHURL%/TatsukiRcodeMisc.R][R Code Miscellaneous]]
   * [[%ATTACHURL%/RFunctions0.R][RFunctions0.R]]: RFunctions0.R
-->

Topic attachments
I	Attachment	Action	Size	Date	Who	Comment
csv	Beatlesongs072108.csv	manage	10.3 K	10 Oct 2008 - 17:18	TatsukiKoyama
pdf	BeatlesongsPoster.pdf	manage	66.8 K	10 Oct 2008 - 17:11	TatsukiKoyama
pdf	BeatlesongsPoster2.pdf	manage	66.9 K	10 Oct 2008 - 18:05	TatsukiKoyama
pdf	Poster3.pdf	manage	69.2 K	22 Nov 2016 - 16:31	TatsukiKoyama
R	RFunctions0.R	manage	27.0 K	19 May 2017 - 12:09	TatsukiKoyama
R	RFunctions071817.R	manage	11.1 K	18 Jul 2017 - 03:07	TatsukiKoyama
R	RFunctions1.R	manage	28.5 K	29 Oct 2020 - 15:45	TatsukiKoyama
R	RFunctions2.R	manage	16.3 K	29 Oct 2020 - 15:44	TatsukiKoyama
R	RFunctions3.R	manage	12.8 K	29 Oct 2020 - 15:45	TatsukiKoyama
R	TatsukiRcodeFormatSum.R	manage	3.4 K	11 May 2017 - 00:05	TatsukiKoyama
R	TatsukiRcodeMisc.R	manage	5.7 K	11 May 2017 - 00:05	TatsukiKoyama
R	TextBox.R	manage	13.2 K	11 Apr 2014 - 11:49	TatsukiKoyama
r	TextBoxExample.r	manage	2.2 K	10 Oct 2008 - 16:33	TatsukiKoyama
R	Three3.R	manage	2.7 K	11 Feb 2020 - 11:32	TatsukiKoyama
R	USEFUL.R	manage	32.7 K	11 May 2017 - 00:05	TatsukiKoyama
r	USEFUL.r	manage	32.7 K	11 Nov 2008 - 09:20	TatsukiKoyama
csv	ap21.csv	manage	10.9 K	03 Aug 2011 - 12:11	TatsukiKoyama	needed in computation on 3+3 designs: http://data.vanderbilt.edu/~koyamat/brew/t33.html
csv	ap31.csv	manage	60.9 K	03 Aug 2011 - 12:11	TatsukiKoyama
csv	ap32.csv	manage	90.8 K	03 Aug 2011 - 12:12	TatsukiKoyama
csv	ap41.csv	manage	264.1 K	03 Aug 2011 - 12:13	TatsukiKoyama
csv	ap42.csv	manage	325.0 K	03 Aug 2011 - 12:13	TatsukiKoyama
csv	ap51.csv	manage	1017.9 K	03 Aug 2011 - 12:13	TatsukiKoyama
csv	ap52.csv	manage	1122.6 K	03 Aug 2011 - 12:13	TatsukiKoyama
csv	ap53.csv	manage	1996.2 K	03 Aug 2011 - 12:13	TatsukiKoyama
csv	ap61.csv	manage	3622.9 K	03 Aug 2011 - 12:13	TatsukiKoyama
csv	ap62.csv	manage	3713.3 K	03 Aug 2011 - 12:14	TatsukiKoyama
csv	ap63.csv	manage	5637.9 K	03 Aug 2011 - 12:14	TatsukiKoyama
R	beatlesongs2012.R	manage	14.2 K	14 May 2020 - 09:58	TatsukiKoyama
csv	beatlesongs2012.csv	manage	15.4 K	03 Mar 2012 - 20:48	TatsukiKoyama
png	dspExample.png	manage	7.1 K	06 Jan 2011 - 14:56	TatsukiKoyama
R	f1Scores.R	manage	3.4 K	01 Aug 2022 - 15:13	TatsukiKoyama
R	f1score.R	manage	3.1 K	01 Aug 2022 - 14:49	TatsukiKoyama
png	jmplotExample1.png	manage	8.9 K	07 Dec 2011 - 16:48	TatsukiKoyama
png	kmplotExample.png	manage	11.6 K	17 Dec 2011 - 16:41	TatsukiKoyama
css	koyama.css	manage	0.3 K	03 Jan 2012 - 13:26	TatsukiKoyama
png	tplotExample.png	manage	6.1 K	07 Dec 2011 - 13:04	TatsukiKoyama
R	twostage2018Web.R	manage	0.1 K	16 Jun 2020 - 13:17	TatsukiKoyama
R	twostage2020Web.R	manage	5.3 K	05 Aug 2021 - 12:04	TatsukiKoyama

Topic revision: r166 - 01 Aug 2022, TatsukiKoyama

Main

Department Home Page

Biostatistics Graduate Program

Vanderbilt University Medical Center

Biostatistics Webs
- Archive
- Main
- Sandbox
- System

Copyright © 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback