Movatterモバイル変換

Adjust Color in HSV Space

Description

Adjust hue, saturation, value, and/or opacity.

Usage

adjust_hsv(col, h=NULL, s=NULL, v=NULL, alpha=NULL)

Arguments

Details

Colors can be specified as a color name, a hexadecimal string, or an integer.

Hue, saturation, value, and opacity are specified as values from 0 to 1, orNULL to leave unchanged.

Value

Adjusted colors in hexadecimal string format.

Author(s)

Arni Magnusson.

See Also

adjustcolor adjusts colors in RGB space.

col2rgb,rgb2hsv, andhsv are theunderlying functions used to convert and adjust the colors.

Examples

col <- "#123456"col2 <- adjust_hsv(col, h=0.1)col3 <- adjust_hsv(col, s=0.1)col4 <- adjust_hsv(col, v=0.7)barplot(rep(1, 4), col=c(col, col2, col3, col4),        names=c("original", "h", "s", "v"))

Add a Axis to a Plot with Rotated Labels

Description

Add a labeled axis to the current plot with rotated text

Usage

angleAxis(side, labels, at = 1:length(labels), srt = 45, adj, xpd = TRUE, ...)

Arguments

Details

This function augments the feature of theaxis functon by allowing the axis labels to be rotated.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

axis

Examples

# create a vector with some values and long labelsvalues <- sample(1:10)names(values) <- sapply(letters[1:10],                         function(x) paste(rep(x, 10), sep="",collapse="")                         ) # barplot labels are too long for the available space, hence some are not plottedbarplot(values)# to add angled labels, tell barplot not to label the x axis, and store the bar locationat <- barplot(values, xaxt="n")# then use angleAxsangleAxis(1, at=at, labels = names(values))# angle counter-clockwise insteadat <- barplot(values, xaxt="n")angleAxis(1, at=at, labels = names(values), srt=-45, adj=0)# put labels at the topoldpar <- par()$marpar(mar=c(1,4,5,2)+0.1)at <- barplot(values, xaxt="n")angleAxis(3, at=at, labels = names(values))par(oldpar)# put labels on the leftoldpar <- par()$marpar(mar=c(5,5,3,2)+0.1)at <- barplot(values, yaxt="n", horiz=TRUE)angleAxis(2, at=at, labels = names(values))par(oldpar)# put labels on the rightoldpar <- par()$marpar(mar=c(2,5,3,5)+0.1)at <- barplot(values, yaxt="n", horiz=TRUE)angleAxis(4, at=at, labels = names(values))par(oldpar)# specify colors for bars and labelsat <- barplot(values, xaxt="n", col=1:10)angleAxis(1, at=at, labels = names(values), col=1:10)

Plot a graphical matrix where each cell contains a dot whose sizereflects the relative magnitude of the corresponding component.

Description

Plot a graphical matrix where each cell contains a dot whose sizereflects the relative magnitude of the corresponding component.

Usage

balloonplot(x, ...)## S3 method for class 'table'balloonplot(x, xlab, ylab, zlab, show.zeros=FALSE,show.margins=TRUE,...)## Default S3 method:balloonplot(x,y,z,                                xlab,                                ylab,                                zlab=deparse(substitute(z)),                                dotsize=2/max(strwidth(19),strheight(19)),                                dotchar=19,                                dotcolor="skyblue",                                text.size=1,                                text.color=par("fg"),                                main,                                label=TRUE,                                label.digits=2,                                label.size=1,                                label.color=par("fg"),                                scale.method=c("volume","diameter"),                                scale.range=c("absolute","relative"),                                colsrt=par("srt"),                                rowsrt=par("srt"),                                colmar=1,                                rowmar=2,                                show.zeros=FALSE,                                show.margins=TRUE,                                cum.margins=TRUE,                                sorted=TRUE,                                label.lines=TRUE,                                fun=function(x)sum(x,na.rm=T),                                hide.duplicates=TRUE,                                ... )

Arguments

Details

This function plots a visual matrix. In eachx,y cell adot is plotted which reflects the relative size of the correspondingvalue ofz. Whenscale.method="volume" the volume ofthe dot is proportional to the relative size ofz. Whenscale.method="diameter", the diameter of the dot is proportional tothe the relative size ofz. The "volume" method is defaultbecause the "diameter" method visually exaggerates differences.

Value

Nothing of interest.

Note

z is expected to be non-negative. The function will stilloperate correctly if there are negative values ofz, but thecorresponding dots will have 0 size and a warning will be generated.

Author(s)

Gregory R. Warnesgreg@warnes.net

References

Function inspired by question posed on R-help by RamonAlonso-Allendeallende@cnb.uam.es.

See Also

plot.table.

bubbleplot provides an alternative interface and visualstyle based on scatterplots instead of tables.

Examples

# Create an Example Data Frame Containing Car x Color datacarnames <- c("bmw","renault","mercedes","seat")carcolors <- c("red","white","silver","green")datavals <- round(rnorm(16, mean=100, sd=60),1)data <- data.frame(Car=rep(carnames,4),                   Color=rep(carcolors, c(4,4,4,4) ),                   Value=datavals )# show the datadata# generate balloon plot with default scalingballoonplot( data$Car, data$Color, data$Value)# show margin label rotation & space expansion, using some long labelslevels(data$Car) <- c("BMW: High End, German","Renault: Medium End, French", "Mercedes: High End, German", "Seat: Imaginary, Unknown Producer")# generate balloon plot with default scalingballoonplot( data$Car, data$Color, data$Value, colmar=3, colsrt=90)# Create an example using tablexnames <- sample( letters[1:3], 50, replace=2)ynames <- sample( 1:5, 50, replace=2)tab <- table(xnames, ynames)balloonplot(tab)# Example of multiple classification variabls using the Titanic datalibrary(datasets)data(Titanic)dframe <- as.data.frame(Titanic) # convert to 1 entry per row formatattach(dframe)balloonplot(x=Class, y=list(Survived, Age, Sex), z=Freq, sort=TRUE)# colorize: surviors lightblue, non-survivors: greyColors <- TitanicColors[,,,"Yes"] <- "skyblue"Colors[,,,"No"] <- "grey"colors <- as.character(as.data.frame(Colors)$Freq)balloonplot(x=list(Age,Sex),            y=list(Class=Class,                   Survived=reorder.factor(Survived,new.order=c(2,1))                   ),            z=Freq,            zlab="Number of Passengers",            sort=TRUE,            dotcol = colors,            show.zeros=TRUE,            show.margins=TRUE)

Plot x-y Points with Locally Smoothed Mean and Standard Deviation

Description

Plot x-y points with curves for locally smoothed mean and standard deviation.

Usage

bandplot(x,...)## S3 method for class 'formula'bandplot(x, data, subset, na.action, ...,           xlab=NULL, ylab=NULL, add = FALSE, sd = c(-2:2),           sd.col=c("magenta", "blue", "red", "blue", "magenta"),           sd.lwd=c(2, 2, 3, 2, 2),  sd.lty=c(2, 1, 1, 1, 2),           method = "frac", width = 1/5, n=50)## Default S3 method:bandplot(x, y, ..., add = FALSE, sd = c(-2:2),           sd.col=c("magenta", "blue", "red", "blue", "magenta"),           sd.lwd=c(2, 2, 3, 2, 2),  sd.lty=c(2, 1, 1, 1, 2),           method = "frac", width = 1/5, n=50)

Arguments

Details

bandplot was created to look for changes in the mean orvariance of scatter plots, particularly plots of regression residuals.

The local mean and standard deviation are calculated by calling'wapply'. By default, bandplot asks wapply to smooth using intervalsthat include the nearest 1/5 of the data. See the documentation ofthat function for details on the algorithm.

Value

Invisibly returns a list containing the x,y points plotted for each line.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

wapply,lowess

Examples

# fixed mean, changing variancex <- 1:1000y <- rnorm(1000, mean=1, sd=1 + x/1000 )bandplot(x,y)bandplot(y~x)# fixed varance, changing meanx <- 1:1000y <- rnorm(1000, mean=x/1000, sd=1)bandplot(x,y)## changing mean and variance#x <- abs(rnorm(500))y <- rnorm(500, mean=2*x, sd=2+2*x)# the changing mean and dispersion are hard to see whith the points alone:plot(x,y )# regression picks up the mean trend, but not the change in variancereg <- lm(y~x)summary(reg)abline(reg=reg, col="blue", lwd=2)# using bandplot on the original data helps to show the mean and# variance trendbandplot(y ~ x)# using bandplot on the residuals helps to see that regression removes# the mean trend but leaves the trend in variabilitybandplot(predict(reg),resid(reg))

Enhanced Bar Plots

Description

An enhancement of the standard barplot() function. Creates a bar plotwith vertical or horizontal bars. Can plot confidence intervals for eachbar, a lined grid behind the bars, change plot area color andlogarithmic axes may be used.

Usage

## Default S3 method:barplot2(height, width = 1, space = NULL,        names.arg = NULL, legend.text = NULL, beside = FALSE,        horiz = FALSE, density = NULL, angle = 45,        col = NULL, prcol = NULL, border = par("fg"),        main = NULL, sub = NULL, xlab = NULL, ylab = NULL,        xlim = NULL, ylim = NULL, xpd = TRUE, log = "",        axes = TRUE, axisnames = TRUE,        cex.axis = par("cex.axis"), cex.names = par("cex.axis"),        inside = TRUE, plot = TRUE, axis.lty = 0, offset = 0,        plot.ci = FALSE, ci.l = NULL, ci.u = NULL,        ci.color = "black", ci.lty = "solid", ci.lwd = 1, ci.width = 0.5,        plot.grid = FALSE, grid.inc = NULL,        grid.lty = "dotted", grid.lwd = 1, grid.col = "black",        add = FALSE, panel.first = NULL, panel.last = NULL, ...)

Arguments

Details

This is a generic function, it currently only has a default method.A formula interface may be added eventually.

Value

A numeric vector (or matrix, whenbeside = TRUE), saymp, giving the coordinates ofall the bar midpointsdrawn, useful for adding to the graph.

Ifbeside is true, usecolMeans(mp) for themidpoints of eachgroup of bars, see example.

Note

Prior toR 1.6.0,barplot behaved as ifaxis.lty = 1,unintentionally.0 (zero) and NA values inheight will not be plotted ifusing logarithmic scales.If there are NA values inheight andbeside = FALSE,values after the NA will not be plotted in stacked bars.

Author(s)

Original barplot() by R-Core. Enhancements by MarcSchwartz.

See Also

plot(..., type = "h"),dotchart,hist.

Examples

tN <- table(Ni <- rpois(100, lambda = 5))r <- barplot2(tN, col = 'gray')#- type = "h" plotting *is* `bar'plotlines(r, tN, type = 'h', col = 'red', lwd = 2)barplot2(tN, space = 1.5, axisnames = FALSE,        sub = "barplot2(..., space = 1.5, axisnames = FALSE)")data(VADeaths, package = "datasets")barplot2(VADeaths, plot = FALSE)barplot2(VADeaths, plot = FALSE, beside = TRUE)mp <- barplot2(VADeaths) # defaulttot <- colMeans(VADeaths)text(mp, tot + 3, format(tot), xpd = TRUE, col = "blue")barplot2(VADeaths, beside = TRUE,        col = c("lightblue", "mistyrose", "lightcyan",                "lavender", "cornsilk"),        legend = rownames(VADeaths), ylim = c(0, 100))title(main = "Death Rates in Virginia", font.main = 4)# Example with confidence intervals and gridhh <- t(VADeaths)[, 5:1]mybarcol <- "gray20"ci.l <- hh * 0.85ci.u <- hh * 1.15mp <- barplot2(hh, beside = TRUE,        col = c("lightblue", "mistyrose",                "lightcyan", "lavender"),        legend = colnames(VADeaths), ylim = c(0, 100),        main = "Death Rates in Virginia", font.main = 4,        sub = "Faked 95 percent error bars", col.sub = mybarcol,        cex.names = 1.5, plot.ci = TRUE, ci.l = ci.l, ci.u = ci.u,        plot.grid = TRUE)mtext(side = 1, at = colMeans(mp), line = -2,      text = paste("Mean", formatC(colMeans(hh))), col = "red")box()# Example with horizontal bars, grid and logarithmic x axisbarplot2(1:10 , log = "x", plot.grid = TRUE, grid.inc = 10,        xlim = c(0.5, 20), horiz = TRUE, cex.axis = 0.9,        prcol = "gray95")box()# Bar shading examplebarplot2(VADeaths, angle = 15 + 10 * 1:5, density = 20, col = "black",        legend = rownames(VADeaths))title(main = list("Death Rates in Virginia", font = 4))# border :barplot2(VADeaths, border = "dark blue")

Produce a Boxplot Annotated with the Number of Observations

Description

This funcntion usesboxplot to produce a boxplot which is then annotated with the number of observations in each group.

Usage

boxplot2(..., top=FALSE, shrink=1, textcolor=NULL)

Arguments

Note

This function replacesboxplot.n, which has been deprecatedavoid potential problems with S3 method dispatching.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

boxplot,text

Examples

data(state)# n's at bottomboxplot2( state.area ~ state.region)# n's at topboxplot2( state.area ~ state.region, top=TRUE)# small red textboxplot2( state.area ~ state.region, shrink=0.8, textcolor="red")

Bubble Plot

Description

Draw a bubble plot, a scatterplot with varying symbol sizes andcolors, or add points to existing plots. A variety of input formatsare supported, including vectors, matrices, data frames, formulas,etc.

Usage

bubbleplot(x, ...)## Default S3 method:bubbleplot(x, y, z, std=TRUE, pow=0.5, add=FALSE,           rev=FALSE, type="p", ylim=NULL, xlab=NULL, ylab=NULL,           pch=c(16,1), cex.points=1, col="black", bg=par("bg"), ...)## S3 method for class 'formula'bubbleplot(formula, data, subset, na.action=NULL, ...)

Arguments

Details

Thestd standardization setsz = abs(z) / mean(abs(z)).

Thepow = 0.5 (square root) is a good default, where azvalue of 2 has twice the area of 1. See example #2 below for anexception, where thez value is tree circumference andtherefore proportional to the tree diameter.

Thepch,col, andbg arguments can be be vectorsof length 2, where positivez values are drawn withpch[1],col[1],bg[1] and negativezvalues are drawn withpch[2],col[2], andbg[2].

Author(s)

Arni Magnusson.

See Also

points is the underlying function used to draw thebubbles.

symbols can also draw bubbles, but does not handlenegativez values or have convenience features such aspow andrev.

balloonplot provides an alternative interface and visualstyle based on tables instead of scatterplots.

Examples

catch.t <- xtabs(Catch~Year+Age, catch.d)              # example tablecatch.m <- as.matrix(as.data.frame(unclass(catch.t)))  # example matrix# 1  Formulabubbleplot(Catch~Age+Year, data=catch.d)# Use rev=TRUE to get same layout as crosstab matrix:print(catch.m)bubbleplot(Catch~Age+Year, data=catch.d, rev=TRUE, las=1)# 2  Data framebubbleplot(catch.d)bubbleplot(Orange)# Visualize tree transverse section at breast heightbubbleplot(Orange, pow=1, cex=2, pch=21,           col="darkred", bg="peru", lwd=1.5)# 3  Matrix or tablebubbleplot(catch.m)bubbleplot(catch.t)# 4  Positive and negative valuesbubbleplot(catch.r)bubbleplot(Resid~Age+Year, catch.r, subset=Age %in% 4:9,           rev=TRUE, xlim=c(3.5,9.5), cex=1.3)# Residuals from orange tree modellibrary(nlme)fm <- nlme(circumference~phi1/(1+exp(-(age-phi2)/phi3)),           fixed=phi1+phi2+phi3~1, random=phi1~1|Tree,           data=Orange, start=c(phi1=200,phi2=800,phi3=400))bubbleplot(residuals(fm)~Tree+age, Orange)bubbleplot(residuals(fm)~Tree+age, Orange, cex=2.5, pch=16,           col=c("dodgerblue","orange"))# 5  Richter magnitude, amplitude, and energy releasebubbleplot(mag~long+lat, quakes, pch=1)bubbleplot(10^mag~long+lat, quakes, cex=1.2, col=gray(0, 0.3))bubbleplot(sqrt(1000)^mag~long+lat, quakes, cex=1.2, col=gray(0, 0.3))bubbleplot(sqrt(1000)^mag~long+lat, quakes, cex=1.2, col="#FF00004D")

Catch at Age and Residuals

Description

Catch-at-age observed data and model residuals from Icelandic saitheassessment.

Usage

catch.dcatch.r

Format

Data frame containing three columns:

Details

The data are from Tables 8.2 and 8.6 in the ICES (2015) fish stockassessment of Icelandic saithe.

Source

ICES (2015) Report of the North-Western Working Group (NWWG).ICES CM 2015/ACOM:07, pp. 240–246.

See Also

bubbleplot is an effective way to visualize these data.

Examples

catch.t <- xtabs(Catch~Year+Age, catch.d)catch.m <- as.matrix(as.data.frame(unclass(catch.t)))# 1  Formulabubbleplot(Catch~Age+Year, data=catch.d)# Use rev=TRUE to get same layout as crosstab matrix:print(catch.m)bubbleplot(Catch~Age+Year, data=catch.d, rev=TRUE, las=1)# 2  Data framebubbleplot(catch.d)# 3  Matrix or tablebubbleplot(catch.m)bubbleplot(catch.t)# 4  Positive and negative valuesbubbleplot(catch.r)bubbleplot(Resid~Age+Year, catch.r, subset=Age %in% 4:9,           rev=TRUE, xlim=c(3.5,9.5), cex=1.3)

Create 2-dimensional empirical confidence regions

Description

Create 2-dimensional empirical confidence regions from provided data.

Usage

ci2d(x, y = NULL,     nbins=51, method=c("bkde2D","hist2d"),     bandwidth, factor=1.0,     ci.levels=c(0.50,0.75,0.90,0.95,0.975),     show=c("filled.contour","contour","image","none"),     col=topo.colors(length(breaks)-1),     show.points=FALSE,     pch=par("pch"),     points.col="red",     xlab, ylab,      ...)## S3 method for class 'ci2d'print(x, ...)

Arguments

Details

This function utilizes eitherKernSmooth::bkde2D orgplots::hist2d to estmate a 2-dimensional density of the datapassed as an argument. This density is then used to create and(optionally) display confidence regions.

Whenbandwidth is ommited andmethod="bkde2d",KernSmooth::dpik is appled in x and y dimensions to select thebandwidth.

Value

Aci2d object consisting of a list containing (at least) thefollowing elements:

Note

Confidence intervals generated by ci2d areapproximate, andare subject to biases and/or artifacts induced by the binning orkernel smoothing method, bin locations, bin sizes, and kernel bandwidth.

Theconf2d function in ther2d2 package may create a moreaccurate confidence region, and reports the actual proportion ofpoints inside the region.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

bkde2D,conf2d,dpik,hist2d

Examples

   ####   ## Basic usage    ####   data(geyser, package="MASS")   x <- geyser$duration   y <- geyser$waiting   # 2-d confidence intervals based on binned kernel density estimate   ci2d(x,y)                   # filled contour plot   ci2d(x,y, show.points=TRUE) # show original data   # image plot   ci2d(x,y, show="image")   ci2d(x,y, show="image", show.points=TRUE)   # contour plot   ci2d(x,y, show="contour", col="black")   ci2d(x,y, show="contour", col="black", show.points=TRUE)   ####   ## Control Axis scales   ####   x <- rnorm(2000, sd=4)   y <- rnorm(2000, sd=1)   # 2-d confidence intervals based on binned kernel density estimate   ci2d(x,y)   # 2-d confidence intervals based on 2d histogram   ci2d(x,y, method="hist2d", nbins=25)    # Require same scale for each axis, this looks oval   ci2d(x,y, range.x=list(c(-20,20), c(-20,20)))   ci2d(x,y, method="hist2d", same.scale=TRUE, nbins=25) # hist2d    ####   ## Control smoothing and binning    ####   x <- rnorm(2000, sd=4)   y <- rnorm(2000, mean=x, sd=2)   # Default 2-d confidence intervals based on binned kernel density estimate   ci2d(x,y)   # change the smoother bandwidth   ci2d(x,y,        bandwidth=c(sd(x)/8, sd(y)/8)       )   # change the smoother number of bins   ci2d(x,y, nbins=10)   ci2d(x,y)   ci2d(x,y, nbins=100)   # Default 2-d confidence intervals based on 2d histogram   ci2d(x,y, method="hist2d", show.points=TRUE)   # change the number of histogram bins   ci2d(x,y, nbin=10, method="hist2d", show.points=TRUE )   ci2d(x,y, nbin=25, method="hist2d", show.points=TRUE )   ####   ## Perform plotting manually   ####   data(geyser, package="MASS")   # let ci2d handle plotting contours...   ci2d(geyser$duration, geyser$waiting, show="contour", col="black")   # call contour() directly, show the 90 percent CI, and the mean point    est <- ci2d(geyser$duration, geyser$waiting, show="none")   contour(est$x, est$y, est$cumDensity,           xlab="duration", ylab="waiting",           levels=0.90, lwd=4, lty=2)   points(mean(geyser$duration), mean(geyser$waiting),         col="red", pch="X")   ####   ## Extract confidence region values   ###   data(geyser, package="MASS")   ## Empirical 90 percent confidence limits   quantile( geyser$duration, c(0.05, 0.95) )   quantile( geyser$waiting, c(0.05, 0.95) )   ## Bivariate 90 percent confidence region   est <- ci2d(geyser$duration, geyser$waiting, show="none")   names(est$contours) ## show available contours   ci.90 <- est$contours[names(est$contours)=="0.9"]  # get region(s)   ci.90 <- rbind(ci.90[[1]],NA, ci.90[[2]], NA, ci.90[[3]]) # join them   print(ci.90)                  # show full contour   range(ci.90$x, na.rm=TRUE)    # range for duration   range(ci.90$y, na.rm=TRUE)    # range for waiting   ####   ## Visually compare confidence regions    ####   data(geyser, package="MASS")   ## Bivariate smoothed 90 percent confidence region   est <- ci2d(geyser$duration, geyser$waiting, show="none")   names(est$contours) ## show available contours   ci.90 <- est$contours[names(est$contours)=="0.9"]  # get region(s)   ci.90 <- rbind(ci.90[[1]],NA, ci.90[[2]], NA, ci.90[[3]]) # join them   plot( waiting ~ duration, data=geyser,         main="Comparison of 90 percent confidence regions" )   polygon( ci.90, col="green", border="green", density=10)   ## Univariate Normal-Theory 90 percent confidence region   mean.x <- mean(geyser$duration)   mean.y <- mean(geyser$waiting)   sd.x <- sd(geyser$duration)   sd.y <- sd(geyser$waiting)   t.value <- qt(c(0.05,0.95), df=length(geyser$duration), lower=TRUE)   ci.x <- mean.x +  t.value* sd.x   ci.y <- mean.y +  t.value* sd.y   plotCI(mean.x, mean.y,          li=ci.x[1],          ui=ci.x[2],          barcol="blue", col="blue",          err="x",          pch="X",          add=TRUE )   plotCI(mean.x, mean.y,          li=ci.y[1],          ui=ci.y[2],          barcol="blue", col="blue",          err="y",          pch=NA,          add=TRUE )#   rect(ci.x[1], ci.y[1], ci.x[2], ci.y[2], border="blue",#        density=5,#        angle=45,#        col="blue" )   ## Empirical univariate 90 percent confidence region   box <- cbind( x=quantile( geyser$duration, c(0.05, 0.95 )),                  y=quantile( geyser$waiting, c(0.05, 0.95 )) )   rect(box[1,1], box[1,2], box[2,1], box[2,2], border="red",        density=5,        angle=-45,        col="red" )   ## now a nice legend   legend( "topright", legend=c("       Region type",                                "Univariate Normal Theory",                                "Univarite Empirical",                                "Smoothed Bivariate"),           lwd=c(NA,1,1,1),           col=c("black","blue","red","green"),           lty=c(NA,1,1,1)         )   ####   ## Test with a large number of points   ####   ## Not run:    x <- rnorm(60000, sd=1)   y <- c( rnorm(40000, mean=x, sd=1),           rnorm(20000, mean=x+4, sd=1) )   hist2d(x,y)   ci <- ci2d(x,y)   ci   ## End(Not run)

Convert color names to hex RGB strings

Description

Convert color names to hex RGB strings

Usage

col2hex(cname)

Arguments

Value

Character vector giving the hex color code translation of the providedcolor names.

Author(s)

Gregory R. Warnes

See Also

col2rgb,colors,rgb

Examples

col2hex(c("red","yellow","lightgrey"))

Generate a smoothly varying set of colors

Description

colorpanel generate a set of colors that variessmoothly.redgreen,greenred,bluered, andredblue generate red-black-green, green-black-red,red-white-blue, and blue-white-red colorbars, respectively.colors

Usage

colorpanel(n, low, mid, high)redgreen(n)greenred(n)bluered(n)redblue(n)

Arguments

Details

The values forlow, mid, high can be given ascolor names ("red"), plot color index (2), andHTML-style RGB, ("#FF0000").

Ifmid is supplied, then the returned color panel will consistofn - floor(n/2) HTML-style RGB elements which vary smoothlybetweenlow andmid, then betweenmid andhigh. Note that ifn is even, the colormid willoccur twice at the center of the sequence.

Ifmid is omitted, the color panel will varysmoothly beweenlow andhigh.

Value

Vector of HTML-style RGB colors.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

colors

Examples

showpanel <- function(col){  image(z=matrix(1:100, ncol=1), col=col, xaxt="n", yaxt="n" )}par(mfrow=c(3,3))# two colors only:showpanel(colorpanel(8,low="red",high="green"))# three colorsshowpanel(colorpanel(8,"red","black","green"))# note the duplicatation of black at the center, using an odd# number of elements resolves this:showpanel(colorpanel(9,"red","black","green"))showpanel(greenred(64))showpanel(redgreen(64))showpanel(bluered(64))showpanel(redblue(64))

Defunct functions

Description

These functions are defunct and have been removed from the gplots package.

Usage

   boxplot.n(..., top=FALSE, shrink=1, textcolor=NULL)   plot.lm2(            x,            which = 1:5,            caption = c("Residuals vs Fitted", "Normal Q-Q plot",              "Scale-Location plot", "Cook's distance plot"),            panel = panel.smooth,            sub.caption = deparse(x$call),            main = "",            ask,            ...,            id.n = 3,            labels.id = names(residuals(x)),            cex.id = 0.75,            band=TRUE,            rug=TRUE,            width=1/10,            max.n=5000            )  smartlegend(x = c("left", "center", "right"),             y = c("top", "center", "bottom"),             ...,             inset = 0.05)

Arguments

Details

These functions are no longer available. Please refer to the manual pagefor the replacement function:

• boxplot.n has been replaced byboxplot2

• plot.lm2 has been replaced bylmplot2

• smartlegend is no longer needed because relativepositioning has been implemented inlegend.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

boxplot2,lmplot2,legend,Defunct

Enhanced Heat Map

Description

A heat map is a false color image (basicallyimage(t(x))) with a dendrogram added to the left sideand/or to the top. Typically, reordering of the rows and columnsaccording to some set of values (row or column means) within therestrictions imposed by the dendrogram is carried out.

This heatmap provides a number of extensions to the standard Rheatmap function.

Usage

heatmap.2 (x,           # dendrogram control           Rowv = TRUE,           Colv=if(symm)"Rowv" else TRUE,           distfun = dist,           hclustfun = hclust,           dendrogram = c("both","row","column","none"),           reorderfun = function(d, w) reorder(d, w),           symm = FALSE,           # data scaling           scale = c("none","row", "column"),           na.rm=TRUE,           # image plot           revC = identical(Colv, "Rowv"),           add.expr,           # mapping data to colors           breaks,           symbreaks=any(x < 0, na.rm=TRUE) || scale!="none",           # colors           col="heat.colors",           # block sepration           colsep,           rowsep,           sepcolor="white",           sepwidth=c(0.05,0.05),           # cell labeling           cellnote,           notecex=1.0,           notecol="cyan",           na.color=par("bg"),           # level trace           trace=c("column","row","both","none"),           tracecol="cyan",           hline=median(breaks),           vline=median(breaks),           linecol=tracecol,           # Row/Column Labeling           margins = c(5, 5),           ColSideColors,           RowSideColors,           cexRow = 0.2 + 1/log10(nr),           cexCol = 0.2 + 1/log10(nc),           labRow = NULL,           labCol = NULL,           srtRow = NULL,           srtCol = NULL,           adjRow = c(0,NA),           adjCol = c(NA,0),           offsetRow = 0.5,           offsetCol = 0.5,           colRow = NULL,           colCol = NULL,           # color key + density info           key = TRUE,           keysize = 1.5,           density.info=c("histogram","density","none"),           denscol=tracecol,           symkey = any(x < 0, na.rm=TRUE) || symbreaks,           densadj = 0.25,           key.title = NULL,           key.xlab = NULL,           key.ylab = NULL,           key.xtickfun = NULL,           key.ytickfun = NULL,           key.par=list(),           # plot labels           main = NULL,           xlab = NULL,           ylab = NULL,           # plot layout           lmat = NULL,           lhei = NULL,           lwid = NULL,           # extras           extrafun=NULL,           ...           )

Arguments

.

Details

If eitherRowv orColv are dendrograms they are honored(and not reordered).Otherwise, dendrograms are computed asdd <- as.dendrogram(hclustfun(distfun(X))) whereX iseitherx ort(x).

If either is a vector (of “weights”) then the appropriatedendrogram is reordered according to the supplied values subject tothe constraints imposed by the dendrogram, byreorder(dd, Rowv), in the row case.If either is missing, as by default, then the ordering of thecorresponding dendrogram is by the mean value of the rows/columns,i.e., in the case of rows,Rowv <- rowMeans(x, na.rm=na.rm).If either isNULL,no reordering will be done forthe corresponding side.

Ifscale="row" (orscale="col") the rows (columns) arescaled to have mean zero and standard deviation one. There is someempirical evidence from genomic plotting that this is useful.

The default colors range from red to white (heat.colors) andare not pretty. Consider using enhancements such as theRColorBrewer package,https://cran.r-project.org/package=RColorBrewerto select better colors.

By default four components will be displayed in the plot. At the topleft is the color key, top right is the column dendrogram, bottom leftis the row dendrogram, bottom right is the image plot. WhenRowSideColor or ColSideColor are provided, an additional row or columnis inserted in the appropriate location. This layout can beoverriden by specifiying appropriate values forlmat,lwid, andlhei.lmat controls the relativepostition of each element, whilelwid controls the columnwidth, andlhei controls the row height. See the help page forlayout for details on how to use thesearguments.

Value

Invisibly, a list with components

Note

The original rows and columns are reordered to match the dendrogramsRowv andColv (if present).

heatmap.2() useslayout to arragent the plotelements. Consequentially, it cannot be used in a multicolumn/row layout usinglayout(...),par(mfrow=...) or(mfcol=...).

Author(s)

Andy Liaw, original; R. Gentleman, M. Maechler, W. Huber,G. Warnes, revisions.

See Also

image,hclust

Examples

 data(mtcars) x  <- as.matrix(mtcars) rc <- rainbow(nrow(x), start=0, end=.3) cc <- rainbow(ncol(x), start=0, end=.3) ## ## demonstrate the effect of row and column dendrogram options ## heatmap.2(x)                    ## default - dendrogram plotted and reordering done. heatmap.2(x, dendrogram="none") ##  no dendrogram plotted, but reordering done. heatmap.2(x, dendrogram="row")  ## row dendrogram plotted and row reordering done. heatmap.2(x, dendrogram="col")  ## col dendrogram plotted and col reordering done. heatmap.2(x, keysize=2)         ## default - dendrogram plotted and reordering done. heatmap.2(x, Rowv=FALSE, dendrogram="both") ## generates a warning! heatmap.2(x, Rowv=NULL, dendrogram="both")  ## generates a warning! heatmap.2(x, Colv=FALSE, dendrogram="both") ## generates a warning! ## Reorder dendrogram by branch means rather than sums heatmap.2(x, reorderfun=function(d, w) reorder(d, w, agglo.FUN = mean) ) ## Color branches of dendrograms by cluster membership using dendextend: library(dendextend)  # for color_branches full <- heatmap.2(x) # we use it to easily get the dendrograms heatmap.2(x,           Rowv=color_branches(full$rowDendrogram, k = 3),          Colv=color_branches(full$colDendrogram, k = 2)) # Look at the vignette for more details: # https://cran.r-project.org/web/packages/dendextend/vignettes/dendextend.html ## plot a sub-cluster using the same color coding as for the full heatmap full <- heatmap.2(x) heatmap.2(x, Colv=full$colDendrogram[[2]], breaks=full$breaks)  # column subset heatmap.2(x, Rowv=full$rowDendrogram[[1]], breaks=full$breaks)  # row subset heatmap.2(x, Colv=full$colDendrogram[[2]],              Rowv=full$rowDendrogram[[1]], breaks=full$breaks)  # both ## Show effect of row and column label rotation heatmap.2(x, srtCol=NULL) heatmap.2(x, srtCol=0,   adjCol = c(0.5,1) ) heatmap.2(x, srtCol=45,  adjCol = c(1,1)   ) heatmap.2(x, srtCol=135, adjCol = c(1,0)   ) heatmap.2(x, srtCol=180, adjCol = c(0.5,0) ) heatmap.2(x, srtCol=225, adjCol = c(0,0)   ) ## not very useful heatmap.2(x, srtCol=270, adjCol = c(0,0.5) ) heatmap.2(x, srtCol=315, adjCol = c(0,1)   ) heatmap.2(x, srtCol=360, adjCol = c(0.5,1) ) heatmap.2(x, srtRow=45, adjRow=c(0, 1) ) heatmap.2(x, srtRow=45, adjRow=c(0, 1), srtCol=45, adjCol=c(1,1) ) heatmap.2(x, srtRow=45, adjRow=c(0, 1), srtCol=270, adjCol=c(0,0.5) ) ## Show effect of offsetRow/offsetCol (only works when srtRow/srtCol is ## not also present) heatmap.2(x, offsetRow=0, offsetCol=0) heatmap.2(x, offsetRow=1, offsetCol=1) heatmap.2(x, offsetRow=2, offsetCol=2) heatmap.2(x, offsetRow=-1, offsetCol=-1) heatmap.2(x, srtRow=0, srtCol=90, offsetRow=0, offsetCol=0) heatmap.2(x, srtRow=0, srtCol=90, offsetRow=1, offsetCol=1) heatmap.2(x, srtRow=0, srtCol=90, offsetRow=2, offsetCol=2) heatmap.2(x, srtRow=0, srtCol=90, offsetRow=-1, offsetCol=-1) ## Show how to use 'extrafun' to replace the 'key' with a scatterplot lmat <- rbind( c(5,3,4), c(2,1,4) ) lhei <- c(1.5, 4) lwid <- c(1.5, 4, 0.75) myplot <- function() {             oldpar <- par("mar")             par(mar=c(5.1, 4.1, 0.5, 0.5))             plot(mpg ~ hp, data=x)           } heatmap.2(x, lmat=lmat, lhei=lhei, lwid=lwid, key=FALSE, extrafun=myplot) ## show how to customize the color key heatmap.2(x,           key.title=NA, # no title           key.xlab=NA,  # no xlab           key.par=list(mgp=c(1.5, 0.5, 0),                        mar=c(2.5, 2.5, 1, 0)),           key.xtickfun=function() {                 breaks <- parent.frame()$breaks                 return(list(                      at=parent.frame()$scale01(c(breaks[1],                                                  breaks[length(breaks)])),                      labels=c(as.character(breaks[1]),                               as.character(breaks[length(breaks)]))                      ))           }) heatmap.2(x,          breaks=256,          key.title=NA,          key.xlab=NA,          key.par=list(mgp=c(1.5, 0.5, 0),                       mar=c(1, 2.5, 1, 0)),          key.xtickfun=function() {               cex <- par("cex")*par("cex.axis")               side <- 1               line <- 0               col <- par("col.axis")               font <- par("font.axis")               mtext("low", side=side, at=0, adj=0,                     line=line, cex=cex, col=col, font=font)               mtext("high", side=side, at=1, adj=1,                     line=line, cex=cex, col=col, font=font)               return(list(labels=FALSE, tick=FALSE))          }) ## ## Show effect of z-score scaling within columns, blue-red color scale ## hv <- heatmap.2(x, col=bluered, scale="column", tracecol="#303030") ### ## Look at the return values ### names(hv) ## Show the mapping of z-score values to color bins hv$colorTable ## Extract the range associated with white hv$colorTable[hv$colorTable[,"color"]=="#FFFFFF",] ## Determine the original data values that map to white whiteBin <- unlist(hv$colorTable[hv$colorTable[,"color"]=="#FFFFFF",1:2]) rbind(whiteBin[1] * hv$colSDs + hv$colMeans,       whiteBin[2] * hv$colSDs + hv$colMeans ) ## ## A more decorative heatmap, with z-score scaling along columns ## hv <- heatmap.2(x, col=cm.colors(255), scale="column",       RowSideColors=rc, ColSideColors=cc, margin=c(5, 10),       xlab="specification variables", ylab= "Car Models",       main="heatmap(<Mtcars data>, ..., scale=\"column\")",         tracecol="green", density="density") ## Note that the breakpoints are now symmetric about 0 ## Color the labels to match RowSideColors and ColSideColors hv <- heatmap.2(x, col=cm.colors(255), scale="column",         RowSideColors=rc, ColSideColors=cc, margin=c(5, 10),       xlab="specification variables", ylab= "Car Models",       main="heatmap(<Mtcars data>, ..., scale=\"column\")",         tracecol="green", density="density", colRow=rc, colCol=cc,         srtCol=45, adjCol=c(0.5,1)) data(attitude) round(Ca <- cor(attitude), 2) symnum(Ca) # simple graphic # with reorder heatmap.2(Ca,  symm=TRUE, margin=c(6, 6), trace="none" ) # without reorder heatmap.2(Ca, Rowv=FALSE, symm=TRUE, margin=c(6, 6), trace="none" ) ## Place the color key below the image plot heatmap.2(x, lmat=rbind( c(0, 3), c(2,1), c(0,4) ), lhei=c(1.5, 4, 2 ) ) ## Place the color key to the top right of the image plot heatmap.2(x, lmat=rbind( c(0, 3, 4), c(2,1,0 ) ), lwid=c(1.5, 4, 2 ) ) ## For variable clustering, rather use distance based on cor(): data(USJudgeRatings) symnum( cU <- cor(USJudgeRatings) ) hU <- heatmap.2(cU, Rowv=FALSE, symm=TRUE, col=topo.colors(16),              distfun=function(c) as.dist(1 - c), trace="none") ## The Correlation matrix with same reordering: hM <- format(round(cU, 2)) hM # now with the correlation matrix on the plot itself heatmap.2(cU, Rowv=FALSE, symm=TRUE, col=rev(heat.colors(16)),             distfun=function(c) as.dist(1 - c), trace="none",             cellnote=hM) ## genechip data examples ## Not run:  library(affy) data(SpikeIn) pms <- SpikeIn@pm # just the data, scaled across rows heatmap.2(pms, col=rev(heat.colors(16)), main="SpikeIn@pm",              xlab="Relative Concentration", ylab="Probeset",              scale="row") # fold change vs "12.50" sample data <- pms / pms[, "12.50"] data <- ifelse(data>1, data, -1/data) heatmap.2(data, breaks=16, col=redgreen, tracecol="blue",               main="SpikeIn@pm Fold Changes\nrelative to 12.50 sample",               xlab="Relative Concentration", ylab="Probeset") ## End(Not run)  ## You can get an interactive version of heatmap.2 using heatmaply: ## Not run:  data(mtcars) x  <- as.matrix(mtcars) library(heatmaply) # just use heatmaply instead of heatmap.2 heatmaply(x)  # The default are slightly different, but it supports most of the same arguments. # If you want the dendrograms to match perfectly, use this: ## Color branches of dendrograms by cluster membership using dendextend: full <- heatmap.2(x) # we use it to easily get the dendrograms heatmaply(x,           Rowv=color_branches(full$rowDendrogram, k = 3),          Colv=color_branches(full$colDendrogram, k = 2)) # Look at the vignette for more details: # https://cran.r-project.org/web/packages/heatmaply/vignettes/heatmaply.html ## End(Not run)

Compute and Plot a 2-Dimensional Histogram

Description

Compute and plot a 2-dimensional histogram.

Usage

hist2d(x,y=NULL, nbins=200, same.scale=FALSE, na.rm=TRUE, show=TRUE,       col=c("black", heat.colors(12)), FUN=base::length, xlab, ylab,       ... )## S3 method for class 'hist2d'print(x, ...)

Arguments

Details

This fucntion creates a 2-dimensional histogram by cutting the x andy dimensions intonbins sections. A 2-dimensional matrix isthen constucted which holds the counts of the number of observed (x,y) pairsthat fall into each bin. Ifshow=TRUE, this matrix is thenthen passed toimage for display.

Value

A list containing 5 elements:

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

image,persp,hist,freq2d

Examples

   ## example data, bivariate normal, no correlation   x <- rnorm(2000, sd=4)   y <- rnorm(2000, sd=1)   ## separate scales for each axis, this looks circular   hist2d(x,y)   ## same scale for each axis, this looks oval   hist2d(x,y, same.scale=TRUE)   ## use different ## bins in each dimension   hist2d(x,y, same.scale=TRUE, nbins=c(100,200) )   ## use the hist2d function to create an h2d object   h2d <- hist2d(x,y,show=FALSE, same.scale=TRUE, nbins=c(20,30))   ## show object summary   h2d   ## object contents   str(h2d)   ## perspective plot   persp( h2d$x, h2d$y, h2d$counts,          ticktype="detailed", theta=30, phi=30,          expand=0.5, shade=0.5, col="cyan", ltheta=-30)   ## for contour (line) plot ...   contour( h2d$x, h2d$y, h2d$counts, nlevels=4 )   ## for a filled contour plot ...   filled.contour( h2d$x, h2d$y, h2d$counts, nlevels=4,                   col=gray((4:0)/4) )

Plots to assess the goodness of fit for the linear model objects

Description

Plots to assess the goodness of fit for the linear model objects

Usage

   lmplot2(            x,            which = 1:5,            caption = c("Residuals vs Fitted", "Normal Q-Q plot",              "Scale-Location plot", "Cook's distance plot"),            panel = panel.smooth,            sub.caption = deparse(x$call),            main = "",            ask = interactive() && nb.fig < length(which)            && .Device != "postscript",            ...,            id.n = 3,            labels.id = names(residuals(x)),            cex.id = 0.75,            band=TRUE,            rug=TRUE,            width=1/10,            max.n=5000            )

Arguments

Note

This function replacesplot.lm2, which has been deprecatedto avoid potential problems with S3 method dispatching.

Author(s)

Gregory R. Warnesgreg@warnes.net and NitinJainnitin.jain@pfizer.com

See Also

plot.lm

Examples

ctl <- rnorm(100, 4)trt <- rnorm(100, 4.5)group <- gl(2,100,200, labels=c("Ctl","Trt"))weight <- c(ctl, trt)wt.err <- rnorm(length(weight), mean=weight, sd=1/2)x <- lm(weight ~ group + wt.err)lmplot2(x)lmplot2(x, which=1,   width=1/3)lmplot2(x, which=1:3, width=1/3)

Scatter Plot Smoothing

Description

Thelowess function performs the computations for theLOWESS smoother (see the reference below).lowess returns a an object containing componentsx andy which give the coordinates of the smooth.The smooth can then be added to a plot of the originalpoints with the functionlines.

Alternatively,plot can be called directly on the objectreturned fromlowess and the 'lowess' method forplotwill generate a scatterplot of the original data with alowessline superimposed.

Finally, theplotLowess function both calculates thelowess smooth and plots the original data with alowesssmooth.

Usage

lowess(x, ...)## Default S3 method:lowess(x, y=NULL, f=2/3, iter=3L, delta=0.01 *       diff(range(x)), ...)## S3 method for class 'formula'lowess(formula,data=parent.frame(), ..., subset, f=2/3,       iter=3L, delta=.01*diff(range(mf[-response])))## S3 method for class 'lowess'plot(x, y, ..., col.lowess="red", lty.lowess=2)plotLowess(formula, data=parent.frame(), ..., subset=parent.frame(),           col.lowess="red", lty.lowess=2  )

Arguments

References

Cleveland, W. S. (1979)Robust locally weighted regression and smoothing scatterplots.J. Amer. Statist. Assoc.74, 829–836.

Cleveland, W. S. (1981)LOWESS: A program for smoothing scatterplots by robust locally weightedregression.The American Statistician,35, 54.

See Also

loess (in packagemodreg), a newerformula based version oflowess (with different defaults!).

Examples

data(cars)## x,y method#plot(cars$speed, cars$dist, main="lowess(cars)")lines(lowess(cars$speed, cars$dist), col=2)lines(lowess(cars$speed, cars$dist, f=.2), col=3)legend(5, 120, c(paste("f=", c("2/3", ".2"))), lty=1, col=2:3)## formula method: plot, then calculate the lowess smoother,#                 then add smooth to the plot#plot(dist ~ speed, data=cars, main="lowess(cars)")lines(lowess(dist ~ speed, data=cars), col=2, lty=2)lines(lowess(dist ~ speed, data=cars, f=.2), col=3) # smaller bandwithlegend(5, 120, c(paste("f=", c("2/3", ".2"))), lty=1, col=2:3)## formula method: calculate lowess() smoother, then call plot()#                  on the lowess object#lw <- lowess(dist ~ speed, data=cars)plot(lw, main="lowess(cars)"  )## formula method: calculate and plot in a single command#plotLowess(dist ~ speed, data=cars, main="lowess(cars)")

Create an OpenOffice style plot

Description

An extension of barplot2. Creates bar- and line-plots mimicking thestyle of OpenOffice plots. This utility can plot the values next toeach point or bar as well as confidence intervals.

Usage

ooplot(data, ...)## Default S3 method:ooplot(data, width=1, space=NULL, names.arg=NULL,                            legend.text=NULL, horiz=FALSE,                            density=NULL, angle=45, kmg="fpnumkMGTP",                            kmglim=TRUE,                            type=c("xyplot", "linear", "barplot", "stackbar"),                            col=heat.colors(NC), prcol=NULL,                            border=par("fg"), main=NULL, sub=NULL,                            xlab=NULL, ylab=NULL, xlim=NULL, ylim=NULL,                            xpd=TRUE, log="", axes=TRUE,                            axisnames=TRUE, prval=TRUE, lm=FALSE,                           cex.axis=par("cex.axis"),                            cex.names=par("cex.axis"),                           cex.values=par("cex"),inside=TRUE,                            plot=TRUE, axis.lty=0, plot.ci=FALSE,                            ci.l=NULL, ci.u=NULL, ci.color="black",                            ci.lty="solid", ci.lwd=1, plot.grid=FALSE,                            grid.inc=NULL, grid.lty="dotted",                            grid.lwd=1, grid.col="black", add=FALSE,                            by.row=FALSE, ...)

Arguments

Details

Plot units are automatically scaled to SI units based on themaximum value present, according to the set of units specified bycharacters in thekmg parameter. These letters are interpretedas

P

peta = 1E15

T

tera = 1E12

G

giga = 1E09

M

mega = 1E06

k

kilo = 1E03

m

milli= 1E-03

u

micro= 1E-06

n

nano = 1E-09

p

pico = 1E-12

f

femto= 1E-15

with the default being "fpnumkMGTP" (all of these units). Forexample, if the largest value plotted is 1243000, it would bepresented as 1.234M.

Value

A numeric vector (or matrix, whenbeside = TRUE), saymp, giving the coordinates ofall the bar midpointsdrawn, useful for adding to the graph.

Ifbeside is true, usecolMeans(mp) for themidpoints of eachgroup of bars, see example.

Author(s)

Lodewijk Bonebakkerbonebakker@comcast.net with modificationsby Gregory R. Warnesgreg@warnes.net.Based on barplot2().

See Also

plot,boxplot

Examples

     data(VADeaths, package = "datasets")     VADeaths <- cbind( Age=c(50,55,60,65,70), VADeaths)     mp <- ooplot(VADeaths) # default     mp <- ooplot(VADeaths, type="xyplot")  # same as default     mp <- ooplot(VADeaths, type="linear")  # linear scale     mp <- ooplot(VADeaths, type="linear", log="y") # log scale on y axis     mp <- ooplot(VADeaths, type="barplot") # barplot     mp <- ooplot(VADeaths, type="stackbar") # stacked     tot <- colMeans(VADeaths[,-1])     ooplot(VADeaths,              col = c("lightblue", "mistyrose", "lightcyan", "lavender"),             legend = colnames(VADeaths)[-1], ylim = c(0, 100),             type="barplot", cex.values=0.75)     title(main = "Death Rates in Virginia", font.main = 4)     ##     ## Capability demo     ##     ## examples for the ooplot routine     ##     ## create some test data     test1 <- data.frame(x=c(0,1,2,3,4), lin=c(0,1,2,3,4))     test2 <- data.frame(x=c(0,1,2,3,4), par=c(0,1,4,9,16))     test3 <- data.frame(x=c(-2,-1,0,1,2),y2=c(4,1,0,1,4))     ## single line test example     test1f <- test1     ## two column example     test2f <- merge(test1,test2,by.x="x",all=TRUE,sort=TRUE)     ## three column example     test3f <- merge(test2f,test3,by.x="x",all=TRUE,sort=TRUE)     ## subset, single row, example     test5r <- test3f[5,]            ##     ## xyplot, linear, barplot, stackbar     dev.off()     mat <- matrix(c(1:16),4,4,byrow=TRUE)     layout(mat)          ooplot(test1f,type="barplot",col=c("red"))     title(main="barplot")     ooplot(test2f,type="barplot",col=c("red","blue"))     ooplot(test3f,type="barplot",col=c("red","blue","green"))     ooplot(test5r,type="barplot",col=c("red","blue","green"))          ooplot(test1f,type="xyplot",col=c("red"))     title(main="xyplot")     ooplot(test2f,type="xyplot",col=c("red","blue"))     ooplot(test3f,type="xyplot",col=c("red","blue","green"))     ooplot(test5r,type="xyplot",col=c("red","blue","green"))          ooplot(test1f,type="linear",col=c("red"))     title(main="linear")     ooplot(test2f,type="linear",col=c("red","blue"))     ooplot(test3f,type="linear",col=c("red","blue","green"))     ooplot(test5r,type="linear",col=c("red","blue","green"))          ooplot(test1f,type="stackbar",col=c("red"))     title(main="stackbar")     ooplot(test2f,type="stackbar",col=c("red","blue"))     ooplot(test3f,type="stackbar",col=c("red","blue","green"))     ooplot(test5r,type="stackbar",col=c("red","blue","green"))     # restore default layout (1 plot/page)     layout(1)

Plot multiple variables on the same region, with appropriate axes

Description

overplot graphs a set of variables defined on the same x-rangebut which have varying y-ranges on the same plotting area. For eachset of y-values it uses a different color and line-type and and drawsa correspondingly colored and line-typed axis.panel.overplotis used byoverplot to draw the individual graphs.

Usage

overplot(formula, data = parent.frame(), same.scale = FALSE, xlab, ylab,         xlim, ylim, min.y, max.y, log = "", panel = "panel.overplot",         subset, plot = TRUE, groups, main, f = 2/3, ...)

Arguments

Details

This function essentially performs

tmp <- split(data, z)

for(i in levels(z))

plot( x ~ y, data=tmp[[z]] )

except that all of the plots are shown on the same plotting region andvarying scales for each value of z are handled nicely.

Value

A copy of the data split by the conditioning variable.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

interaction.plot,coplot for alternative visualizations of 3-way data.

Examples

# Example teratogenicity rtPCR datadata(rtPCR)# same scaleoverplot( RQ ~ Conc..ug.ml. | Test.Substance,         data=rtPCR,         subset=Detector=="ProbeType 1" & Conc..ug.ml. > 0,         same.scale=TRUE,         log="xy",         f=3/4,         main="Detector=ProbeType 1",         xlab="Concentration (ug/ml)",         ylab="Relative Gene Quantification"         )# different scales, but force lower limit to 0.01 overplot( RQ ~ Conc..ug.ml. | Test.Substance,         data=rtPCR,         subset=Detector=="ProbeType 8" & Conc..ug.ml. > 0,         log="xy",         f=3/4,         main="Detector=ProbeType 8",         xlab="Concentration (ug/ml)",         ylab="Relative Gene Quantification",         min.y=0.01         )

Plot Error Bars and Confidence Intervals

Description

Given a set of x and y values and interval width or upper and lower bounds,plot the points with error bars. This can be a useful tool forvisualizing confidence intervals.

Usage

plotCI(x, y = NULL, uiw, liw = uiw, ui, li, err='y', ylim=NULL,       xlim=NULL, type="p",  col=par("col"), barcol=col,       pt.bg = par("bg"),  sfrac = 0.01, gap=1, lwd=par("lwd"),       lty=par("lty"), labels=FALSE, add=FALSE, xlab, ylab,  minbar,       maxbar, ... )

Arguments

Author(s)

Original version by Bill Venableswvenable@attunga.stats.adelaide.edu.au posted to r-help onSep. 20, 1997. Enhanced version posted to r-help by Ben Bolkerben@zoo.ufl.edu on Apr. 16, 2001. This version was modifiedand extended by Gregory R. Warnesgreg@warnes.net. Additional changessuggested by Martin Maechlermaechler@stat.math.ethz.chintegrated on July 29, 2004.

See Also

plotmeans provides an enhanced wrapper toplotCI.

Examples

  # plot means and  data(state)  tmp   <- split(state.area, state.region)  means <- sapply(tmp, mean)  stdev <- sqrt(sapply(tmp, var))  n     <- sapply(tmp,length)  ciw   <- qt(0.975, n) * stdev / sqrt(n)  # plain  plotCI(x=means, uiw=ciw)  # prettier  plotCI(x=means, uiw=ciw, col="black", barcol="blue", lwd=1)  # give mean values  plotCI(x=means, uiw=ciw, col="black", barcol="blue",         labels=round(means,-3), xaxt="n", xlim=c(0,5) )  axis(side=1, at=1:4, labels=names(tmp), cex=0.7)  # better yet, just use plotmeans ... #  plotmeans( state.area ~ state.region )

Plot Group Means and Confidence Intervals

Description

Plot group means and confidence intervals.

Usage

plotmeans(formula, data=NULL, subset, na.action,          bars=TRUE, p=0.95, minsd=0, minbar, maxbar,          xlab=names(mf)[2], ylab=names(mf)[1], mean.labels=FALSE,          ci.label=FALSE, n.label=TRUE, text.n.label="n=",          digits=getOption("digits"), col="black", barwidth=1,          barcol="blue", connect=TRUE, ccol=          col, legends=names(means), xaxt, use.t=TRUE,          lwd=par("lwd"), ...)

Arguments

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

plotCI,boxplot

Examples

  # library(gplots)  # show comparison with boxplot  data(state)  plotmeans(state.area ~ state.region)  # show some color and mean labels  plotmeans(state.area ~ state.region,             mean.labels=TRUE, digits=-3,             col="red", connect=FALSE)  # show how to specify which means should be connected  plotmeans(state.area ~ state.region, connect=list(1:2, 3:4),            ccol="red", pch=7 )  # more complicated example showing how to show an interaction  data(esoph)  par(las=2,                        # use perpendicular axis labels      mar=c(10.1,4.1,4.1,2.1),      # create enough space for long x labels      mgp=c(8,1,0)                  # move x axis legend down to avoid overlap      )  plotmeans(ncases/ncontrols ~ interaction(agegp , alcgp, sep ="   "),            connect=list(1:6,7:12,13:18,19:24),            barwidth=2,            col="dark green",            data=esoph,            xlab="Age Group and Alcohol Consumption",            ylab="# Cases / # Controls",            ylim = c(-.9,1.4),            main=c("Fraction of Cases for by Age and Alcohol Consumption",                   "Ile-et-Vilaine Esophageal Cancer Study")            )  abline(v=c(6.5, 12.5, 18.5), lty=2)

Makes a half or full normal plot for the effects from an aov model

Description

Makes a half or full normal plot for the effects from a modelinheriting fromclassaov. One can interactively label the points in the plot.

Usage

## S3 method for class 'aov'qqnorm(y, full=FALSE, label=FALSE, omit=NULL,           xlab=paste(if (full) "" else "Half", " Normal plot"),            ylab="Effects", ...)

Arguments

Details

Produces a (half) normal plot of the effects from an AOV model. Theidea behind the plot is that most effects will be small or null, andthis effects can be used as a basis for estimation of theexperimental variance. This small effects will show up in the plot asa straight line, other effects can be judged against this as abackground. Heavily used by Box, Hunter & Hunter, which attributesthe idea to Daniel.

Value

Iflabel=TRUE, the vector of points identified, else nothing ofinterest.

Author(s)

Kjetil Halvorsenkjetil@entelnet.bo

References

Box, Hunter and Hunter: Statistics for Experimenters. An Introductionto Design, Data Analysis and Model Building. Wiley.
Daniel, C (1976): Applications of Statistics to IndustrialExperimentation. Wiley.
Daniel, C (1959): Use of half-normal plot in interpreting factorialtwo-level experiments.Technometrics.1, 149.

Examples

library(MASS)data(npk)npk.aov <- aov(yield ~ block + N*P*K, npk)qqnorm(npk.aov)## interactive labeling of points.  Click mouse on points to show label.if (dev.interactive()) qqnorm(npk.aov, omit=2:6, label=TRUE)

Reorder the Levels of a Factor

Description

Reorder the levels of a factor

Usage

## S3 method for class 'factor'reorder(x, X, FUN, ..., order=is.ordered(x), new.order, sort=mixedsort)

Arguments

Details

This function changes the order of the levels of a factor. It can doso via three different mechanisms, depending on whether,XandFUN,new.order orsort are provided.

IfXandFun are provided: The data inXis grouped by the levels ofx andFUN is applied.The groups are then sorted by this value, and the resulting order isused for the new factor level names.

Ifnew.order is a numeric vector, the new factor level namesare constructed by reordering the factor levels according to thenumeric values. Ifnew.order is a chraccter vector,new.order gives the list of new factor level names. In eithercase levels omitted fromnew.order will become missing(NA) values.

Ifsort is provided (as it is by default): The new factor levelnames are generated by calling the function specified bysortto the existing factor levelnames. Withsort=mixedsort(the default) the factor levels are sorted so that combined numericand character strings are sorted in according to character rules onthe character sections (including ignoring case), and the numericrules for the numeric sections. Seemixedsortfor details.

Value

A new factor with reordered levels

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

factorandreorder

Examples

   # Create a 4 level example factor   trt <- factor( sample( c("PLACEBO", "300 MG", "600 MG", "1200 MG"),                  100, replace=TRUE ) )   summary(trt)   # Note that the levels are not in a meaningful order.   # Change the order to something useful..   # - default "mixedsort" ordering   trt2 <- reorder(trt)   summary(trt2)   # - using indexes:   trt3 <- reorder(trt, new.order=c(4, 2, 3, 1))   summary(trt3)   # - using label names:   trt4 <- reorder(trt, new.order=c("PLACEBO", "300 MG", "600 MG", "1200 MG"))   summary(trt4)   # - using frequency   trt5 <- reorder(trt, X=rnorm(100), FUN=mean)   summary(trt5)   # Drop out the '300 MG' level   trt6 <- reorder(trt, new.order=c("PLACEBO", "600 MG", "1200 MG"))   summary(trt6)

Undocumented functions

Description

These functions are undocumented. Some are internal and not intendedfor direct use. Others simply haven't been documented yet.

Usage

residplot(model, formula, ...)

Arguments

Details

These functions are undocumented. Some are internal and not intendedfor direct use. Others simply haven't been documented yet.

Author(s)

Gregory R. Warnesgreg@warnes.net

Rich Color Palettes

Description

Create a vector ofn colors that are perceptually equidistantand in an order that is easy to interpret.

Usage

rich.colors(n, palette="temperature", alpha=1.0, rgb=FALSE, plot=FALSE)

Arguments

Value

A character vector of color codes.

Author(s)

Arni Magnusson.

See Also

rgb,rainbow,heat.colors.

Examples

m <- abs(matrix(1:120+rnorm(120), nrow=15, ncol=8))opar <- par(bg="gray", mfrow=c(1,2))matplot(m, type="l", lty=1, lwd=3, col=rich.colors(8))matplot(m, type="l", lty=1, lwd=3, col=rich.colors(8,"blues"))par(opar)barplot(rep(1,100), col=rich.colors(100), space=0, border=0, axes=FALSE)barplot(rep(1,20), col=rich.colors(40)[11:30]) # choose subsetplot(m, rev(m), ylim=c(120,0), pch=16, cex=2,     col=rich.colors(200,"blues",alpha=0.6)[1:120]) # semitransparentrich.colors(100, plot=TRUE)  # describe rgb recipepar(mfrow=c(2,2))barplot(m, col=heat.colors(15), main="\nheat.colors")barplot(m, col=1:15, main="\ndefault palette")barplot(m, col=rich.colors(15), main="\nrich.colors")barplot(m, col=rainbow(15), main="\nrainbow")par(opar)

Teratogenesis rtPCR data

Description

rtPCR data for experiments investigating a variety of markers forcharacterizing teratogenicity.

Usage

data(rtPCR)

Format

A data frame with 1672 observations on the following 21 variables.

PlateID

a factor with levelsA0027002 throughA0054019

Test.Substance

a factor with levelsCompound A throughCompound H

Teratogenicity.in.vivo

a factor with levelsNonStrongWeak / Moderate

Sample

a factor with levelsSample 1 -Sample 152

Rep..

a factor with levelsRep 1 -Rep 21

Label

a factor with levelsCtrl,Neg. CtrlP1 -P9,No Vehicle Ctrl, andPos. Ctrl

Conc..ug.ml.

a numeric vector

Detector

a factor with levelsProbeType 1 -ProbeType 17

Avg.delta.Ct

a numeric vector

delta.Ct.SD

a numeric vector

delta.delta.Ct

a numeric vector

RQ

a numeric vector

X..RQ

a numeric vector

X100..Custom..

a numeric vector

X100...Custom..

a numeric vector

Custom..

a numeric vector

Custom...1

a numeric vector

RQ.Min

a numeric vector

RQ.Max

a numeric vector

Threshold

a numeric vector

Details

TBA

Source

Anonymized data.

Examples

data(rtPCR)# same scaleoverplot( RQ ~ Conc..ug.ml. | Test.Substance,         data=rtPCR,         subset=Detector=="ProbeType 7" & Conc..ug.ml. > 0,         same.scale=TRUE,         log="xy",         f=3/4,         main="Detector=ProbeType 7",         xlab="Concentration (ug/ml)",         ylab="Relative Gene Quantification"         )# different scales, but force lower limit to 0.01 overplot( RQ ~ Conc..ug.ml. | Test.Substance,         data=rtPCR,         subset=Detector=="ProbeType 7" & Conc..ug.ml. > 0,         log="xy",         f=3/4,         main="Detector=ProbeType 7",         xlab="Concentration (ug/ml)",         ylab="Relative Gene Quantification",         min.y=0.01         )

Send textual R output to a graphics device

Description

Divert R's standard text output to a graphics device.

Usage

sinkplot(operation = c("start", "plot", "cancel"), ...)

Arguments

Details

This function allows the printed output of R commands to be capturedand displayed on a graphics device.

The capture process is started by callingsinkplot("start"). Now R commands can be executed and allprinted output (except errors) will be captured. When the desiredtext has been capturedsinkplot("plot") can be called toactually display the output.sinkplot("cancel") can be used toabort the output capture without plotting.

The current implementation does not allowsinkplot to be nested.

Value

Invisibly returns a character vector containing one element for eachline of the captured output.

Author(s)

Gregory R. Warnesgreg@warnes.net

References

Functionality requested by Kevin Wrightkwright@eskimo.com in the R-devel newlist postinghttps://stat.ethz.ch/pipermail/r-devel/2004-January/028483.html.

See Also

capture.output,textplot

Examples

   set.seed(12456)   x <- factor(sample( LETTERS[1:5], 50, replace=TRUE))   y <- rnorm(50, mean=as.numeric(x), sd=1)   ## construct a figure showing a box plot of the data, followed by an   ## analysis of variance table for the data   layout(cbind(1:2), heights=c(2,1))   boxplot(y~x, col="darkgreen")   sinkplot()   anova(lm(y~x))   sinkplot("plot",col="darkgreen")

Space points in an x-y plot so they don't overlap.

Description

Space points in an x-y plot so they don't overlap.

Usage

space(x, y, s=1/50, na.rm=TRUE, direction="x")

Arguments

Details

In an x-y plot where at least one variable has discrete levels severalpoints may be plotted at or very near the same coordonates.This makes it difficult to guage the number of points in a specificregion. A common method of resolving this problem is to 'jitter' thepoints by adding random noise.

This function takes a different approach to the same problem.

When there are two or more points with the same (x,y) value (orwithin x+-s[1] and x+-s[2]), it spaces these out in the x directionso that the points are separated by at least distance s.

Another method for dealing with overploting is available in thesunflowerplot function.

Value

list with two components

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

jitter,sunflowerplot

Examples

   x <- rep(1:5, 10)   y <- round(rnorm(length(x),x))   prepar <- par("mfrow")   par(mfrow=c(1,3))   # standard x-y plot: noverlapping points are hidden   plot(x,y)   title("Standard Plot")    # 'spaced' plot: overlapping points are spread out and visible   plot(space(x,y))   title("Plot with 'space'")   # 'spaced' plot: overlapping points are spread out along y and visible   plot(space(x,y, direction='y'))   title("Plot with 'space', direction='y' ")   # 'sunflower' plot, another approach, overlapping points are   # indicated via petals   sunflowerplot(x,y)   title("Sunflower Plot")      par(mfrow=prepar)

Display text information in a graphics plot.

Description

This function displays text output in a graphics window. It is theequivalent of 'print' except that the output is displayed as a plot.

Usage

textplot(object, halign="center", valign="center", cex, ...)## Default S3 method:textplot(object, halign=c("center","left","right"),         valign=c("center", "top", "bottom"), cex, ... )## S3 method for class 'character'textplot(object, halign = c("center", "left", "right"),         valign = c("center", "top", "bottom"), cex, fixed.width=TRUE,         cspace=1, lspace=1, mar=c(0, 0, 3, 0) + 0.1,         tab.width = 8, ...)## S3 method for class 'data.frame'textplot(object, halign = c("center", "left", "right"),         valign = c("center", "top", "bottom"), cex, ...)## S3 method for class 'matrix'textplot(object, halign = c("center", "left", "right"),         valign = c("center", "top", "bottom"), cex, cmar = 2,         rmar = 0.5, show.rownames = TRUE, show.colnames = TRUE,         hadj = 1, vadj = 1, mar = c(1, 1, 4, 1) + 0.1,         col.data = par("col"), col.rownames = par("col"),          col.colnames = par("col"), ...)

Arguments

Details

A new plot is created and the object is displayedusing the largest font that will fit on in the plotting region. Thehalign andvalign parameters can be used to control thelocation of the string within the plotting region.

For matrixes and vectors a specialized textplot function is available,which plots each of the cells individually, with column widths setaccording to the sizes of the column elements. If present, row andcolumn labels will be displayed in a bold font.

Value

The character scaling factor (cex) used.

Author(s)

Gregory R. Warnesgreg@warnes.net

See Also

plot,text,capture.output

Examples

## Not run: ### simple examples# show R version informationtextplot(version)# show the alphabet as a single stringtextplot( paste(letters[1:26], collapse=" ") )# show the alphabet as a matrixtextplot( matrix(letters[1:26], ncol=2))### Make a nice 4 way display with two plots and two text summariesdata(iris)par(mfrow=c(2,2))plot( Sepal.Length ~ Species, data=iris, border="blue", col="cyan",      main="Boxplot of Sepal Length by Species" )plotmeans( Sepal.Length ~ Species, data=iris, barwidth=2, connect=FALSE,           main="Means and 95% Confidence Intervals\nof Sepal Length by Species")info <- sapply( split(iris$Sepal.Length, iris$Species),                function(x) round(c(Mean=mean(x), SD=sd(x), N=nrow(x)),2) )textplot( info, valign="top"  )title("Sepal Length by Species")reg <- lm( Sepal.Length ~ Species, data=iris )textplot( capture.output(summary(reg)), valign="top")title("Regression of Sepal Length by Species")par(mfrow=c(1,1))### Show how to control text colorcols <- c("red", "green", "magenta", "forestgreen")mat <- cbind(name=cols, t(col2rgb(cols)), hex=col2hex(cols))textplot(mat,         col.data=matrix(cols, nrow=length(cols), byrow=FALSE, ncol=5),         )### Show how to manually tune the character sizedata(iris)reg <- lm( Sepal.Length ~ Species, data=iris )text <- capture.output(summary(reg))# do the plot and capture the character size usedtextplot(text, valign="top")# see what size was usedcex# now redo the plot at 80% sizetextplot( text, valign="top", cex=cex*0.80)## End(Not run)

Plot a Venn diagram

Description

Plot a Venn diagrams for up to 5 sets

Usage

venn(data, universe=NA, small=0.7, showSetLogicLabel=FALSE,     simplify=FALSE, show.plot=TRUE, intersections=TRUE, names,     ...)## S3 method for class 'venn'plot(x, y, ..., small=0.7, showSetLogicLabel=FALSE,     simplify=FALSE)

Arguments

Details

data should be either a named list of vectors containingcharacter string names ("GeneAABBB", "GeneBBBCY", .., "GeneXXZZ") orindexes of group intersections (1, 2, .., N), or a data frame containingindicator variables (TRUE, FALSE, TRUE, ..) for group intersectionship.Group names will be taken from the component list element or columnnames.

Value

Invisibly returns an object of class "venn", containing:

• A matrix of all possible sets of groups, and the observed countof items belonging to each The fist column contains observedcounts, subsequent columns contain 0-1 indicators of groupintersectionship.

• Ifintersections=TRUE, the attributeintersectionswill be a list of vectors containing the names of the elementsbelonging to each subset.

Author(s)

Steffen Moeller, with cleanup and packaging by Gregory R. Warnes.

Examples

#### Example using a list of item names belonging to the## specified group.#### construct some fake gene names..oneName <- function() paste(sample(LETTERS,5,replace=TRUE),collapse="")geneNames <- replicate(1000, oneName())##GroupA <- sample(geneNames, 400, replace=FALSE)GroupB <- sample(geneNames, 750, replace=FALSE)GroupC <- sample(geneNames, 250, replace=FALSE)GroupD <- sample(geneNames, 300, replace=FALSE)input  <-list(GroupA,GroupB,GroupC,GroupD)inputtmp <- venn(input)attr(tmp, "intersections")#### Example using a list of item indexes belonging to the## specified group.##GroupA.i <- which(geneNames %in% GroupA)GroupB.i <- which(geneNames %in% GroupB)GroupC.i <- which(geneNames %in% GroupC)GroupD.i <- which(geneNames %in% GroupD)input.i  <-list(A=GroupA.i,B=GroupB.i,C=GroupC.i,D=GroupD.i)input.ivenn(input.i)#### Example using a data frame of indicator ('f'lag) columns##GroupA.f <- geneNames %in% GroupAGroupB.f <- geneNames %in% GroupBGroupC.f <- geneNames %in% GroupCGroupD.f <- geneNames %in% GroupDinput.df <- data.frame(A=GroupA.f,B=GroupB.f,C=GroupC.f,D=GroupD.f)head(input.df)venn(input.df)## smaller set to create empty groupingssmall <- input.df[1:20,]venn(small, simplify=FALSE) # with empty groupingsvenn(small, simplify=TRUE)  # without empty groupings## Capture group counts, but don't plottmp <- venn(input, show.plot=FALSE)tmp## Show internal binary group labelsvenn(input, showSetLogicLabel=TRUE)## Limit  universetmp <- venn(input, universe=geneNames[1:100])tmp#### Example to determine which elements are in A and B but not in## C and D using the 'intersections' attribute.##tmp <- venn(input, intersection=TRUE)isect <- attr(tmp, "intersection")# Look at all of the subsetsstr(isect)# Extract and combine the subsets of interest..AandB <- unique(c(isect$A, isect$B, isect$'A:B'))# and look at the resultsstr(AandB)#### The full set of elements of each intersection is provided in the## "interesections" attribute.##a<-venn(list(1:5,3:8), show.plot=FALSE)intersections<-attr(a,"intersections")print(intersections)# $A# [1] "1" "2"## $B# [1] "6" "7" "8"## $`A:B`# [1] "3" "4" "5"

Compute the Value of a Function Over a Local Region Of An X-Y Plot

Description

This function applies the specified function to the sets of y valuesthat are defined by overlapping "windows" in the x-dimension. Forexample, settingfun=mean returns local means, while settingfun=function(x) sqrt(var(x)) returns local estimates ofthe standard deviation.

Usage

wapply(x, y, fun=mean, method="range", width, n=50, drop.na=TRUE,       pts, ...)

Arguments

Details

Two basic techniques are available for determining what points fallwithin the same x-neighborhood. The first technique uses a window witha fixed width in the x-dimension and is is selected bysettingmethod="width" ormethod="range". Formethod="width" thewidth argument is an absolutedistance in the x-dimension. Formethod="range", the width isexpressed as a fraction of the x-range. In both cases,ptsspecifies the points at which evaluation offun occurs. Whenpts is omitted,n x values equally spaced along the xrange are used.

The second technique uses windows containing k neighboring points. The(x,y) pairs are sorted by the x-values and the nearest k/2 points withhigher x values and the k/2 nearest points with lower x values areincluded in the window. Whenmethod="nobs", k equalswidth (actually 2*floor(width/2) ). Whenmethod="fraction",width specifies what fraction of thetotal number of points should be included. The actual number of pointsincluded in each window will be floor(n*frac/2)*2. Regardless of thevalue ofpts, the functionfun will be evaluated at allx locations.

Value

Returns a list with components

Author(s)

Gregory R. Warnesgreg@warnes.net

Examples

#show local mean and inner 2-sd interval to help diagnose changing mean#or variance structurex <- 1:1000y <- rnorm(1000, mean=1, sd=1 + x/1000 )plot(x,y)lines(wapply(x,y,mean),col="red")CL <- function(x,sd) mean(x)+sd*sqrt(var(x))lines(wapply(x,y,CL,sd= 1),col="blue") lines(wapply(x,y,CL,sd=-1),col="blue") lines(wapply(x,y,CL,sd= 2),col="green")lines(wapply(x,y,CL,sd=-2),col="green")#show local mean and inner 2-sd interval to help diagnose changing mean#or variance structurex <- 1:1000y <- rnorm(1000, mean=x/1000, sd=1)plot(x,y)lines(wapply(x,y,mean),col="red")CL <- function(x,sd) mean(x)+sd*sqrt(var(x))lines(wapply(x,y,CL,sd= 1,method="fraction",width=1/20),col="blue")lines(wapply(x,y,CL,sd=-1,method="fraction",width=1/20),col="blue")lines(wapply(x,y,CL,sd= 2,method="nobs",width=250),col="green")lines(wapply(x,y,CL,sd=-2,method="nobs",width=250),col="green")