[R] lattice: double y - problem changing axis color after doubleYScale
Duncan Mackay
dulcalma at bigpond.com
Tue Sep 24 11:32:33 CEST 2013
Hi Anna
I have never used doubleYScale and I am not sure what you want but try this
You need to do this first if columns not correct saves doing later - I
forgot to mention it for Week but as you have given the order for the
factors do it as well
SummPdata[,1] <- as.numeric(SummPdata[,1])
SummNdata[,1] <- as.numeric(SummNdata[,1])
SummPdata[,3] <- factor(SummPdata[,3], levels = c('B1', 'H2', 'H3', 'H4'))
SummNdata[,3] <- factor(SummNdata[,3], levels = c('B1', 'H2', 'H3', 'H4'))
As the data has a similar structure combine data
summData = rbind(SummPdata, SummNdata)
summData$element <- rep(c("P","N"), each = 32)
Just quickly cut and pasted the colours etc
useOuterStrips(strip = strip.custom(#factor.levels = ,
par.strip.text =
list(fontfamily="serif" )),
strip.left = strip.custom(#factor.levels = ,
par.strip.text =
list(fontfamily="serif" )),
xyplot(mean ~ Week | Station*interaction(element,fyear),
data= summData,
cex=1.1,
par.settings = list(strip.background = list(col =
"transparent"),
add.text = list(fontfamily = "serif")),
#strip = strip.custom(#factor.levels = ,
# par.strip.text =
list(fontfamily="serif" )),
scales = list(alternating = c(3,3),
fontfamily = "serif"),
ylab = list(text = "Pc", fontfamily = "serif"),
xlab=list(text = "Week", fontfamily = "serif"),
key = list(text = list(labels = c("P","N")),
points = list(pch = c(2,20), col =
c("red","blue")),
lines = list(lty = c(1,2)), col =
c("red","blue")),
type = "o",
ylim = c(0,70),
subscripts = TRUE,
panel = function(x,y, subscripts, ...){
pnl = panel.number()
panel.xyplot(x,y, type = "o", col =
c("red","blue")[ifelse(pnl <9, 1,2)], lty = c(1,2)[ifelse(pnl <9, 1,2)], pch
= c(2,20)[ifelse(pnl <9, 1,2)])
# sd
# upper
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y+ summData[subscripts,
"sd"]/2, angle = 90, length = 0.04, col = c("red","blue")[ifelse(pnl <9,
1,2)])
# lower
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y- summData[subscripts,
"sd"]/2, angle = 90, length = 0.04, col = c("red","blue")[ifelse(pnl <9,
1,2)])
}# panel
)#
) ## useOuterStrips
You can fill in the LHS strip with a vector of names as required
If you want to amend the ylimits see ?combineLimits in the latticeExtra
package
Duncan
-----Original Message-----
From: Anna Zakrisson Braeunlich [mailto:anna.zakrisson at su.se]
Sent: Tuesday, 24 September 2013 18:28
To: Duncan Mackay
Subject: RE: [R] lattice: double y - problem changing axis color after
doubleYScale
OK, now the code works - thank you. The error bars are added beautifully!
As I am using doubleYScale, I am having the problem that the first y-axis is
covered by the left hand strip - how can I solve this? I really much prefer
using outer.strip() and have one strip vertically on the plot. It saves plot
space.
I am still having the problem of colourful y-axis when using doubleYScale()
- how can I make them black?
I have the same problem with the error bars - they are also default colour
no matter what I tell them to be.
Is there a way around this except Adobe Photoshop?
Here is the working script with dummy data (you can just run the lot if you
have time). I really appreciate your code - thanks! ...and sorry for being a
bit daft and sending a reply before thinking it through properly.
Ndata <- data.frame(
Ncellpercent = rnorm(400, mean = rep(c(14, 18, 65), each = 40),
sd = rep(c(1, 3, 6), each = 40)),
fyear = rep(c('2007', '2008'), each = 100*2),
Station = sample(c('B1', 'H2', 'H3', 'H4'), 400, replace = TRUE),
Week = sample(c('19', '21', '23', '25'), 400, replace = TRUE))
Pdata <- data.frame(
Ppercentcell = rnorm(400, mean = rep(c(4, 17, 22), each = 40),
sd = rep(c(0.1, 0.2, 0.4), each = 40)),
fyear = rep(c('2007', '2008'), each = 100*2),
Station = sample(c('B1', 'H2', 'H3', 'H4'), 400, replace = TRUE),
Week = sample(c('19', '21', '23', '25'), 400, replace = TRUE))
SummNdata <- ddply(Ndata, .(Week, fyear, Station), summarise,
mean = mean(Ncellpercent),
sd = sd(Ncellpercent))
names(Pdata)
SummPdata <- ddply(Pdata, .(Week, fyear, Station), summarise,
mean = mean(Ppercentcell),
sd = sd(Ppercentcell)) SummPdata
plotNtest <- useOuterStrips(strip = strip.custom(factor.levels =
c("B1", "H2", "H3", "H4"),
par.strip.text =
list(fontfamily="serif" )),
strip.left = strip.custom(factor.levels = c("2007", "2008",
"2009"),
par.strip.text =
list(fontfamily="serif" )),
xyplot(mean ~ Week | Station*fyear, data= SummNdata,
col="black",
pch=1,
cex=1.1,
lty=1,
par.settings = list(strip.background = list(col =
"transparent"),
add.text = list(fontfamily =
"serif"
)),
strip = strip.custom(factor.levels = c("B1",
"H2", "H3", "H4"),
par.strip.text =
list(fontfamily="serif" )),
scales = list(alternating = c(3,3),
fontfamily = "serif"),
ylab = list(text = "Nc", fontfamily = "serif"),
xlab=list(text = "Week", fontfamily = "serif"),
key = list(text = list(labels = c("Pc","Nc")),
points = list(pch = c(1,2)),
lines = list(lty = c(1,2))),
type = "o",
ylim = c(0,70),
subscripts = TRUE,
panel = function(x,y, subscripts, ...){
panel.xyplot(x,y, ...)
# sd
# upper
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y+ SummNdata[subscripts,
"sd"]/2, angle = 90,
length = 0.04)
# lower
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y- SummNdata[subscripts,
"sd"]/2, angle = 90,
length = 0.04)
panel.xyplot(x+0.1, SummNdata[subscripts,"mean"],
type ="o",pch = 20, lty =1, col = "black")
# sd
# upper
panel.arrows(x0 = x+0.1, x1 = x+0.1,
y0 = SummNdata[subscripts,"mean"], y1
= SummNdata[subscripts,"mean"] +
SummNdata[subscripts, "sd"]/2, angle = 90,
length = 0.04)
# lower
panel.arrows(x0 = x+0.1, x1 = x+0.1,
y0 = SummNdata[subscripts,"mean"], y1
= SummNdata[subscripts,"mean"]-
SummNdata[subscripts, "sd"]/2, angle = 90,
length = 0.04)
}# panel
)#
) ## useOuterStrips
plotPtest <- useOuterStrips(strip = strip.custom(factor.levels =
c("B1", "H2", "H3", "H4"),
par.strip.text =
list(fontfamily="serif"
)),
strip.left = strip.custom(factor.levels = c("2007",
"2008", "2009"),
par.strip.text =
list(fontfamily="serif"
)),
xyplot(mean ~ Week | Station*fyear, data= SummPdata,
col="black",
pch=2,
cex=1.1,
lty=2,
par.settings = list(strip.background =
list(col =
"transparent"),
add.text = list(fontfamily
= "serif"
)),
strip = strip.custom(factor.levels =
c("B1", "H2", "H3", "H4"),
par.strip.text =
list(fontfamily="serif" )),
scales = list(alternating = c(3,3),
fontfamily = "serif"),
ylab = list(text = "Pc", fontfamily =
"serif"),
xlab=list(text = "Week", fontfamily =
"serif"),
type = "o",
ylim = c(0,70),
subscripts = TRUE,
panel = function(x,y, subscripts, ...){
panel.xyplot(x,y, ...)
# sd
# upper
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y+
SummPdata[subscripts,
"sd"]/2, angle = 90, length = 0.04)
# lower
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y-
SummPdata[subscripts,
"sd"]/2, angle = 90, length = 0.04)
panel.xyplot(x+0.1,
SummPdata[subscripts,"mean"],
type ="o",pch = 20, lty =1, col
= "black")
# sd
# upper
panel.arrows(x0 = x+0.1, x1 = x+0.1,
y0 =
SummPdata[subscripts,"mean"], y1
= SummPdata[subscripts,"mean"]
+ SummPdata[subscripts, "sd"]/2, angle = 90,
length = 0.04)
# lower
panel.arrows(x0 = x+0.1, x1 = x+0.1,
y0 =
SummPdata[subscripts,"mean"], y1
= SummPdata[subscripts,"mean"]-
SummPdata[subscripts, "sd"]/2, angle = 90,
length = 0.04)
}# panel
)#
) ## useOuterStrips
plotNtest
plotPtest
doubleYScale(plotNtest, plotPtest, add.ylab2 = TRUE)
Can I make the
Anna Zakrisson Braeunlich
PhD student
Department of Ecology, Environment and Plant Sciences Stockholm University
Svante Arrheniusv. 21A
SE-106 91 Stockholm
Sweden/Sverige
Lives in Berlin.
For paper mail:
Katzbachstr. 21
D-10965, Berlin - Kreuzberg
Germany/Deutschland
E-mail: anna.zakrisson at su.se
Tel work: +49-(0)3091541281
Mobile: +49-(0)15777374888
LinkedIn: http://se.linkedin.com/pub/anna-zakrisson-braeunlich/33/5a2/51b
><((((º>`•. . • `•. .• `•. . ><((((º>`•. . • `•. .• `•. .><((((º>`•. . •
>`•. .• `•. .><((((º>
________________________________________
From: Duncan Mackay [dulcalma at bigpond.com]
Sent: 22 September 2013 05:00
To: Anna Zakrisson Braeunlich
Cc: R
Subject: RE: [R] lattice: double y - problem changing axis color after
doubleYScale
Hi Anna
I am not sure what you want but the following should get you part of the way
I prefer to keep all my parameters within the function rather than use
themes.
useOuterStrips(strip = strip.custom(#factor.levels = ,
par.strip.text =
list(fontfamily="serif" )),
strip.left = strip.custom(#factor.levels = ,
par.strip.text =
list(fontfamily="serif" )), xyplot(mean ~ Week | Station*fyear, data=
SummPdata,
col="black",
pch=2,
cex=1.1,
lty=2,
par.settings = list(strip.background = list(col =
"transparent"),
add.text = list(fontfamily = "serif"
)),
#strip = strip.custom(#factor.levels = ,
# par.strip.text =
list(fontfamily="serif" )),
scales = list(alternating = c(3,3),
fontfamily = "serif"),
ylab = list(text = "Pc", fontfamily = "serif"),
xlab=list(text = "Week", fontfamily = "serif"),
key = list(text = list(labels = c("P","N")),
points = list(pch = c(2,20)),
lines = list(lty = c(1,2))),
type = "o",
ylim = c(0,70),
subscripts = TRUE,
panel = function(x,y, subscripts, ...){
panel.xyplot(x,y, ...)
# sd
# upper
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y+ SummPdata[subscripts,
"sd"]/2, angle = 90, length = 0.04)
# lower
panel.arrows(x0 = x, x1 = x,
y0 = y, y1 = y- SummPdata[subscripts,
"sd"]/2, angle = 90, length = 0.04)
panel.xyplot(x+0.1, SummNdata[subscripts,"mean"],
type ="o",pch = 20, lty =1, col = "black")
# sd
# upper
panel.arrows(x0 = x+0.1, x1 = x+0.1,
y0 = SummNdata[subscripts,"mean"], y1
= SummNdata[subscripts,"mean"] + SummNdata[subscripts, "sd"]/2, angle = 90,
length = 0.04)
# lower
panel.arrows(x0 = x+0.1, x1 = x+0.1,
y0 = SummNdata[subscripts,"mean"], y1
= SummNdata[subscripts,"mean"]- SummNdata[subscripts, "sd"]/2, angle = 90,
length = 0.04)
}# panel
)#
) ## useOuterStrips
Note that I offset the N data so that the error bars are more legible
For sd I just divided them by 2 so you will have to work out what you
require
Regards
Duncan
Duncan Mackay
Department of Agronomy and Soil Science
University of New England
Armidale NSW 2351
-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On
Behalf Of Anna Zakrisson Braeunlich
Sent: Thursday, 19 September 2013 20:21
To: r-help at r-project.org
Subject: [R] lattice: double y - problem changing axis color after
doubleYScale
Hi,
I have had some troubles using doubleYScale. No matter what I try, I cant
manage to change the color of the y-axis in the end. I have to produce a
black and white plot. There is also something I do not understand regarding
fontfamilyj="serif" when using it in:
strip=strip.custom()
Maybe someone has a better idea for defining which line and dots belong to
which y-axis when not using a colorcode than the one I had.
I have annotated my questions in the code below.
Thank you for your time!
Here is some dummy data:
Ndata <- data.frame(
Ncellpercent = rnorm(400, mean = rep(c(14, 18, 65), each = 40),
sd = rep(c(1, 3, 6), each = 40)),
fyear = rep(c('2007', '2008'), each = 100*2),
Station = sample(c('B1', 'H2', 'H3', 'H4'), 400, replace = TRUE),
Week = sample(c('19', '21', '23', '25'), 400, replace = TRUE))
Pdata <- data.frame(
Ppercentcell = rnorm(400, mean = rep(c(4, 17, 22), each = 40),
sd = rep(c(0.1, 0.2, 0.4), each = 40)),
fyear = rep(c('2007', '2008'), each = 100*2),
Station = sample(c('B1', 'H2', 'H3', 'H4'), 400, replace = TRUE),
Week = sample(c('19', '21', '23', '25'), 400, replace = TRUE))
SummNdata <- ddply(Ndata, .(Week, fyear, Station), summarise,
mean = mean(Ncellpercent),
sd = sd(Ncellpercent))
names(Pdata)
SummPdata <- ddply(Pdata, .(Week, fyear, Station), summarise,
mean = mean(Ppercentcell),
sd = sd(Ppercentcell))
library(lattice)
library(latticeExtra)
library(HH)
font.settings <- list( font = 1, cex = 1.2, fontfamily = "serif")
my.theme <- list(
par.xlab.text = font.settings,
par.ylab.text = font.settings,
axis.text = font.settings,
par.sub=font.settings)
plotN <- xyplot(mean ~ Week | Station*fyear,
col="black",
pch=1,
cex=1.1,
lty=1,
strip = strip.custom(bg = 'white', style=1), # why can I not
use fontfamily="serif" here ???
key=list(text=list(c(""),
col=c("black")),
points=list(pch=1, lty=1, cex=1.5,
col=c("black")),
columns=1, border=F,
x = 0.02, y = 0.55, corner = c(2, 2),
title="", cex.title=1.3),
ylab = ("Nc"),
xlab="Week",
data= SummNdata,type="o",
par.settings = my.theme) plotN # I would like to add the
standard deviations (sd) to the plot. I have tried some stuff, # but for
some reason, it does not seem to work. How would I go about this?
plotP <- xyplot(mean ~ Week | Station*fyear,
col="black",
pch=2,
cex=1.1,
lty=2,
strip = strip.custom(bg = 'white', style=1), # why can I not
use fontfamily="serif" here ???
key=list(text=list(c(""),
col=c("black")),
points=list(pch=1, lty=1, cex=1.5,
col=c("black")),
columns=1, border=F,
x = 0.2, y = 0.2, corner = c(2, 2),
title="", cex.title=1.3),
ylab = ("Pc"),
xlab="Week",
data= SummPdata,type="o",
par.settings = my.theme) plotP
doubleYScale(plotN, plotP, add.ylab2 = TRUE) #Why can I not change the axis
color by adding to this argument?
# I want the y1 and y2 axes to be defined not by color, but by shape and
linetype.
# I have managed to draw the shapes (defined by Nc and Pc) by the y1 and y2
axes, but I do not manage to get the lines # though the shape - ideas?
# Alternative ways that are not based on color (I have to do this black and
white).
# Is there possible to add shapes to the axis text? such as:
# --O-- Nc on the left y-axis (but with lty=1: I could not do a non-dotted
line on the keybord).
with kind regards
Anna Zakrisson Braeunlich
PhD student
Department of Ecology, Environment and Plant Sciences Stockholm University
Svante Arrheniusv. 21A
SE-106 91 Stockholm
Sweden/Sverige
Lives in Berlin.
For paper mail:
Katzbachstr. 21
D-10965, Berlin - Kreuzberg
Germany/Deutschland
E-mail: anna.zakrisson at su.se
Tel work: +49-(0)3091541281
Mobile: +49-(0)15777374888
LinkedIn: http://se.linkedin.com/pub/anna-zakrisson-braeunlich/33/5a2/51b
><((((:>`?. . ? `?. .? `?. . ><((((:>`?. . ? `?. .? `?. .><((((:>`?. . ?
>`?. .? `?. .><((((:>
[[alternative HTML version deleted]]
More information about the R-help
mailing list