[R] set axis limit in lattice
Duncan Mackay
dulcalma at bigpond.com
Fri May 9 06:17:09 CEST 2014
Luigi
I suspect that it is in accessing the correct groups within the panel function
I suggest that printing the values before plotting within the panel function may give you some idea.
Without specific data I cannot suggest anything else
Duncan
-----Original Message-----
From: Luigi Marongiu [mailto:marongiu.luigi at gmail.com]
Sent: Wednesday, 7 May 2014 22:37
To: Duncan Mackay
Subject: Re: [R] set axis limit in lattice
Dear Duncan,
thank you for the reply. I have fixed the axis, bu still I don't
understand why i cannot draw the segments on my real data whereas i
could on the example when the panel function is quite general. igra is
the outcome that i rendered as positivity in the example. i also
change it into numeric values (1/0) to resemble more closely the
example but still the segments are ignored.
may i ask you for a tip on the possible causes?
best regards
luigi
On Thu, May 1, 2014 at 3:00 AM, Duncan Mackay <dulcalma at bigpond.com> wrote:
> If you just want to limit the axis just insert as arguments
> for the x axis
> xlim = c(min, max),
> and for the y
> ylim = c(min,max) ,
> where min and max correspond to the appropriate min and max for the x and y axes
>
> As for the panel.groups you mention medians where I was using means - if you want medians you will have to amend means to medians
>
> Make sure for the groups you have the right mode numeric/ factor ?
> I do not know what igra is
>
> Duncan
>
> -----Original Message-----
> From: Luigi Marongiu [mailto:marongiu.luigi at gmail.com]
> Sent: Thursday, 1 May 2014 03:54
> To: Duncan Mackay
> Subject: Re: [R] set axis limit in lattice
>
> Dear Duncan,
> these solutions are actually splitting the drawing in different panels
> and the scale=free is giving different limits to each panel. I still
> need the relation=same but just to input myself the fixed limits of
> the axis. It should go into the scale argument but i don't know the
> actual syntax.
> As for the segments, you were perfectly right, there was a problem
> with my script so the medians did not match the actual data. Your
> solution using panel.groups works on the example, but when I copied it
> to the actual dataframe I am working with it did not work despite the
> fact of being of quite a general syntax. could you please give a look
> at the dataframe I have attached and see where I made a mistake in
> applying your solution? what I wrote is:
>
> my.data.IP10<-subset(my.data, gene=="i" & stimulus != "Unstimulated")
>
> stripplot(
> my.data.IP10$ratio ~
> factor(my.data.IP10$igra)|factor(my.data.IP10$stimulus, levels =
> c("ESAT6","CFP10","Rv3615c", "Rv2654","Rv3879", "Rv3873","PHA")),
> my.data,
> group = my.data.IP10$igra,
> hor=F,
> layout = c(7,1),
> scales = list(x = list(draw=FALSE)),
> jitter.data=TRUE,
> pch=c(1,16),
> col="black",
> ylab=expression(bold("Ratio GOI/RG")),
> xlab=expression(bold("Antigen stimulation")),
> main="IP10 expression",
> par.settings = list(strip.background=list(col="light grey")),
> par.strip.text=list(font=2),
> key = list(space="top", columns=2, text=list(c("IGRA Negative",
> "IGRA Positive"), col="black"), points=list(pch=c(1,16),
> col="black")),
>
> panel = panel.superpose,
> panel.groups = function(x, y, group.number, ...) {
>
> panel.segments(x0= x-0.25, x1=x+0.25,
> y0=mean(y), y1=mean(y), lty = c(1:2)[group.number],
> lwd=c(2:1)[group.number], col = c(1:2)[group.number])
>
> panel.stripplot(x,y, ...)
> }
> )
>
>
> ### Many thanks
> ### Best regards
> Luigi
>
>
> On Wed, Apr 30, 2014 at 12:15 AM, Duncan Mackay <dulcalma at bigpond.com> wrote:
>> Hi Luigi
>>
>> This produces plenty of white space for the y axes
>>
>> Y<-max(my.data$copy)
>>
>> stripplot(
>> copy ~ factor(positivity)|factor(stimulation, levels = c("Unst.",
>> "ESAT6","CFP10","Rv3615c", "Rv2654","Rv3879", "Rv3873","PHA")),
>> my.data,
>> group = positivity,
>> hor=F,
>> layout = c(8,1),
>> scales = list(x = list(at = c(1,2), labels = c("N","P")),
>> y= list(relation = "free")),
>> jitter.data=TRUE, pch=c(16,1), col="black",
>> ylab=expression(bold("Copy")),
>> xlab=expression(bold("Stimulation")),
>> main="Plot",
>> par.settings = list(strip.background=list(col="white")),
>> par.strip.text=list(font=2),
>> key = key.plot,
>> panel = function(x, y, ...)
>> {
>> pnl = panel.number()
>> k<-0
>> for (i in 1:2){
>> datme<-subset(datmeA, positivity==i-1)
>> w<-i+k
>> panel.segments(x0 = X[w], y0 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],
>> x1 = X[w+1], y1 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],lwd = 2, col = "black")
>> k=k+1
>>
>> panel.stripplot(x,y, ...)
>> }
>> }
>> )
>> You could limit the y-axis to your levels
>> stripplot(
>> copy ~ factor(positivity)|factor(stimulation, levels = c("Unst.",
>> "ESAT6","CFP10","Rv3615c", "Rv2654","Rv3879", "Rv3873","PHA")),
>> my.data,
>> group = positivity,
>> hor=F,
>> layout = c(8,1),
>> scales = list(x = list(at = c(1,2), labels = c("N","P")),
>> y= list(relation = "free",
>> limits = list(c(0, max panel 1),
>> c(0, max panel 2),
>> ...
>> c(0, max panel 8)),
>> )),
>> jitter.data=TRUE, pch=c(16,1), col="black",
>> ylab=expression(bold("Copy")),
>> xlab=expression(bold("Stimulation")),
>> main="Plot",
>> par.settings = list(strip.background=list(col="white")),
>> par.strip.text=list(font=2),
>> key = key.plot,
>> panel = function(x, y, ...)
>> {
>> pnl = panel.number()
>> k<-0
>> for (i in 1:2){
>> datme<-subset(datmeA, positivity==i-1)
>> w<-i+k
>> panel.segments(x0 = X[w], y0 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],
>> x1 = X[w+1], y1 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],lwd = 2, col = "black")
>> k=k+1
>>
>> panel.stripplot(x,y, ...)
>> }
>> }
>> )
>>
>> Sorting the stimulation groups shows that some of them can be loosely grouped so reordering the factors
>>
>> stripplot(copy ~ factor(positivity)|factor(stimulation, levels = c("PHA","Rv2654","CFP10","Unst.","Rv3615c", "Rv3873","ESAT6", "Rv3879")),
>> my.data,
>> group = positivity,
>> hor=F,
>> layout = c(2,4),
>> scales = list(x = list(at = c(1,2), labels = c("N","P")),
>> y= list(relation = "free")),
>> jitter.data=TRUE, pch=c(16,1), col="black",
>> ylab=expression(bold("Copy")),
>> xlab=expression(bold("Stimulation")),
>> main="Plot",
>> par.settings = list(strip.background=list(col="white")),
>> par.strip.text=list(font=2),
>> key = key.plot,
>> panel = function(x, y, ...)
>> {
>> pnl = panel.number()
>> k<-0
>> for (i in 1:2){
>> datme<-subset(datmeA, positivity==i-1)
>> w<-i+k
>> panel.segments(x0 = X[w], y0 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],
>> x1 = X[w+1], y1 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],lwd = 2, col = "black")
>> k=k+1
>>
>> panel.stripplot(x,y, ...)
>> }
>> }
>> )
>>
>> or using panel.groups
>>
>> stripplot(copy ~ factor(positivity)|factor(stimulation, levels = c("PHA","Rv2654","CFP10","Unst.","Rv3615c", "Rv3873","ESAT6", "Rv3879")),
>> my.data,
>> group = positivity,
>> hor=F,
>> layout = c(2,4),
>> scales = list(x = list(at = c(1,2), labels = c("N","P")),
>> y= list(relation = "free")),
>> jitter.data=TRUE, pch=c(16,1), col="black",
>> ylab=expression(bold("Copy")),
>> xlab=expression(bold("Stimulation")),
>> main="Plot",
>> par.settings = list(strip.background=list(col="white")),
>> par.strip.text=list(font=2),
>> key = key.plot,
>> panel = panel.superpose,
>> panel.groups = function(x, y, group.number, ...) {
>>
>> panel.segments(x0= x-0.25, x1=x+0.25, y0=mean(y), y1=mean(y), lty = c(1:2)[group.number],
>> lwd=c(2:1)[group.number], col = c(1:2)[group.number])
>>
>> panel.stripplot(x,y, ...)
>> }
>> )
>>
>> Regards
>>
>> Duncan Mackay
>>
>> -----Original Message-----
>> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Luigi Marongiu
>> Sent: Wednesday, 30 April 2014 03:58
>> To: r-help at r-project.org
>> Subject: [R] set axis limit in lattice
>>
>> Dear all,
>>
>> I would like to set the axis of a figure using the max() so to have
>> more control on the limits of the axis -- this because in the actual
>> case more independent figures are generated from the same dataframe
>> and these must have the same axis scale.
>>
>> Since the figure is generated using lattice device, the syntax is too
>> complicated for me: when I place the xlim=c(0,Y) in the arguments —
>> where Y is obtained using max(data) — there is no result. In basic R
>> this argument would work virtually in any position.
>>
>> Any tip?
>>
>> Best regards
>>
>> Luigi
>>
>> ####CODE
>> ### open plot library
>> library(lattice)
>> my.data<-structure(list(
>> column_1 = 1:120,
>> column_2 = structure(c(
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8,
>> 1,2,3,4,5,6,7,8), .Label = c("Unst.", "ESAT6", "CFP10", "Rv3615c",
>> "Rv2654", "Rv3879", "Rv3873", "PHA"), class = "factor"),
>> column_3 = structure(c(
>> 0,0,0,0,0,0,0,0,
>> 0,0,0,0,0,0,0,0,
>> 0,0,0,0,0,0,0,0,
>> 0,0,0,0,0,0,0,0,
>> 0,0,0,0,0,0,0,0,
>> 0,0,0,0,0,0,0,0,
>> 0,0,0,0,0,0,0,0,
>> 1,1,1,1,1,1,1,1,
>> 1,1,1,1,1,1,1,1,
>> 1,1,1,1,1,1,1,1,
>> 1,1,1,1,1,1,1,1,
>> 1,1,1,1,1,1,1,1,
>> 1,1,1,1,1,1,1,1,
>> 1,1,1,1,1,1,1,1,
>> 0,0,0,0,0,0,0,0)),
>> column_4 = c(
>> 192.0519108,183.6403531,53.46798757,83.60638077,69.60749873,159.4706861,256.8765622,499.2899303,
>> 2170.799076,1411.349719,2759.472348,2098.973397,2164.739515,1288.676574,1611.486543,6205.229575,
>> 870.7424981,465.9967135,191.8962375,864.0937485,2962.693675,1289.259137,2418.651212,7345.712517,
>> 0,168.1198893,674.4342961,101.1575401,47.81596237,0,0,1420.793922,
>> 142.6871331,5.466468742,291.9564635,80.73914133,73.02239621,64.47806871,144.3543635,3167.959757,
>> 3164.748333,1092.634557,28733.20269,1207.87783,729.6090973,151.8706088,241.2466141,9600.963594,
>> 1411.718287,12569.96285,1143.254476,6317.378481,16542.27718,79.68025792,1958.495138,7224.503437,
>> 208.4382941,69.48609769,656.691151,0.499017582,7114.910926,187.6296174,41.73980805,8930.784541,
>> 4.276752185,0.432300363,60.89228665,1.103924786,0.490686366,1.812993239,7.264531581,1518.610307,
>> 2172.051528,595.8513744,17141.84336,589.6565971,1340.287628,117.350942,593.7034054,24043.61463,
>> 0,81.83292179,1539.864321,36.41722958,8.385131047,161.7647376,65.21615696,7265.573875,
>> 97.84753179,154.051827,0.613835842,10.06138851,45.04879285,176.8284258,18795.75462,30676.769,
>> 5780.34957,944.2200834,2398.235596,1083.393165,2541.714557,1251.670895,1547.178549,1792.679176,
>> 3067.988416,8117.210173,23676.02226,8251.937547,17360.80494,18563.61561,16941.865,31453.96708,
>> 2767.493803,4796.33016,12292.93705,3864.657567,9380.673835,14886.44683,8457.88646,26050.47191)),
>> .Names = c("row", "stimulation", "positivity", "copy"), row.names =
>> c(NA, -120L),
>> class = "data.frame")
>>
>> key.plot<-list(
>> space="top", columns=2,
>> text=list(c("Positive", "Negative"), col="black"),
>> points=list(pch=c(16,1), col="black"))
>>
>> datmeA <- aggregate(copy ~ positivity+stimulation, my.data, median, na.rm = T)
>>
>> X<-c(0.7, 1.3, 1.7, 2.3)
>>
>> Y<-max(copy)
>>
>> stripplot(
>> copy ~ factor(positivity)|factor(stimulation, levels = c("Unst.",
>> "ESAT6","CFP10","Rv3615c", "Rv2654","Rv3879", "Rv3873","PHA")),
>> my.data,
>> group = positivity,
>> hor=F,
>> layout = c(8,1),
>> scales = list(x = list(at = c(1,2), labels = c("N","P"))),
>> jitter.data=TRUE, pch=c(16,1), col="black",
>> ylab=expression(bold("Copy")),
>> xlab=expression(bold("Stimulation")),
>> main="Plot",
>> par.settings = list(strip.background=list(col="white")),
>> par.strip.text=list(font=2),
>> key = key.plot,
>> panel = function(x, y, ...)
>> {
>> pnl = panel.number()
>> k<-0
>> for (i in 1:2){
>> datme<-subset(datmeA, positivity==i-1)
>> w<-i+k
>> panel.segments(x0 = X[w], y0 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],
>> x1 = X[w+1], y1 = datme[datme[,2]==
>> levels(datme[,2])[pnl],"copy"],lwd = 2, col = "black")
>> k=k+1
>>
>> panel.stripplot(x,y, ...)
>> }
>> }
>> )
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>
More information about the R-help
mailing list