An embedded and charset-unspecified text was scrubbed... Name: not available Url: https://stat.ethz.ch/pipermail/r-help/attachments/20050410/fc61772c/attachment.pl
Plotting the occassional second label
3 messages · Lisbeth Riis, Peter Dalgaard, Rich FitzJohn
"Lisbeth Riis" <lriis at scarabconsult.com> writes:
Dear useRs, I'm trying to plot spray quantities against dates, and label the points on the plot. Basically quite simple, but sometimes two chemicals have been used and are listed in separate rows in the table as below; then the labels are written on top of each other.
spray
SprayDate PD Trt Qwater Qai 1 2005-03-09 Spidermites Pegasus 1300 1.04 2 2005-03-10 Powdery MKP 800 0.40 3 2005-03-10 Powdery Nimrod 250 EC 800 2.40 4 2005-03-12 Spidermites Pegasus 1300 1.04 5 2005-03-16 Spidermites Pegasus 1300 1.04 6 2005-03-17 Powdery Stroby 50 WG 800 0.40 7 2005-03-21 Spidermites Dynamec 1.8 EC 1250 0.62 8 2005-03-21 Spidermites Apollo 50 SC 1250 0.62 9 2005-03-30 Whiteflies Pegasus 1000 0.60 I came up with the work around below which gave me my plot for some of the data (despite warnings). The idea was to split the second chemical for each date into a separate file and plot the text a little higher:
spray1 <- spray[spray$PD=="Spidermites", ] spray2 <- spray1[spray1$SprayDate==unique(spray1$SprayDate), ]
Warning message:
longer object length
is not a multiple of shorter object length in: spray1$SprayDate == unique(spray1$SprayDate)
I don't think this does what I think you think it does... Consider
x <- c(1,1,2,3) x == unique(x)
[1] TRUE FALSE FALSE FALSE
Warning message:
longer object length
is not a multiple of shorter object length in: x == unique(x)
unique(x) is c(1,2,3) so you end up with c(1,1,2,3)==c(1,2,3,1)
Were you perhaps intending
!duplicated(x)
[1] TRUE FALSE TRUE TRUE ?
O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907
Another option is to combine the sprays used as a single string, and
plot that directly, rather than subsetting the data. This has the
advantages of not having to worry about how much to offset the second
label by, and should also work if you got more than two sprays per
day.
spray$SprayDate <- as.Date(spray$SprayDate)
spray.sub <- spray[spray$PD=="Spidermites",]
## Collapse, so there is only a single row per date
spray.col <- spray.sub[unique(tapply(spray.sub$SprayDate,
spray.sub$SprayDate)),]
## And paste together any treatments used on a single day, separated
## by a newline
txt <- tapply(spray.sub$Trt, spray.sub$SprayDate, paste,
collapse="\n")
plot(spray.col$SprayDate, spray.col$Qwater,
xlim=c(as.Date("2005-03-08"), as.Date("2005-03-24")),
ylim=c(0,1500))
text(spray.col$SprayDate, spray.col$Qwater, txt, pos=4, srt=45)
If you wanted the order of the chemicals used changed, you could
insert a new function into the tapply() call, e.g.
tapply(spray.sub$Trt, spray.sub$SprayDate,
function(x) paste(rev(x), collapse="\n"))
Cheers,
Rich
On Apr 10, 2005 8:24 PM, Lisbeth Riis <lriis at scarabconsult.com> wrote:
Dear useRs, I'm trying to plot spray quantities against dates, and label the points on the plot. Basically quite simple, but sometimes two chemicals have been used and are listed in separate rows in the table as below; then the labels are written on top of each other.
...
Does anyone have a better way of splitting the second reading for each date off, or another way of printing two labels for one plotted point?
Rich FitzJohn rich.fitzjohn <at> gmail.com | http://homepages.paradise.net.nz/richa183 You are in a maze of twisty little functions, all alike