Dear List:
I'm creating a series of barplots using Sweave that must assume a
standard format. This is student achievement data and the x-axis must
include all grades 3 to 8. In some cases, the data for a grade (or more
than one grade) are missing in the vector math.bar, but are never
missing for the vector apmxpmeet. The following sample code illustrates
the issue.
Using the code below to create the plot works fine. However, the
following command is designed to place the data onto the plot:
text(math.barplot, graph.max+5, format(tot), xpd = TRUE, col =
c("blue",
"orange") )
This does work, but, it also places 'NA's on the plot. Is there a way to
place the data onto the plot such that the numbers appear, but for the
NAs not to appear? I've searched through ?text but was not able to find
a solution.
math.bar <- c(NA,NA,NA,NA,55,48)
apmxpmeet <- c(47, 50, 49, 50, 49, 46)
par(ps=10)
math.bar <- rbind(math.bar, apmxpmeet)
math.barplot <- barplot(math.bar, beside=T,
col=c('blue','orange'),
names=c('Grade \n 3','Grade \n 4','Grade \n
5','Grade \n 6','Grade \n
7','Grade \n 8'),
ylim=c(0,120), ylab="Percentage", xlab="Grade Level")
tot <- round(math.bar,digits=0)
graph.max <- max(math.bar, apmxpmeet, na.rm=T)
text(math.barplot, graph.max+5, format(tot), xpd = TRUE, col =
c("blue",
"orange") )
tmp$sch_nameS05 <- as.character(tmp$sch_nameS05)
legend(2,120,legend=(c(tmp$sch_nameS05, "Average")),
fill=c("blue","orange"))
Thanks,
Harold
[[alternative HTML version deleted]]
On Wed, 2005-08-24 at 13:06 -0400, Doran, Harold wrote:> Dear List: > > I'm creating a series of barplots using Sweave that must assume a > standard format. This is student achievement data and the x-axis must > include all grades 3 to 8. In some cases, the data for a grade (or more > than one grade) are missing in the vector math.bar, but are never > missing for the vector apmxpmeet. The following sample code illustrates > the issue. > > Using the code below to create the plot works fine. However, the > following command is designed to place the data onto the plot: > > text(math.barplot, graph.max+5, format(tot), xpd = TRUE, col = c("blue", > "orange") ) > > This does work, but, it also places 'NA's on the plot. Is there a way to > place the data onto the plot such that the numbers appear, but for the > NAs not to appear? I've searched through ?text but was not able to find > a solution. > > math.bar <- c(NA,NA,NA,NA,55,48) > apmxpmeet <- c(47, 50, 49, 50, 49, 46) > > par(ps=10) > math.bar <- rbind(math.bar, apmxpmeet) > math.barplot <- barplot(math.bar, beside=T, col=c('blue','orange'), > names=c('Grade \n 3','Grade \n 4','Grade \n 5','Grade \n 6','Grade \n > 7','Grade \n 8'), > ylim=c(0,120), ylab="Percentage", xlab="Grade Level") > tot <- round(math.bar,digits=0) > graph.max <- max(math.bar, apmxpmeet, na.rm=T) > text(math.barplot, graph.max+5, format(tot), xpd = TRUE, col = c("blue", > "orange") ) > tmp$sch_nameS05 <- as.character(tmp$sch_nameS05) > legend(2,120,legend=(c(tmp$sch_nameS05, "Average")), > fill=c("blue","orange")) > > Thanks, > HaroldHarold, Get rid of the 'format(tot)' in your text() call: text(math.barplot, graph.max+5, tot, xpd = TRUE, col = c("blue", "orange")) By using format(tot), you are coercing the values to text. Thus:> format(tot)[,1] [,2] [,3] [,4] [,5] [,6] math.bar "NA" "NA" "NA" "NA" "55" "48" apmxpmeet "47" "50" "49" "50" "49" "46" versus:> tot[,1] [,2] [,3] [,4] [,5] [,6] math.bar NA NA NA NA 55 48 apmxpmeet 47 50 49 50 49 46 In the first case, text() will draw "NA", whereas in the second, the NA is ignored. Also, your code above is not entirely reproducible, since:> tmp$sch_nameS05 <- as.character(tmp$sch_nameS05)Error: Object "tmp" not found HTH, Marc Schwartz
Thanks, Marc. It works. My apologies for not making the code reproducible. -----Original Message----- From: Marc Schwartz (via MN) [mailto:mschwartz at mn.rr.com] Sent: Wednesday, August 24, 2005 1:29 PM To: Doran, Harold Cc: r-help at stat.math.ethz.ch Subject: Re: [R] Remove NAs from Barplot On Wed, 2005-08-24 at 13:06 -0400, Doran, Harold wrote:> Dear List: > > I'm creating a series of barplots using Sweave that must assume a > standard format. This is student achievement data and the x-axis must > include all grades 3 to 8. In some cases, the data for a grade (or > more than one grade) are missing in the vector math.bar, but are never> missing for the vector apmxpmeet. The following sample code > illustrates the issue. > > Using the code below to create the plot works fine. However, the > following command is designed to place the data onto the plot: > > text(math.barplot, graph.max+5, format(tot), xpd = TRUE, col = > c("blue", > "orange") ) > > This does work, but, it also places 'NA's on the plot. Is there a way > to place the data onto the plot such that the numbers appear, but for > the NAs not to appear? I've searched through ?text but was not able > to find a solution. > > math.bar <- c(NA,NA,NA,NA,55,48) > apmxpmeet <- c(47, 50, 49, 50, 49, 46) > > par(ps=10) > math.bar <- rbind(math.bar, apmxpmeet) math.barplot <- > barplot(math.bar, beside=T, col=c('blue','orange'), names=c('Grade \n > 3','Grade \n 4','Grade \n 5','Grade \n 6','Grade \n 7','Grade \n 8'), > ylim=c(0,120), ylab="Percentage", xlab="Grade Level") tot <- > round(math.bar,digits=0) graph.max <- max(math.bar, apmxpmeet, > na.rm=T) text(math.barplot, graph.max+5, format(tot), xpd = TRUE, col > = c("blue", > "orange") ) > tmp$sch_nameS05 <- as.character(tmp$sch_nameS05) > legend(2,120,legend=(c(tmp$sch_nameS05, "Average")), > fill=c("blue","orange")) > > Thanks, > HaroldHarold, Get rid of the 'format(tot)' in your text() call: text(math.barplot, graph.max+5, tot, xpd = TRUE, col = c("blue", "orange")) By using format(tot), you are coercing the values to text. Thus:> format(tot)[,1] [,2] [,3] [,4] [,5] [,6] math.bar "NA" "NA" "NA" "NA" "55" "48" apmxpmeet "47" "50" "49" "50" "49" "46" versus:> tot[,1] [,2] [,3] [,4] [,5] [,6] math.bar NA NA NA NA 55 48 apmxpmeet 47 50 49 50 49 46 In the first case, text() will draw "NA", whereas in the second, the NA is ignored. Also, your code above is not entirely reproducible, since:> tmp$sch_nameS05 <- as.character(tmp$sch_nameS05)Error: Object "tmp" not found HTH, Marc Schwartz