subset not sticking, all data shows up in plot

[Henry Drexler]

Hello,

When I plot my data, I am getting all the data points for the data set as text labels and all are showing up in the legend, but only those from my subset should be showing up.
Perhaps I am missing something simple.

I have attached the dataset.

Here is my R script:

library(ggplot2)
library(maps)
double <- read.csv(file="z_plain.csv",header=TRUE)
double$month <- as.Date(as.character(double$month), "%m/%d/%Y")
double$rolling_rate <- double$rolling_rate*100
blast_ps <- subset(double,double$type == "typeA")
p <- qplot(month,rolling_rate,data=blast_ps,shape=15,color=report,ylab="rolling_rate (percent)") + geom_line() + geom_text(aes(label=double$rolling_rate),position = "jitter")
p


As you can see I have blast_ps <- subset(double,double$type == "typeA") so only a,b,c should be shown in the key and only its text data values should be shown in the plot - although it is showing all the data values.

dput(double)
structure(list(report = structure(c(1L, 1L, 1L, 1L, 2L, 2L, 2L, 
2L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 5L, 5L, 5L, 5L, 6L, 6L, 6L, 
6L), .Label = c("a", "b", "c", "d", "e", "f"), class = "factor"), 
    type = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
    1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L), .Label = c("typeA", 
    "typeB"), class = "factor"), month = structure(c(15248, 15279, 
    15309, 15340, 15248, 15279, 15309, 15340, 15248, 15279, 15309, 
    15340, 15248, 15279, 15309, 15340, 15248, 15279, 15309, 15340, 
    15248, 15279, 15309, 15340), class = "Date"), rolling_rate = c(10.4, 
    10.1, 9.8, 9.6, 29, 26.8, 26.3, 25.3, 39.8, 38.1, 37.6, 36, 
    7.9, 7.2, 6.6, 6.6, 7.7, 7.2, 7.1, 7.3, 12.1, 11.3, 11.2, 
    11.3)), .Names = c("report", "type", "month", "rolling_rate"
), row.names = c(NA, -24L), class = "data.frame")


dput(blast_ps)
structure(list(report = structure(c(1L, 1L, 1L, 1L, 2L, 2L, 2L, 
2L, 3L, 3L, 3L, 3L), .Label = c("a", "b", "c", "d", "e", "f"), class = "factor"), 
    type = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
    1L, 1L), .Label = c("typeA", "typeB"), class = "factor"), 
    month = structure(c(15248, 15279, 15309, 15340, 15248, 15279, 
    15309, 15340, 15248, 15279, 15309, 15340), class = "Date"), 
    rolling_rate = c(10.4, 10.1, 9.8, 9.6, 29, 26.8, 26.3, 25.3, 
    39.8, 38.1, 37.6, 36)), .Names = c("report", "type", "month", 
"rolling_rate"), row.names = c(NA, 12L), class = "data.frame")

Thank you in advance for any insights.

[Justin Haynes]

look at levels(double$type) and levels blast_ps$type. Because you are
subsetting a factor, it carries all the levels of the factor along.

blast_ps$type <- droplevels(blast_ps$type)
blast_ps$report <- droplevels(blast_ps$report)

you can also use ?str to examine structures which can be incredibly useful.

From there you should be able to plot like you wanted.

Hope that helps,

Justin

2012/3/7 Henry Drexler <alon...@gmail.com>:

> --
> You received this message because you are subscribed to the ggplot2 mailing
> list.
> Please provide a reproducible example: http://gist.github.com/270442
>
> To post: email ggp...@googlegroups.com
> To unsubscribe: email ggplot2+u...@googlegroups.com
> More options: http://groups.google.com/group/ggplot2

[Art]

Hello Henry,

Using the ggplot structure is sometimes easier when adding components to the graph.

I've modified your geom_text to reference the subset data.frame and I've added colour grouping to the geom_line argument to graph the a, b, and c reports.

See script below.

Hope that helps.

Best,

Art 

p <- ggplot(blast_ps, aes(month, rolling_rate)) +

  geom_line(aes(colour = report)) +

  geom_text(aes(label=blast_ps$rolling_rate),position = "jitter")

p

[Henry Drexler]

Justin, thank you for the hints, I will read up on those.