Properly aligning directlabels with ggplot2

Question

I am absolutely new to directlabel + ggplot2 package. So, I'd appreciate any help. I want to properly align my labels for the linegraph--I've spent hours trying to learn directlabel, and I think with SO's help I can make this work.

Here's my data:

structure(list(Rio.Olympics.Sports.Participating.Team = structure(c(2L, 
2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 4L, 
4L, 4L, 4L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 3L, 
3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L), .Label = c("American Gymnastics", 
"American Swimmers", "Boxing", "European Gymnastics", "Running"
), class = "factor"), Calendar.Quarter = structure(c(16071, 16161, 
16252, 16344, 16436, 16526, 16617, 16709, 16801, 16892, 16983, 
17075, 16071, 16161, 16252, 16344, 16436, 16526, 16617, 16709, 
16801, 16892, 16983, 17075, 16071, 16161, 16252, 16344, 16436, 
16526, 16617, 16709, 16801, 16892, 16983, 17075, 16071, 16161, 
16252, 16344, 16436, 16526, 16617, 16709, 16801, 16892, 16983, 
17075, 16071, 16161, 16252, 16344, 16436, 16526, 16617, 16709, 
16801, 16892, 16983, 17075), class = "Date"), Randomized.Viewers = c(49, 
45, 51, 55, 47, 48, 54, 57, 53, 50, 52, 58, 32, 29, 33, 40, 34, 
36, 31, 39, 37, 30, 35, 41, 5, 1, 25, 46, 38, 4, 56, 27, 21, 
43, 42, 44, 2, 59, 3, 10, 60, 7, 14, 24, 13, 16, 17, 28, 15, 
6, 19, 23, 11, 12, 20, 22, 9, 8, 18, 26)), .Names = c("Rio.Olympics.Sports.Participating.Team", 
"Calendar.Quarter", "Randomized.Viewers"), row.names = c(NA, 
-60L), class = "data.frame")

Now, I wrote the following code to plot a simple linegraph:

d<-ggplot(data = p, aes(x=Calendar.Quarter,y=Randomized.Viewers)) +
       geom_line(aes(color = Rio.Olympics.Sports.Participating.Team))  +
       scale_color_manual(name = "Sports Type",values = c("blue", "gray51","gray51","gray51","gray51"),guide = FALSE)
 direct.label(d, list('last.qp', cex=.75,dl.trans(x=x-3)))

I used the following reference guide for maxvar.qp http://directlabels.r-forge.r-project.org/docs/

The output of this graph is:

As we can see all the points have moved to the left. While the top-3 lines seems to be okay, but "Running" and "Boxing" have moved a lot.

I have two questions:

a) Is there anyway I can place labels on top of linegraphs and toward the end of the linegraph, especially for "Running" and "Boxing"? I read this thread Labelling points with ggplot2 and directlabels but I wasn't quite sure what's going on---I am a beginner with R have started using R only for about 2-3 weeks. My hypothesis is that the poster is using for loops, but I have too many rows in my actual data, and I want to stay away from for loop. I am looking for vectorized operation, if possible.

b) From this thread How to use custom names for labels with ggplot2 + directlabels, it seems that DirectLabels package takes values from Column names of the table. Is there anyway I can modify the label names? ggplot2 provides an API for this. However, I am not sure whether directlabels has such API as well. If not, I would use colnames to overwrite the column names before calling directlabel function. I'd appreciate any thoughts on this.

PS: I am asking two questions because I strongly believe they are related to the same data and problem. In the past, I have been criticized for creating multiple questions on the same dataset. So, I'm using my best judgment here.

Names are taken from factor levels which you map to color. Function levels() can modify names for you. — Roman Luštrik
@RomanLuštrik Thanks for your help. I've two questions: a) Is levels() provided by directlabel? b) Can you please help me with the first question, if possible? — watchtower

lukeA lukeA · Accepted Answer · 2016-09-04T12:43:06

a) Is there anyway I can place labels on top of linegraphs and toward the end of the linegraph [...]?

For example,

direct.label(d, list('last.qp', cex=.75, hjust = 1, vjust = 0))

gives you

b) [...] Is there anyway I can modify the label names?

For example

d + geom_dl(
  aes(label = letters[1:5][p[,1]]), 
  method = list('last.qp', cex=.75, hjust = 1, vjust = 0)
)

gives you

The mapping is

cbind(levels(p[,1]), letters[1:5])
#      [,1]                  [,2]
# [1,] "American Gymnastics" "a" 
# [2,] "American Swimmers"   "b" 
# [3,] "Boxing"              "c" 
# [4,] "European Gymnastics" "d" 
# [5,] "Running"             "e"

You can adjust it as you wish.

Properly aligning directlabels with ggplot2

1 Answers