looping figure creation, but needing to input different species name each iteration
58 views
Skip to first unread message
henn...@d.umn.edu
Dec 17, 2015, 5:30:09 PM12/17/15
to ggplot2
I posted yesterday and received a couple of responses (Thanks to those who did), but my needs have evolved slightly.
I am creating figures for 75 species at nine different park units and may have need to do this several times, so I'm really in need of a method to loop this process to create consistent figures quickly.
The looping itself was quite easy to figure out, however I'm having difficulties creating a title that satisfies three criteria:
a. I need to italicize the species name which is contained within the figure title (i.e.Mean Suitability in Anonymous National Park for Canus lupus by Year)
b. It should be able to wrap to at least two rows so the title is not wider than the figure itself.
c. Since I am looping this, I need to draw the species name to be italicized from the dataset that I am creating the figure from.
the below code is what I've figured out so far. The part that I cannot for the life of me figure out is how to reference the variable i (which represents the genus species name) to be italicized. Any help anyone could provide would be greatly, greatly appreciated. Thanks.
for( i in unique(park$species)) {
myplot<- subset(park, species==i)
my_title <- paste("Mean Suitability in Anonymous National\nPark for", i, "by Year")
plotprint=ggplot(myplot, aes(x=year, y=MEAN, group=rcp))+geom_line(size=1,aes(colour= factor(rcp)))+geom_point(aes(colour=factor(rcp)))+labs(title = my_title, x="Projection Year", y="Mean Suitability\nValue")+scale_x_continuous(breaks=c(2015,2025,2050,2075,2095))+scale_y_continuous(limits = c(0,1),breaks=c(0.2,0.4,0.6,0.8,1.0))+theme(panel.grid.major = element_blank(),panel.grid.minor=element_blank(),panel.background = element_rect(fill = "white"), plot.title = element_text(vjust=1.5, size=14),axis.title.y=element_text(vjust=1),panel.border = element_rect(colour = "black", fill=NA, size=1),axis.title.x=element_text(vjust=-0.1),axis.text.x=element_text(vjust=0.2, colour="black"), axis.text.y=element_text(colour="black"))+scale_colour_manual(name="factor", breaks=c("group1","group2"), labels=c("Group 1","Group 2"),values=c("group1"="blue","group2"="red"))
ggsave(paste0(i,".png"), width=6, height=2.5,plotprint)
}
species rcp year park MEAN
species_genus group1 2010 ANON 0.544363995
species_genus group2 2010 ANON 0.474398504
species_genus group1 2015 ANON 0.497079159
species_genus group2 2015 ANON 0.571342493
species_genus group1 2020 ANON 0.539996331
species_genus group2 2020 ANON 0.522597844
species_genus group1 2025 ANON 0.464497998
species_genus group2 2025 ANON 0.479889333
species_genus group1 2030 ANON 0.539434493
species_genus group2 2030 ANON 0.516314661
species_genus group1 2035 ANON 0.445646996
species_genus group2 2035 ANON 0.466872344
species_genus group1 2040 ANON 0.461644833
species_genus group2 2040 ANON 0.382267167
species_genus group1 2045 ANON 0.371176163
species_genus group2 2045 ANON 0.299473159
species_genus group1 2050 ANON 0.316472329
species_genus group2 2050 ANON 0.143133335
species_genus group1 2055 ANON 0.354656503
species_genus group2 2055 ANON 0.161598501
species_genus group1 2060 ANON 0.167542666
species_genus group2 2060 ANON 0.143555
species_genus group1 2065 ANON 0.138111168
species_genus group2 2065 ANON 0.127673333
species_genus group1 2070 ANON 0.121445348
species_genus group2 2070 ANON 0.126145499
species_genus group1 2075 ANON 0.342944995
species_genus group2 2075 ANON 0.127160334
species_genus group1 2080 ANON 0.163886666
species_genus group2 2080 ANON 0.136789334
species_genus group1 2085 ANON 0.157619332
species_genus group2 2085 ANON 0.116568
species_genus group1 2090 ANON 0.157446332
species_genus group2 2090 ANON 0.131710666
species_genus group1 2095 ANON 0.155704334
species_genus group2 2095 ANON 0.106103182
Tom Philippi
Dec 17, 2015, 6:26:07 PM12/17/15
to henn...@d.umn.edu, ggplot2
One approach would be to use plotmath, something like:
my_title <- expression(paste("Mean Suitability in Anonymous National\nPark for",italics(species)i, "by Year"))
Plotmath lets you include many mathematical symbols as well as italicize, underline, superscript & subscript in title & label strings.
?expression and ?plotmath and perhaps googling "plotmath ggplot2" should get you what you need.
Tom Philippi
Quantitative Ecologist & Data Therapist
Inventory and Monitoring Program
National Park Service
--
--
You received this message because you are subscribed to the ggplot2 mailing list.
Please provide a reproducible example: https://github.com/hadley/devtools/wiki/Reproducibility
To post: email ggp...@googlegroups.com
To unsubscribe: email ggplot2+u...@googlegroups.com
More options: http://groups.google.com/group/ggplot2
---
You received this message because you are subscribed to the Google Groups "ggplot2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ggplot2+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Dennis Murphy
Dec 17, 2015, 10:20:42 PM12/17/15
to henn...@d.umn.edu, ggplot2
Hi:
1. You need to simplify your plot call - the calls to theme() are
unnecessary for the purposes of your question.
2. Your code got mashed together likely because you sent your mail in
HTML; therefore, it wasn't easy to read.
Complaints aside, the biggest programming problem was to combine
normal and italic fonts in the title across two lines and avoid
getting a large space between the end of the first text chunk and the
italicized text. From your end, you have to distinguish the contents
of an atomic vector from their respective indices/locations. Below is
one way to do what I think you requested. I made up a fake data set to
generate multiple species; you could extend the idea to species-park
combinations if that is of interest.
# Fake data set - year is numeric rather than factor, so a group
# aesthetic for it is unnecessary.
DF <- data.frame(
species = factor(rep(paste0("species_", LETTERS[1:3]), each = 10)),
rcp = c("group1", "group2"),
year = rep(2011:2015, each = 2),
MEAN = runif(30, 0.3, 0.6))
library(ggplot2)
# define an atomic vector of species names
species_names <- as.character(unique(DF$species))
# A function to create a plot for an individual species. It takes a
# data frame DF and a character object species_name as formal arguments
plotfun <- function(DF, species_name)
{
d <- DF[DF$species == species_name, ] # get the data subset
# Manufacture an expression that creates a two-line title
my_title <- bquote(atop(paste("Mean Suitability for ",
italic(.(species_name)), " by Year"),
"Anonymous National Park"))
# create and return the ggplot
ggplot(d, aes(x = year, y = MEAN, colour = rcp)) +
geom_point() +
geom_line() +
labs(title = my_title, x = "Projection Year",
y="Mean Suitability\nValue", colour = "factor")
}
# Apply the above function to each species, return a list of ggplot objects
L <- lapply(species_names,
function(x) plotfun(DF, x))
# Cycle through the list of plots and save them to individual files
for(i in seq_along(species_names))
{
filename <- paste(species_names[i], "png", sep = ".")
ggsave(filename, L[[i]], width = 6, height = 2.5)
}
Creating axis titles and main titles in a plot can sometimes be a
challenge, particularly when one needs to insert non-standard text
such as mathematics or text in a modified font. The challenge here was
to create a main title with two lines that mixes normal and italic
text. \n works when given a text string without 'embellishments', but
unexpected things can happen if your input is an expression rather
than a string. I made two adjustments: (i) I used atop() because it
can take an expression or a string as an argument; (ii) I modified the
title so that the italicized text was on the first line inside an
expression().
The code above cycles through each species name, first taking the
appropriate data subset and then producing and returning a plot with a
two-line main title. The lapply() call produces a list of plots, one
per species. The loop at the end cycles through the list, saving each
plot to a file in png format. I find it useful to save the plots in a
list in case you need to reference them later. If you don't want the
list of plots, assign the ggplot() output to an object, say p, inside
plotfun() and make its last line
ggsave(p, file = paste(species_name, "png", sep = ".", width = 6, height = 2.5))
HTH,
Dennis
1. You need to simplify your plot call - the calls to theme() are
unnecessary for the purposes of your question.
2. Your code got mashed together likely because you sent your mail in
HTML; therefore, it wasn't easy to read.
Complaints aside, the biggest programming problem was to combine
normal and italic fonts in the title across two lines and avoid
getting a large space between the end of the first text chunk and the
italicized text. From your end, you have to distinguish the contents
of an atomic vector from their respective indices/locations. Below is
one way to do what I think you requested. I made up a fake data set to
generate multiple species; you could extend the idea to species-park
combinations if that is of interest.
# Fake data set - year is numeric rather than factor, so a group
# aesthetic for it is unnecessary.
DF <- data.frame(
species = factor(rep(paste0("species_", LETTERS[1:3]), each = 10)),
rcp = c("group1", "group2"),
year = rep(2011:2015, each = 2),
MEAN = runif(30, 0.3, 0.6))
library(ggplot2)
# define an atomic vector of species names
species_names <- as.character(unique(DF$species))
# A function to create a plot for an individual species. It takes a
# data frame DF and a character object species_name as formal arguments
plotfun <- function(DF, species_name)
{
d <- DF[DF$species == species_name, ] # get the data subset
# Manufacture an expression that creates a two-line title
my_title <- bquote(atop(paste("Mean Suitability for ",
italic(.(species_name)), " by Year"),
"Anonymous National Park"))
# create and return the ggplot
ggplot(d, aes(x = year, y = MEAN, colour = rcp)) +
geom_point() +
geom_line() +
labs(title = my_title, x = "Projection Year",
y="Mean Suitability\nValue", colour = "factor")
}
# Apply the above function to each species, return a list of ggplot objects
L <- lapply(species_names,
function(x) plotfun(DF, x))
# Cycle through the list of plots and save them to individual files
for(i in seq_along(species_names))
{
filename <- paste(species_names[i], "png", sep = ".")
ggsave(filename, L[[i]], width = 6, height = 2.5)
}
Creating axis titles and main titles in a plot can sometimes be a
challenge, particularly when one needs to insert non-standard text
such as mathematics or text in a modified font. The challenge here was
to create a main title with two lines that mixes normal and italic
text. \n works when given a text string without 'embellishments', but
unexpected things can happen if your input is an expression rather
than a string. I made two adjustments: (i) I used atop() because it
can take an expression or a string as an argument; (ii) I modified the
title so that the italicized text was on the first line inside an
expression().
The code above cycles through each species name, first taking the
appropriate data subset and then producing and returning a plot with a
two-line main title. The lapply() call produces a list of plots, one
per species. The loop at the end cycles through the list, saving each
plot to a file in png format. I find it useful to save the plots in a
list in case you need to reference them later. If you don't want the
list of plots, assign the ggplot() output to an object, say p, inside
plotfun() and make its last line
ggsave(p, file = paste(species_name, "png", sep = ".", width = 6, height = 2.5))
HTH,
Dennis
Roman Luštrik
Dec 18, 2015, 3:25:46 AM12/18/15
to Dennis Murphy, henn...@d.umn.edu, ggplot2
Another good option for concatenating text is `sprintf` function. Wrap that into `expression` and you're all set.
Cheers,
Roman
--
In God we trust, all others bring data.
henn...@d.umn.edu
Dec 21, 2015, 12:27:24 PM12/21/15
to ggplot2, henn...@d.umn.edu
Hi Dennis,
I'll be aware of your suggestions next time. Thanks.
Your suggestions worked quite well. The key to modifying what I had done to get it all to work was the addition of nesting atop inside of bquote.
myplot<- subset(anon, species==i)
my_title<-bquote(atop(paste("Mean Suitability for ", italic(.(i)), " by Year"), "Anonymous National Park"))
thanks so much for you help. This will be a huge time saver.
Kris
Reply all
Reply to author
Forward
0 new messages