How to plot data containing NA's as "gaps" in geom_line

[Thiago V. dos Santos]

Hi all,

I'm having trouble plotting data containing NA's with ggplot.

I have several variables on my data.frame df1. Basically, I would like to create time-series (geom_line) plots of two of them: "latent" and "LE.W.M2.".

Latent line should be black, solid, and LE.W.M2. should be black and dashed. Also, I need a legend identifying each line.

However, LE.W.M2. contains some NA values. This is not a problem, as long as the NA's are shown as gaps in the lines. I would expect ggplot2 do this automatically, but it simply refuses to plot this line. Please run the code:

---------------------------------------------------------------------------------------------------------------------------------------------------------------

## Load package to

library(reshape2) # melt data.frames

library(ggplot2)  # create plots

library(scales)   # improve dates on graph axis

library(plyr)     # subset data while plotting

library(RCurl)    # download data

# Read data and define NA

link <- getURL('https://dl.dropboxusercontent.com/u/27700634/Bananal_validation_2011-2013.csv')

file <- read.csv(textConnection(link), header=T, sep=';', stringsAsFactors = FALSE, as.is=TRUE)

file[file==-9999] <- NA

# Get date from character

file$date <- as.Date(as.character(file$date), "%m/%d/%Y")

# Open model data

link2 <- getURL('https://dl.dropboxusercontent.com/u/27700634/zday_energybal.txt')

mod <- read.csv(textConnection(link2), sep="", header=T)

tmp <- paste(mod$year,"-", mod$month, "-", mod$day, sep='')

mod$date <- as.Date(tmp)

# Reorder DF and discard old date characteres

mod <- mod[c(9,4,5,6,7,8)]

# Merge datasets by date

data <- merge (file, mod, by='date')

names(data)

# Then melt by date

df1 <- melt (data, id='date')

#################################################################

# Create Plot

#################################################################

p <- ggplot(df1, aes(x=date, y=value, colour=variable)) + 

       scale_x_date(breaks = '1 month', labels = date_format('%b-%y'),

                                limits = as.Date(c('2012-01-01','2013-01-01'))) +

       theme_bw(16) + 

       theme(panel.margin = unit(0,'lines'), axis.title.x = element_blank()) 

# Here I create the two geom_lines, subsetting desired data "one the fly"

comparison <- p + geom_line(subset = .(variable == 'latent'),   linetype='solid', colour='black') +

                                 geom_line(subset = .(variable == 'LE.W.M2.'), linetype='dashed',colour='black') +

                                 ylab (expression(paste("LE (","W.",m^-2,")", sep="")))

print(comparison) # But only 'latent' line is shown, perhaps because LW.W.M2. contains some NA's

---------------------------------------------------------------------------------------------------------------------------------------------------------------

How can I plot LE.W.M2. as a geom_line with gaps corresponding to NA values?

How can I add a legend identifying each line?

Thanks in advance,

--

Thiago V. dos Santos
PhD student
Land and Atmospheric Science
University of Minnesota
http://www.laas.umn.edu/CurrentStudents/MeettheStudents/ThiagodosSantos/index.htm
Phone: (612) 323 9898

[Janesh Devkota]

The problem here is with the naming of your variable. Because you have too many “.” In your variable names. I removed “.” From your variable names and it seems to be working now.

 

# Create a new variable1 by removing dots

df1$variable1 <- as.factor(gsub(pattern="\\.",replacement="",x=df1$variable))

 

# Now plot your data you will see two lines

p <- ggplot(subset(df1,variable1==c("NETWM2","latent")),na.rm=T) + geom_line(aes(date,value,linetype=variable1))

p

 

Best,

Janesh  Devkota

--
--
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/groups/opt_out.

[thi_v...@yahoo.com.br]

Hi Janesh, Thanks a lot, your tip did the trick. Is there any way to insert a legend and control the linetype of each line? I have been trying it with scale_color_manual and it didn't work. Best, Thiago.

---

From: Janesh Devkota <janesh....@gmail.com>;
To: 'Thiago V. dos Santos' <thi_v...@yahoo.com.br>; <ggp...@googlegroups.com>;
Subject: RE: How to plot data containing NA's as "gaps" in geom_line
Sent: Wed, Feb 19, 2014 2:14:35 AM

[Janesh Devkota]

Hello Thiago, 

You can use scale_linetype_manual to control the linetype of each line. 

Best,

Janesh Devkota

[thi_v...@yahoo.com.br]

Janesh, Thanks again, I made it. Best, Thiago.

---

From: Janesh Devkota <janesh....@gmail.com>;

To: thi_v...@yahoo.com.br <thi_v...@yahoo.com.br>;
Cc: ggp...@googlegroups.com <ggp...@googlegroups.com>;
Subject: Re: RE: How to plot data containing NA's as "gaps" in geom_line
Sent: Thu, Feb 20, 2014 5:54:09 PM