Error in plot.new() : figure margins too large, Scatter plot

134
votes

I've looked in different questions for a solution and I've tried what was suggested but I have not found a solution to make it work.

Everytime I want to run this code it always says:

Error in plot.new() : figure margins too large

and I don't know how to fix it. Here is my code:

par(mfcol=c(5,3))
hist(RtBio, main="Histograma de Bio Pappel")
boxplot(RtBio, main="Diagrama de Caja de Bio Pappel")
stem(RtBio)
plot(RtBio, main="Gráfica de Dispersión")

hist(RtAlsea, main="Histograma de Alsea")
boxplot(Alsea, main="Diagrama de caja de Alsea")
stem(RtAlsea)
plot(RtTelev, main="Gráfica de distribución de Alsea")

hist(RtTelev, main="Histograma de Televisa")
boxplot(telev, main="Diagrama de Caja de Televisa")
stem(Telev)
plot(Telev, main="Gráfica de dispersión de Televisa")

hist(RtWalmex, main="Histograma de Walmex")
boxplot(RtWalmex, main="Diagrama de caja de Walmex")
stem(RtWalmex)
plot(RtWalmex, main="Gráfica de dispersión de Walmex")

hist(RtIca, main="Histograma de Ica")
boxplot(RtIca, main="Gráfica de caja de Ica")
stem(RtIca)
plot(RtIca, main="Gráfica de dispersión de Ica")

What can I do?

7
rplotfiguremargins
possible duplicate of Error in plot.new() : figure margins too large in R – Calimo
Margins appear to be too large for your image. This can happen if you have a small plot window. In any case, your description is insufficient to diagnose the problem. We could use a reproducible example or screenshot of your R session with the plot window. – Roman Luštrik
I my case, it helped to debug with a small subset of the data that was to be plotted like plot(df[1,1:3], df2[1,1:3]) - and then I realized that what I actually wanted to do is to plot(unlist(df[1,1:3]), unlist(df2[1,1:3])) Also see: stackoverflow.com/a/17074060/6018688 – fabianegli

7 Answers

188
votes

Every time you are creating plots you might get this error - "Error in plot.new() : figure margins too large". To avoid such errors you can first check par("mar") output. You should be getting: 

[1] 5.1 4.1 4.1 2.1

To change that write:

par(mar=c(1,1,1,1))

This should rectify the error. Or else you can change the values accordingly.

Hope this works for you.

131
votes

This can happen when your plot panel in RStudio is too small for the margins of the plot you are trying to create. Try making expanding it and then run your code again.

RStudio UI causes an error when the plot panel is too small to display the chart: RStudio with the plot panel too small

Simply expanding the plot panel fixes the bug and displays the chart: RStudio with the plot panel expanded

33
votes

Invoking dev.off() to make RStudio open up a new graphics device with default settings worked for me. HTH.

20
votes

If you get this message in RStudio, clicking the 'broomstick' figure "Clear All Plots" in Plots tab and try plot() again.

Moreover Execute the command

graphics.off()
11
votes

Just clear the plots and try executing the code again...It worked for me

7
votes

Just a side-note. Sometimes this "margin" error occurs because you want to save a high-resolution figure (eg. dpi = 300 or res = 300) in R.
In this case, what you need to do is to specify the width and height. (Btw, ggsave() doesn't require this.)

This causes the margin error:

# eg. for tiff()
par(mar=c(1,1,1,1))
tiff(filename =  "qq.tiff",
     res = 300,                                                 # the margin error.
     compression = c( "lzw") )
# qq plot for genome wide association study (just an example)
qqman::qq(df$rawp, main = "Q-Q plot of GWAS p-values", cex = .3)
dev.off()

This will fix the margin error:

# eg. for tiff()
par(mar=c(1,1,1,1))
tiff(filename =  "qq.tiff",
     res = 300,                                                 # the margin error.
     width = 5, height = 4, units = 'in',                       # fixed
     compression = c( "lzw") )
# qq plot for genome wide association study (just an example)
qqman::qq(df$rawp, main = "Q-Q plot of GWAS p-values", cex = .3)
dev.off()
-1
votes

Just run graphics.off() before plotting your data. This instruction solved my error. So, it's harmless to try it before taking a more complex solution.