Export a pie() chart into .png file

rstudio

#1

Hello,
This is how I usually export a chart/plot from ggplot() into a .png file. I am not sure this is a good way but it works pretty well for me so far

> plot_pie = pie(slices,labels = lbls, col = c("red","green","blue"),
+     main = "Sales Dollars Percentage Distribution", cex=0.8)
> plot_pie
NULL
> ggsave('Sales Percentage Comparison.png', plot_pie, width = 15, height = 10, dpi = 300)

Of course, this does not export anything because:

> plot_pie
NULL

I don't need to post any data because this is mainly a coding issue.

How can I tackle this?
Thanks!


#2

I am not sure but I think ggsave can be used only for a ggplot graph and pie is not - it is a function for base R.

For regular plot, you can use grDevices::png() for png graphic devices

png('Sales Percentage Comparison.png', width = 15, height = 10, units = "in")
pie(slices,labels = lbls, col = c("red","green","blue"), main = "Sales Dollars Percentage Distribution", cex=0.8)
dev.off()

I thing this should work; You may have to tweak it though.

EDIT:
I just checked. pie is not like a ggplot function, it does not return an object. So you can use ggsave but by directly providing the plot call.

temp_file <- tempfile(fileext = ".png")
ggplot2::ggsave(
  filename = temp_file,
  plot = pie(rep(1, 24), col = rainbow(24), radius = 0.9),
  width = 15,
  height = 10,
  dpi = 300
)
img <- magick::image_read(temp_file)
magick::image_scale(img, 'x300')
unlink(temp_file)

Created on 2018-06-30 by the reprex package (v0.2.0).


#3
> pie(slices,labels = lbls, col = c("red","green","blue"), main = "Sales Dollars Percentage Distribution", cex=0.8)
> png('Sales Percentage Comparison.png', width = 15, height = 10, units = "in")
Error in .geometry(width, height, units, res) : 
  'res' must be specified unless 'units = "px"'

#4

Solution:

 ggsave('Sales Percentage Comparison.png', pie(slices,labels = lbls, col = c("red","green","blue"), main = "Sales Dollars Percentage Distribution", cex=0.8)
+ , width = 15, height = 10, dpi = 300)

#5

Yes. As in the example I provided, it works that way with ggsave but required ggplot2.

For the other solution, just change the units for width and length or provide a res argument as explained in the error message.