[R-meta] xlim - half text / many studies: row text problemes
P. Roberto Bakker
robertob@kker @ending from gm@il@com
Fri Dec 14 14:35:15 CET 2018
Hi Anne-Wil,
I just wanted to tell that your information below was were usable.
Thank you,
Roberto
Op ma 11 jun. 2018 om 15:17 schreef P. Roberto Bakker <
robertobakker using gmail.com>:
> Hi Anne-Wil,
>
> Thank you so much for your information.
> I will let you know if I have more questions.
>
> All the best (en hartelijke groet ook),
>
> Roberto
>
> 2018-06-11 14:36 GMT+02:00 Anne-Wil Kruijt <mail using awkruijt.nl>:
>
>> Hi Robert and list,
>>
>> The problem is not so much with the internal code of the plot (such that
>> changing the size of, for instance, the axes will solve it), but with the
>> space that is made available for the plot - it is just not high enough for
>> your number of records.
>> I wrote some script with examples, comments, and the solution that I use
>> (printing the plot as a .png with specified dimensions), which I paste
>> below. I hope this solves it for you. Please let me know if it doesn't!
>>
>> All best (hartelijke groet!),
>>
>> Anne-Wil
>>
>>
>> ### start of example script
>>
>> # this is code given as an example of a forest plot in metafor's
>> documentation:
>>
>> ### meta-analysis of the log risk ratios using a random-effects model
>> res <- rma(measure="RR", ai=tpos, bi=tneg, ci=cpos, di=cneg, data=dat.bcg,
>> slab=paste(author, year, sep=", "))
>>
>> ### default forest plot of the log risk ratios and summary estimate
>> forest(res)
>>
>>
>> # now we double up the number of records by creating an 'example data
>> frame' (edf) which consists of
>> # the same data use above copied twice:
>>
>> edf <- rbind(dat.bcg, dat.bcg)
>> res <- rma(measure="RR", ai=tpos, bi=tneg, ci=cpos, di=cneg, data=edf,
>> slab=paste(author, year, sep=", "))
>>
>> # now the forest plot shows those overlapping rows that you have for your
>> data too:
>> forest(res)
>>
>>
>> # the problem is that the number of rows doubles, but the total
>> dimensions made available (by rmarkdown) to plot it in are still the same.
>>
>> # if you're working in a .rmd file, a quick-glance workaround is to hit
>> the left one of the three tiny light gray icons that appear at the top
>> right of the plot output (right below the chunk you're working in). It will
>> open a pop up window that you can resize.
>> # if you're working in 'plain r studio', you can hit the icon on the
>> right of three icons at the top left of the 'viewer pane' to get a
>> resizable pop up window with the plot.
>>
>> # personally, I use the above for quick glances, and use the code below
>> to create the images that I eventually use in papers, presentations etc.
>> # It allows you to set the required resolution and size (in pixels), as
>> well as the background color (in this example it's set to transparent,
>> which can be nice for powerpoints with a coloured background).
>>
>> # if you're working on a mac you may need to install xquartz if you don't
>> have it yet (https://www.xquartz.org/); if you're on windows I don't
>> know if you need additional software (but I suspect not - perhaps you won't
>> even need the dev.off() line on windows either).
>>
>> # it is important (on my computer at least) to run the three lines
>> (png(), plotcall, dev.off()) in one go (select the three, then run, or
>> place them in their own chunk and then run - but don't run them line by
>> line)
>>
>> # these two sets of three lines (and the DPI setting line) will save the
>> same plot (forest(res)) twice, as .pngs in your working directory. You
>> won't see the plots appear in R itself (! - it will only show some text
>> like "null device 1") but you can find them using finder/explorer.
>> # The first set of three lines saves it as a .png with width = 500 and
>> height = 350 (giving slighly overlapping lines), the second set saves it as
>> a .png with width = 500 and height = 600, which is a lot more pleasant.
>> Just toy around with the width and heigth settings untill the plots feels
>> good for you purposes.
>>
>> # enjoy!
>>
>> dpi = 150
>>
>>
>> png(filename = "foresplot500by350.png", width= (500 *(dpi/72)), height=
>> (350 * (dpi/72)), res = dpi, bg = "transparent") # info on the file to save
>> forest(res) # the r code for a plot - could also be ggplot, base plot, etc
>> dev.off() # close the connection with the external writer.
>>
>>
>> png(filename = "foresplot500by600.png", width= (500 *(dpi/72)), height=
>> (600 * (dpi/72)), res = dpi, bg = "transparent")
>> forest(res)
>> dev.off()
>>
>> ### end of example script
>>
>>
>>
>> On 2018-06-11, 13:37, "R-sig-meta-analysis on behalf of P. Roberto
>> Bakker" <r-sig-meta-analysis-bounces using r-project.org on behalf of
>> robertobakker using gmail.com> wrote:
>>
>> Hi,
>>
>> My R page is 3.5.0.
>> My question is about forest() in metafor.
>>
>> Thank you in advance.
>>
>> Best,
>> Roberto
>>
>> 2018-06-08 18:21 GMT+02:00 P. Roberto Bakker <robertobakker using gmail.com
>> >:
>>
>> > Hi,
>> >
>> > My apologies, I forgot telling, the package is metafor/forest
>> >
>> > Best and thank you in advance,
>> >
>> > Roberto
>> >
>> > 2018-06-07 10:11 GMT-04:00 P. Roberto Bakker <
>> robertobakker using gmail.com>:
>> >
>> >> Hi,
>> >>
>> >> When I change 'xlim(-1, 49)' into 'xlim(1, 49)' then the lower
>> half of
>> >> text "RE Model..." disappears.
>> >> How can I correct this?
>> >>
>> >> Reason for changing the 'xlim' is because this meta-analysis
>> contains a
>> >> lot of studies and therefore row texts overlap. A smaller font
>> makes it
>> >> difficult to read
>> >> By lowering the lower xlim the text font is better.
>> >> Is ther a better way?
>> >>
>> >> Best and thank you in advance,
>> >>
>> >> Roberto
>> >>
>> >
>> >
>>
>> [[alternative HTML version deleted]]
>>
>> _______________________________________________
>> R-sig-meta-analysis mailing list
>> R-sig-meta-analysis using r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
>>
>>
>>
>>
>
[[alternative HTML version deleted]]
More information about the R-sig-meta-analysis
mailing list