Box and Whisker plots ?

[SharonEv]

Hi, I have been thinking how best to present the data from here in papers, I really like the graph formats but I am wonder if they would take up too much room. So instead I was thinking that a box and whisker plot from this data may be better. I pretty sure I read for SIAR that it was easy to do? How could I do this?

Thanks Sharon

[EmiliaIOPAN]

Hi, I have also the same issue. The posterior density graphs produced in MixSIAR are really nice, but when I have to present results for few sources and few species, it just take too much space. Can we somehow produce the "proportionbygroup" box plots, with boxes representing bayesian CI? This is a function in SIAR, but can we use it on output data from MixSIAR? "Summary statistic" basicly consist all information that we need for bayesian CI box plots, but how to combine it? Thanks, Emilia

[brianstock]

Good suggestion! Yes, it wouldn't be too hard for me to add a "proportionbygroup" function to MixSIAR - I'll get on it.

For the time being, you can try to modify this code I wrote for another analysis that produces:

[SharonEv]

Great thanks!

[ChrisHarrod]

@brianstock Did the boxplot function ever get included?

[AndrewLJackson]

Like the siar style hdr boxplots of overlapping greys etc? no, I don’t think so. Ideally these need to be rewritten for ggplot. Ive added it to my list for SIBER but its fairly far down the list.

--

Dr Andrew Jackson, PhD, FTCD Associate Professor Irish Research Council Laureate School of Natural Sciences, Department of Zoology Trinity College Dublin, the University of Dublin Dublin 2, Ireland.

+353 1 896 2728 | [email protected]mailto:[email protected] Twitter: @yodacomplexhttps://twitter.com/yodacomplex http://www.tcd.ie/Zoology/research/http://www.tcd.ie/Zoology/research/groups/jackson/groups/jackson/

From: ChrisHarrod [email protected] Reply to: brianstock/MixSIAR [email protected] Date: Thursday 30 July 2020 at 02:25 To: brianstock/MixSIAR [email protected] Cc: Subscribed [email protected] Subject: Re: [brianstock/MixSIAR] Box and Whisker plots ? (#66)

@brianstockhttps://github.com/brianstock Did the boxplot function ever get included?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/brianstock/MixSIAR/issues/66#issuecomment-666020963, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AAZLLMDYMBTCW2XAH5MVVVDR6DDXHANCNFSM4CBU3FHQ.

[brianstock-NOAA]

Sorry - no, I have only provided code examples and illustrated how to access the posteriors for plotting. My rationale is that people are always going to have to modify the plot code to make their own publication-quality figures. But I/we could probably write a function that would accept and pass along ggplot2 arguments and get pretty close. My energy has been devoted elsewhere, and only fixing bugs/issues in MixSIAR (like splitting off the GUI so MixSIAR could remain on CRAN after gWidgets was deprecated).

[ChrisHarrod]

I totally understand. I know your time is very limited. As an end user, I think it would be a really useful addition to the pipeline. I always liked the box credibility plots that @AndrewLJackson et al did in SIAR, but the Boxplot options etc in SIMMR are good too.

If I was handy at coding, I'd offer to help, but I'm not :(

Thanks to all for a really useful tool. I and lots of others very much appreciate your hard work.