distorted aspect ratio in figures with very different x/y axes acaling about rphylopic HOT 11 CLOSED

thanks for the report @KamilSJaron and for including session info!

@dill any ideas? This seems clearly linked to the change made in #26 #28 From this report, seems like perhaps there was a good use case for the old version of add_phylopic_base() that used par() - I wonder if we can sort out the high level use cases here and ideally support all of reasonable ones. Is there just two use cases that we can use a parameter to toggle between?

from rphylopic.

Hi folks, sorry for the delay, I was doing a fieldwork (collecting the right silhouettes). I will take a look shortly.

from rphylopic.

Thanks for checking! I liked the previous specification, but since it was causing issues for people in terms of output we had to switch (rather have functionality in all cases over ease of use) but if you think of a way of getting it working (even just as a quick example, not a full PR) feel free to open an issue and I'll look into it :)

@sckott can the PR be merged now?

from rphylopic.

yes

from rphylopic.

ah bummer, I thought we'd got this sorted. I'll take a look and post back when I have a chance. Sorry about the hassle @KamilSJaron!

from rphylopic.

thanks @dill

from rphylopic.

No worries, thanks to you for making am r interface for phylopic!

from rphylopic.

Sorry for the delay, a bunch of other work stuff came up. I've tried to fix this in this PR.

I've also added an xsize argument to allow for any fine-grained adjustment in case the automatic version doesn't work (again!).

Let me know if this works for you :)

from rphylopic.

@KamilSJaron if you have time can you test the solution in pull request #30 - should be able to install like remotes::install_github("dill/rphylopic")

from rphylopic.

Hi there, I tried and it does work. With xsize it will be always possible to plot a silhouette in any possible size including all distortions.

To be completely honest, I liked the previous behaviour better than this one (specifying always relative coordinates). I will think if there would be an easy way to allow both absolute and relative coordinates / sizes.

from rphylopic.

🍻

from rphylopic.