Highlight multiple intersections about complex-upset HOT 14 CLOSED

Thank you. I got it, needs a tiny change to fix it - will be there in 20 minutes.

from complex-upset.

The potential fix has been just published as v0.5.11 - you can update the package using:

devtools::install_github("krassowski/complex-upset")

Please let me know if the fix works for you. And again, thanks for highlighting this!

from complex-upset.

I am closing this issue now as all the tests passed, but your feedback is still very welcome.

from complex-upset.

Thanks, it works as long as the same aesthetics are passed to each intersection query, i.e.

queries=list(
    upset_query(
      intersect=c('Drama', 'Comedy'),
      color='red',
      fill = 'red',
      only_components=c('intersections_matrix')
    )
    , upset_query(
      intersect=c('Romance', 'Drama'),
      color='yellow',
      fill='yellow',
      only_components=c('intersections_matrix')
    )
  )

works, but omitting either fill='yellow' or color='yellow' from the second query gives the error Error: Unknown colour name: NULL.

I'm quite happy with specifiying the same set of aes's for ech query.

from complex-upset.

Thank you for testing the update. I re-created the reported issue and fixed it in v0.5.12. Huge thanks again, and please do let me know of any other issues you may find.

from complex-upset.

Hi I'm encountering what I think is a related issue. At least, it has to do with specifying upset_queries and getting the "Aesthetics must be length 1 or ..." error from ggplot. I'm trying to specify multiple upset queries with different colors, but there's a point where I added another upset plot with a different color, I start getting that error.

I'm attaching two scripts and the data file to use with the script. The only difference that I can see between the scripts is that I'm specifying a different color and fill for the first upset_query.

I've tried installing the latest version from github and still got the same error.

Archive.zip

Thank you for your help.

from complex-upset.

Hi @dence this is really a lot of code and data in the archive you provided. Could you create a minimal reproducible example please? This would help me to helpy you quickly.

from complex-upset.

Hi, Here is a minimal script that replicates the error. When I change the color in the last upset_query (the one with "plant1") from "red" to "blue", I get the aesthetics error.

binary_DEGs.upsetplot.doesnt_work.R.zip

from complex-upset.

Actually, after playing around with this some more, I think this might be a case of just a very unhelpful error message. I think the error occurs when any of the specified intersections are empty.

from complex-upset.

Thank you for getting back to me. Sorry I did not have time earlier. Unhelpful messages are bad! I fully agree that this is something that should be fixed. Can you reproduce this with the movies example from the readme?

from complex-upset.

Yes, I'll try to get to that this week. Do you want just the R script or would there be some sort of printout or log that would be helpful for you?

from complex-upset.

Just the code that triggers it would be perfect.

from complex-upset.

Hi, I'm attaching two scripts called annoying_errors.works.R and annoying_errors.doesnt_work.R. The I try to plot theplot in the Rstudio console, I get this error.

Error: Aesthetics must be either length 1 or the same as the data (2): colour and fill
Run rlang::last_error() to see where the error occur

Archive.zip
red.

from complex-upset.

I'm running into this issue as well. It's popping up in a situation where I want to create the same upset plot for multiple subsets of my data. In some subsets, certain intersections become empty. Is there a way around this issue, other than manually adjusting the queries for each subset?

from complex-upset.