Closed sjspielman closed 1 year ago
@jharenza I don't see any color loss on my end when opened on my computer in preview, but this can be heavily affected by the viewer you use to look at the files. Maybe the github view (if that's what you're looking at?) is just a little dulled?
So bizarre! It is showing in my preview and adobe for the PDF, but not the panels:
Interesting... When I re-export after making these changes from review shortly, I'll see if there's another setting in there!
compare to top right which is in master:
Mainly seeing in barplots. The others are dull bc of the alpha, and I see that in almost all pdfs. Just that the bars jumped out at me!
CMYK vs. RGB is the first thing I'd check.
@jharenza how do these look to you now? I did two things to try and fix colors -
@jharenza how do these look to you now? I did two things to try and fix colors -
Looking the same to me 😬
@jharenza how do these look to you now? I did two things to try and fix colors -
Looking the same to me 😬
Also looking dull to me...
En route to re-importing PDFs into AI in the first place with the correct color settings already turned on, stay tuned!
See https://github.com/AlexsLemonade/OpenPBTA-analysis/pull/1750/commits/760fca5464fde22b7d9441ef51846feb1479330f! 🎨 🎉
(edit - just saw that ATRT text in D needs to be fixed, doing now, but the colors are better!) (edit 2 - omg and the A legend! i was just excited about the colors apparently...!)
(edit 2 - omg and the A legend! i was just excited about the colors apparently...!)
I was about to comment... The Illustrator!
yay colors!
I guess I didn't realize the colors in some other figures were dull until the barplots came up dull. (They have been dull the whole time, I think, eg figure5 in PR now)...
They have been dull the whole time, I think, eg figure5 in PR now
I'm going to go back and check it all!!!
I think this one is about ready now!
@jashapiro leaving this open in case you wanted another look, lmk!
what happened to my lovely gradient on the co-occurrance score legend?
This is a frustrating indeed! I've played around with some export settings, but nothing is really working. It also seems that when I import the PDF figure into Illustrator in the first place, the gradient is already choppy from the beginning. I'm going to try re-exporting the plot with useDingbats = FALSE
which maaayyy help..
Edit - did not help.. Maybe @envest has a thought or two about preserving smooth gradients in Illustrator?
Maybe @envest has a thought or two about preserving smooth gradients in Illustrator?
Sorry nothing specific... this was my most promising find on StackExchange 🤷
So it's fine in the original PDF but not after integration into Illustrator? huh
Yep, for your viewing pleasure.. edit wow this shows up rather largely....but you get the point!
https://user-images.githubusercontent.com/4701111/234904391-b7835873-6f84-4e74-bc46-e1c916f8f68a.mov
In conclusion, we have decided it is what it is!
Closes #1738
This PR updates Figure 3 with larger text all around. I did some manual changes in illustrator as well as code changes. Code changes include:
mutational-signatures
module in07-plot_cns_fit.Rmd
\n
in one of the columns. It turns outreadr::read_csv()
has no problem figuring out how to read those files in, which is why no "bug" was previously caught! But, we shouldn't be exporting CSV's with inline\n
so I fixed it and re-rendered associated notebooks. The offending CSV isanalyses/mutational-signatures/results/deconstructsigs_exposures_merged.tsv
chromothripsis
module in04-plot-chromothripsis-and-breakpoint-data.Rmd
figures/scripts/fig3-chromothripsis-barplot.R
Let me know what you think of this first go!