Tweak aesthetics of RD-plot; fix SZ3 as normalizer#31
Conversation
| title_fontsize=14, | ||
| ) | ||
| plt.xlabel( | ||
| r"Median Compression Ratio Relative to SZ3 ($\uparrow$)", fontsize=16 |
There was a problem hiding this comment.
Should the labels be updated depending on what we normalize by?
There was a problem hiding this comment.
The normalizer is now fixed to be SZ3 for the reason outlined above, so no need to update the label!
There was a problem hiding this comment.
Is it really fixed now since above it looks like the normalizer can be configured?
There was a problem hiding this comment.
It's fixed here:
I still let the normalize function be more flexible because we might want to change the normalizer in the future.
There was a problem hiding this comment.
My issues is that the plot_aggregated_rd_curve function is public and so cannot assume that it is only called in this one place where SZ3 is fixed. It should either be private (though I'd like to keep it exposed) or also take a normalizer name parameter
| if ispdf: | ||
| # Saving a PDF with the alternative code below leads to a corrupted file. | ||
| # Hence, we use the default savefig method. | ||
| plt.savefig(outfile, dpi=300) |
There was a problem hiding this comment.
At least we should add a warning comment here saying that virtual UPath's are only support with the second method
|
@treigerm Could you show a before and after of the RD-plot? |
|
Apologies for the delay on getting back to this PR! This is before the changes: |
|
Could you open an issue about the PDF corruption, I'd like to track the conflict between supporting virtual paths in the online lab and producing PDFs so that we can resolve it later |
2 participants
This PR does three things:
wbwas leading to corrupted PDFs (I couldn't really figure out why, saving PNGs works fine)