@wcbdata TY.

Map colors: standard #matplotlib sequential color gradient, for which "the lightness value increases monotonically": one presumes to make differences legible even with various color-blind profiles.

https://matplotlib.org/stable/tutorials/colors/colormaps.html

Explainer: took building entire library (#AnnoPlot, now pre-alpha) to allow for detailed annotations, as matplotlib doesn't support simultaneous word-wrapping and tight margins in same figure.

Cf. capacity trend viz earlier in thread.

@justyourluck Have been wanting to generate regional versions, but as of yet haven't even gotten around to the jiggery-pokery that would add outlying states and territories.

Unfortunately, #AnnoPlot charting library development is on hold as attend to other life matters. Ideally would release library for others to work with directly, but that's still a while off.

@Gdac Adding a note to my #AnnoPlot project plan to at some point generate a map of staffing loss since reporting began.

Hypothesis to be tested: population centers lost staff at a higher rate than rural facilities, simply as a function of more frequent unmitigated exposure of medical personnel where populations are denser, thus transmission sustained at higher levels.

Would be quite surprised if this weren't the case.

Still on December hiatus from programming this week—catching up on household projects—so no revisions to charts nor new features for #AnnoPlot dataviz library.

Hoping to get back to working on those in January.

This week has been full of frustrations and aggravations. Thus no closer to returning to new dataviz design.

Looking forward to making some headway on birdsite archival and pachyderm migration tools during early part of December, return to #AnnoPlot library development by Jan.

#dataviz #mastodon

Best laid plans or sumfin'.

Spoons this week went to food prep and stocking up the freezers with meals, before diving back into programming projects from twitter archival and mastodon utility tools to further development of #AnnoPlot & pandemic dataviz.

Daily postings to weekly #ThisIsOurPolio thread continues, but some half a dozen planned new #dataviz charts still remain on backburner for the forseeable future.

https://mastodon.social/@beadsland/109373482409094264

Other than minor legibility tweaks to hospital capacity charts, taking a hiatus from new chart dev this week and next.

Will be focused on #migration to and across #mastodon, including cleaning up incomplete archive received from birdsite and other suchnot.

#AnnoPlot project profile: https://mastodon.social/@beadsland/109333798459653322

#AnnoPlot #dataviz library for #matplotlib

Mostly hacks of working prototype hospital capacity & county choropleth (levels)—break out pediatrics & improve legibility.

Legibility improved also on prototype devpie—this chart. Project todo list has moved out of bullseye to top level grey slice.

#datavis

Previous update: https://twitter.com/beadsland/status/1578308035063476224

@nobodyinperson Anyway, yeah, #AnnoPlot library started to address problem of including chart captions.

#matplotlib docs are quite explicit that one can't do tight margins and word wrapping in the same plot. That didn't suit me, so started building something to allow one to do both.

Up to 5.5K lines of working prototypes and preliminary library modules. Now have the scaffolding of an artists layout loop that jiggles things around until all the elements of a #dataviz play nice together.

@nobodyinperson It can be done, but only by explicitly using a font that provides the appropriate glyphs. There's no automatic failover substitution.

Am working on a #Matplotlib library (#AnnoPlot) that is pretty intensively text oriented. Currently building text out as Packer hierarchies of artists for each word, which will eventually support emoji integration.

Still have to figure out how to properly align character baselines though, as currently sentences and paragraphs are subtly janky. 🤷

Historical tweets re development of #AnnoPlot dataviz library (the #ThisIsOurPolio toolchain) for #matplotlib:

https://twitter.com/search?q=from%3A%40beadsland%20matplotlib&src=typed_query&f=live

#introduction

Generativist—#anarchism-adjacent #criptheory—#lumpentheory of #RentierSociety & #tensegrity fetters of #classposture #kyriarchy, with interest turning increasingly to societal #grief and collective acts of #mourning.

#cripplepunk meets #solarpunk

#abolitionist by way of #STS, #NVC, Pirsig.

#neologism by way of #neurodivergence.

Team #emdash.

Old school #programmer from before dotcom bubble, now developing #AnnoPlot #dataviz library & weekly refresh of #ThisIsOurPolio charts.