Lots of figures in the second chapter of the interpretability report.

Interpretability report figures in progress.

Strangely compelling.

From: https://twitter.com/M_PF/status/869982898153246721

Working on figures for the interpretability report.

This version history interaction will be useful to look at and compare to the churn probability prototype.

From: https://www.figma.com/

I got the new version of the blog to successfully download a cat photo when told.

Slowly making my way through oauth for the new blog.

I like this interaction detail on Jacob Heftmann's photo page. A nice, minimal way to have some branding/info on the site.

From: http://www.jacobheftmann.com/photo

I like this effect of making it look like a smaller page was printed on a larger page (almost like a printing error). I wonder if it could be used effectively on the web to break up long text or to make text fit better with a wide layout (by making the discrepancy look intentional).

From: http://rafrennie.com/

Zeroing on the analyzing indicators. A bunch of previously dead-ended approaches (loading gif, analyzing label, color from churn probability) together.

Lots of nice touches on this landing page. Clean and welcoming but also with some real work in the details. I like the icons at the bottom, very clean.

From: https://abstractapp.com/

Made using the quickdraw data set and openframeworks. So it looks like maybe she defines sections and then it fills in with a certain type of drawing. Whatever packing algorithm is happening is really nice.

From: https://twitter.com/frauzufall/status/867485105434095618

Working on a writing post set-up for the new version of this blog/website.

Text-constructed (probably there is a different phrase for that) dungeon adventure game. I like the colors.

From: https://sites.google.com/site/broguegame/home

Got the waterfall going. The analyzing animation has the colors shifting at random. I think I'm going to need to go back to a horizontal waterfall for them to feel good.

A more literal simulation of what LIME does, as an experiment for the loading animation. Too confusing. The most literal would be to just perturb one item in a column at a time. But that seems like it would be a little slow and pedantic. I'm going to go back to color experiments, which are less literal but I think will have the right feeling.

Working on a new version of this blog.

Wang tiles can tile a plane aperiodically.

From: https://en.m.wikipedia.org/wiki/Wang_tile

More individual cell behavior for the analyze animation.

Experimenting with 'analyze' animations for customer rows in the churn explainer prototype.