jsvine

An associated paper in Scientific Data: https://www.nature.com/articles/s41597-022-01894-2

Thanks for flagging this confusion. It's helpful to hear how this is perceived. Short answer: The spreadsheet is, indeed, embedded in a broader context. (And has a pointer to that context in the "Notes" tab.) Slightly longer explanation in my main comment on this thread.

That's an interesting suggestion, thanks. Certainly worth looking into. In the meantime, you can consider the newsletter's landing page to be a canonical URL of sorts — it will always link to the structured archive: https://tinyletter.com/data-is-plural

Hi, author/creator here. Very neat to see this on HN, thanks. The spreadsheet is a byproduct of the weekly newsletter I publish: https://tinyletter.com/data-is-plural

The spreadsheet simply contains the text and links from each newsletter edition ... but in a tabular format. (One advantage to the newsletter over the spreadsheet: the links make a bit more sense, since they're associated with specific anchor text.) The "non-structured" archive of previous newsletters can be found here: https://tinyletter.com/data-is-plural/archive

Happy to answer any Qs!

I've been using VisiData for about nine months now, and have found it to be tremendously useful. These days, it's typically the first first tool I reach for when examining/exploring a new dataset.

In fact, I've been liking it so much that I wrote a tutorial for first-time users: https://jsvine.github.io/intro-to-visidata/

(Feedback very much welcome / appreciated.)

Updated, merged, and pushed to PyPi as part of v0.1.0: https://github.com/jsvine/waybackpack/pull/5

Thanks again for the feedback. Really appreciate it — and the existence of the Internet Archive and Wayback Machine.

Hi, Greg! Library author here. I'd be happy to add a configurable UserAgent. Perhaps the default would be a generic "waybackpack" but could be configurable to add contact info for the user. Does that sound about right? Prefer a different approach?

And, yep, the library is intentionally designed only to request one snapshot at a time.

Layperson methodology here: http://www.buzzfeed.com/johntemplon/how-we-used-data-to-inve...

Detailed methodology/code here: https://github.com/BuzzFeedNews/2016-01-tennis-betting-analy...

"Whether you realize it or not, every major online system we use today creates unique URLs for each distinct idea."

Thanks! I haven't yet, but that's a great idea. Will do so shortly.

Agreed. The archive uses the first line of the newsletter as the summary. I've been using the same tagline as the first line. But, for future newsletters, I think I'll make the first line more descriptive. E.g., for the most recent: "Maternity leave, art collections, firearm dealers, New Guinea languages, and aircraft-wildlife strikes." Would that be an improvement?

Good point. I've now updated the text on the landing page, to the extent that's possible with TinyLetter.

TinyLetter archive: https://tinyletter.com/data-is-plural/archive

Structured archive: https://docs.google.com/spreadsheets/d/1wZhPLMCHKJvwOkP4jucl...

Most recent edition, sent yesterday: https://tinyletter.com/data-is-plural/letters/data-is-plural...

TinyLetter archive: https://tinyletter.com/data-is-plural/archive

Structured archive: https://docs.google.com/spreadsheets/d/1wZhPLMCHKJvwOkP4jucl...

Most recent edition, sent this morning: http://tinyletter.com/data-is-plural/letters/data-is-plural-...

The story behind the code: "LAPD underreported serious assaults, skewing crime stats for 8 years" http://www.latimes.com/local/cityhall/la-me-crime-stats-2015...

Added docstrings in v0.0.6: https://github.com/jsvine/spectra/issues/1

Another thanks for flagging, mark-r. I've updated spectra to address this issue more explicitly. Per colormath[0], colors are allowed to go out of gamut. Colors now have `mycolor.rgb` and `mycolor.rgb_clamped` properties; `mycolor.hexcode` uses the clamped values. Are there other ways you'd like to see the out-of-gamut issue addressed? Open to suggestions.

[0]: http://python-colormath.readthedocs.org/en/latest/conversion...

Yes, the library definitely does need some stronger specs. Thanks for flagging. Addressing these two specific issues:

> Is the rgb space actually sRGB?

Yes. Added: https://github.com/jsvine/spectra/commit/7f1ac6f5287340ed356...

> And what is the parameter to brighten/darken/saturate/desaturate?

The parameter for brighten/darken is a positive/negative linear adjustment to the L(ightness) value of the color's Lab representation. (Spectra converts the color to Lab, makes the change, and then converts back to the original color space.)

Likewise, the parameter for saturate/desaturate is a positive/negative linear adjustment to the c(hroma) value of the color's Lch representation.

Added a note here: https://github.com/jsvine/spectra/commit/fd3cbde9338ce388849...

The goal is to, ultimately, have all these details in the docstrings: https://github.com/jsvine/spectra/issues/1

Fair criticism. I should add those docstrings. Issue opened: https://github.com/jsvine/spectra/issues/1

> It looks like a helpful tool for situations where the standard colormaps in matplotlib/pyplot don't work well. Is that an appropriate use case?

Yes, that's definitely the type of use-case I had in mind. Another big one: Choropleth maps.

> I also like seeing nbviewer used for documentation, I haven't seen that very often.

Glad to hear it! It's my first attempt at writing documentation in an IPython notebook. Seemed like a good fit for something so visual.

I hadn't heard of OpenColourIO; it looks fascinating and quite powerful. Thanks for the tip!

Thanks! Colormath does an amazing job handling the mathematical conversions between color spaces, but has (at least to me) a somewhat verbose API. Spectra wraps that API in something simpler, and adds scales, ranges, blending, brightening/darkening, and saturation/desaturation.

Spectra uses grapefruit for parsing HTML names and hexcodes.

> ‘Cloud Face’ is a collection of cloud images that are recognized as human face by a face-detection algorithm. It is a result of computer’s vision error, but they often look like faces to human eyes, too. This work attempts to examine the relation between computer vision and human vision.

Related graphic here: http://www.scientificamerican.com/article/tiny-genetic-diffe...

Demo here: https://jsvine.github.io/notebookjs/demo/

> Throughout the past six years, the project has been stuck in the beta phase. According to McKinsey, "for past 5 years, Release 1.0 is consistently projected to be 24-32 months away."

GitHub repo and PDF here: https://github.com/jesusfv/Comparison-Programming-Languages-...

Author here. Thanks for posting! HN might be interested in the related repos we open-sourced today:

- Data-processing scripts and HTML/JS/CSS for the maps and charts: https://github.com/BuzzFeedNews/2014-06-bikeshare-gender-map...

- Python parsers for the various bikeshare services' published data: https://github.com/BuzzFeedNews/bikeshares

- General guide to getting data from these services: https://github.com/BuzzFeedNews/bikeshare-data-sources

I like this. Reminds me a (tiny) bit of Bruce Nauman's "One Hundred Live and Die": http://www.art21.org/images/bruce-nauman/one-hundred-live-an...

I'm the guy who built that, and I agree. I tinkered with having another "Show Code" button at the top, but my attempts looked bulky and distracting. Still trying to figure out a better way.

In the meantime, pull requests quite welcome: https://github.com/jsvine/reporter