Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
nrjames
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
36 ms
·
361.
▲
by
nrjames
10y ago
While gaming oriented, there's a great group of people who discuss pretty much everything over at the Quarter to Three forums. https://forum.quartertothree.com/
362.
▲
by
nrjames
10y ago
This reminds me of Sweave for R.
363.
▲
by
nrjames
10y ago
The long history of Python support in the scientific community has produced a rich catalog of domain-specific analysis libraries. Warts and all, it's simply not possible, in most languages, to achieve the analytical productivity you ge
364.
▲
by
nrjames
10y ago
This is great. You know what would really take it to the next level? Tides! It would be awesome if micro-adjustments along the coast showed the times of low and high tides. My parents, who live along the coast, would use it multiple times&#
365.
▲
by
nrjames
10y ago
West with the Night, by Beryl Markham. This is the autobiography of a woman who grew up on a farm in British East Africa (Kenya) in the early 1900s. She eventually became a bush pilot and the first person ever to fly an airplane solo across
366.
▲
by
nrjames
10y ago
Shiffman also has a "Coding Rainbow" series on YouTube that is pretty great. He has a good personality for goofy YouTube programming teacher. https://www.youtube.com/user/shiffman
367.
▲
by
nrjames
10y ago
I'm a big fan of Processing or p5.js for teaching kids that age how to write code. The immediate visual feedback from a line or two of code creates a positive feedback loop that encourages learning. Shiffman's book Learning Proces
368.
▲
by
nrjames
10y ago
I would structure your resume to focus on your projects, giving them detailed descriptions, in which you can describe the languages, libraries, etc... that you used to implement them. That way, you put the technology in the context of the s
369.
▲
by
nrjames
10y ago
Well, that's sort of how d3.js grew.
370.
▲
by
nrjames
10y ago
I did what you are considering for the reasons you mention. I sorely miss the freedom of being self-employed. My employer is extremely flexible, but nevertheless, I almost never have time for my personal projects. I'm far less efficien
371.
▲
by
nrjames
10y ago
My dream is to see one of these graph databases use NetworkX for the python API and the database for the back end / storage.
372.
▲
by
nrjames
10y ago
Is there a reliable python module for working with ArangoDB 3.0?
373.
▲
by
nrjames
10y ago
It is useful for people to understand what goes into developing software in the same way that it is useful for people to learn geometry in high school. Code is the most tangible form of algorithmic thinking that is accessible to most people
374.
▲
by
nrjames
10y ago
The Hello World Open was super fun a few years ago; code the AI for a slot car. Their site says they'll hold it again this year: http://helloworldopen.com/
375.
▲
by
nrjames
10y ago
It's almost just like this children's book! Weasels: Elys Dolan http://www.amazon.com/Weasels-Elys-Dolan/dp/0763671002/ref=s...
376.
▲
by
nrjames
10y ago
Not to mention the fact that some research suggests they blast bacteria all around the bathroom/environment. http://www.telegraph.co.uk/news/2016/04/13/dyson-airblades-s...
377.
▲
by
nrjames
10y ago
This is your current best resource: http://www.creativeai.net/ And this is an intro to that site, in a way: https://medium.com/@ArtificialExperience/creativeai-9d4b2346...
378.
▲
by
nrjames
10y ago
You could easily extrapolate from those textures to a Wang tileset, as explained in this old paper from Microsoft Research "Wang Tiles for Image and Texture Generation": https://pdfs.semanticscholar.org/8d96/c
379.
▲
by
nrjames
10y ago
What comes with the PhD positions is a requirement to respond to RFPs with concept notes, proposals, etc... and a lot of high level project management. They provide subject matter expertise from an analytic angle (usually without getting in
380.
▲
by
nrjames
11y ago
I work as a data scientist at a nonprofit research organization (rti.org) and have to apply far more skills than the PhD researchers here. While we fall back on them for depth in subject matter expertise, we also have to understand conceptu
381.
▲
by
nrjames
11y ago
Yep, I'm well aware of all of that because I sit on our "Cloud Technology Advisory Committee," which feels like a scene from Brazil. However, many here work with sensitive PHI and various types of classified data, so the conc
382.
▲
by
nrjames
11y ago
In an enterprise environment, Slack is sometimes considered bad because all of the company data is not under the company's control. I seriously doubt that the company I work for ever would go all in on Slack, precisely for that reason.
383.
▲
by
nrjames
11y ago
Not to be confused with MacKenzie-Childs :) http://www.mackenzie-childs.com/
384.
▲
by
nrjames
11y ago
I like the site but for usability, I would prefer to see all of the hex values for a given palette at once instead of hovering over each color individually.
385.
▲
by
nrjames
11y ago
It's an abstraction layer on top of SQLAlchemy. There might be some performance hits, but none that have made me question whether I should be using it. The convenience of having a more pythonic way to work with data for ETL scripts is
386.
▲
by
nrjames
11y ago
Dataset is awesome. I've been using it consistently for the past several years. It is lovely in ETL scripts, especially considering you can use it just to send regular SQL when needed.
387.
▲
by
nrjames
11y ago
Interesting that Microstrategy also is down quite a bit.
388.
▲
by
nrjames
11y ago
Donate them to the Library of Congress?
389.
▲
by
nrjames
11y ago
Deep learning stuff. We push to servers for deployment but it is very convenient to develop/test on a laptop first.
390.
▲
by
nrjames
11y ago
I would like to purchase a Macbook Pro with a dedicated Nvidia GPU, so that I can use CUDA. The only viable option is a 2 yr old refurbished MBP for $2k. So... what now? I'm looking for a good laptop with an Nvidia chip that will run L
More ›