Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
altairiumblue
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
31.
▲
by
altairiumblue
8y ago
As a complete side note - it's cool to see Altair charts used in the wild :)
32.
▲
by
altairiumblue
8y ago
As someone who doesn't use C++, can anyone explain in a straightforward way why a low level language is going through so many new versions? And is there big difference between the latest version and C++ from 5 or 15 years ago? How do y
33.
▲
by
altairiumblue
8y ago
Your very last point should be heavily against R, not in favour. Having multiple ways to do the same thing is fine when you work by yourself, but in a large team it makes the style of the codebase vary a lot based on who wrote each block of
34.
▲
by
altairiumblue
8y ago
Don't look for sources that are biased in a different way. The solution to spin in the news really shouldn't be spin in the opposite direction - that would just add another source of misinformation. The solution is consuming only
35.
▲
by
altairiumblue
8y ago
Regardless of the answer, I have plans to learn it simply out of curiosity. I'm just not yet convinced that there are any practical benefits to using Julia over the other languages.
36.
▲
by
altairiumblue
8y ago
As someone who has never run into performance problems with R, and also knows how to use Python - is there a good reason to learn Julia?
37.
▲
by
altairiumblue
8y ago
I don't think your post makes any sense. It's on AskHN and you don't have a question. Also... - How did you define and measure progress? - How did you measure the number of people working in the field? - If true, why would an
38.
▲
by
altairiumblue
8y ago
Maybe you got misled by the title, but I don't think this is a good introduction for someone who wants to get started with time series forecasting.
39.
▲
by
altairiumblue
8y ago
Robert Nau's notes on time series are a good start - https://people.duke.edu/~rnau/411home.htm Keep in mind: - not everything that can be represented as time series, should be - not everything that can be represen
40.
▲
by
altairiumblue
8y ago
Stop saying things that make you weak - https://www.youtube.com/watch?v=IAFO3v_Ucio Hacker News is generally sceptical toward Peterson and there are good reasons to be but this is very relevant and something that I have exp
41.
▲
by
altairiumblue
8y ago
- you can set up toggle in a way that isn't distracting - definitely disable the popups, have a keyboard shortcut for start/stop tracking, or use the auto tracking feature if you like it - I've found time tracking with toggl
42.
▲
by
altairiumblue
8y ago
The best places to start for a complete beginner are Precalculus and Hello-World in C. I'm serious about this. Ultimately the job is just software development plus statistics. If you are a software developer, work on your statistics. I
43.
▲
Ask HN: Best Way for s.o Who Knows Statistics to Contribute to Medical Research?
2 points
by
altairiumblue
8y ago
|
0 comments
44.
▲
by
altairiumblue
8y ago
My impression after reading your answers in this thread - you're trying really hard to come up with a use case for blockchain. You didn't start with the goal to create a good publishing platform and naturally get to blockchain as
45.
▲
by
altairiumblue
8y ago
Personal opinion - I really dislike the term and all the overpriced drag-and-drop, "automated", "easy-to-use" tools that are associated with it. Generally, I would choose open-source over licensed software and a programm
46.
▲
by
altairiumblue
8y ago
- Why would content creators choose to use this service over youtube or twitch? - And why would users choose it over the other platforms? - And if they did, how would the experience be different for each party?
47.
▲
by
altairiumblue
8y ago
Can you explain what your product does without using the word blockchain?
48.
▲
The Rise of the Citizen Data Scientist [pdf]
(pages.dataiku.com)
58 points
by
altairiumblue
8y ago
|
3 comments
49.
▲
When Surveillance Meets Incompetence
(techcrunch.com)
3 points
by
altairiumblue
8y ago
|
0 comments
50.
▲
by
altairiumblue
8y ago
Without knowing too much about your situation (and being a bit envious of it), my initial thoughts are: - The company's inefficiency is an opportunity for you - you clearly have ideas on how things can be improved. Develop those into a
51.
▲
by
altairiumblue
8y ago
Why would they be? I'm not being smart - this is the genuine way to approach the problem. At some point you started with no protected classes and if there was a reason that was considered good enough by enough people, then the list w
52.
▲
by
altairiumblue
8y ago
Jupyter has docstrings and autocomplete for dot calls - I just tried your Series.str.contains example in JupyterLab 0.32. And in general my experience with the documentation and autocomplete has been the exact opposite - way better than the
53.
▲
by
altairiumblue
8y ago
That's a cool idea. Just yesterday I posted a thread about finding Discord/Slack groups and it didn't even occur to me that I can start one. However, my impression is still that these groups quickly degenerate into a mix of s
54.
▲
Apple, the iPhone, and the innovator's Dilemma
(wired.com)
1 points
by
altairiumblue
8y ago
|
0 comments
55.
▲
by
altairiumblue
8y ago
Yup. As someone who is in a junior position in the field - I'm torn between, on one hand, riding the wave so to speak and taking advantage of all the buzzwords that I can put on my resume (which is fine with me because I can back them
56.
▲
Ask HN: How do you find Slack/Discord groups of people that do similar work?
2 points
by
altairiumblue
8y ago
|
1 comments
57.
▲
Ask HN: Have you used Alteryx? What are your thoughts on it?
1 points
by
altairiumblue
8y ago
|
1 comments
58.
▲
by
altairiumblue
8y ago
> work out what we're going to do in the new reality whether we like it or not. The reality in question is being denied, ignored or lied about. > Ridiculing half the population as uneducated xenophobes isn't going to achieve
59.
▲
by
altairiumblue
8y ago
It's hard to pin it down as influential but when a new episode comes out, it feels like you're about to hang out with friends that you haven't seen for a few weeks.
60.
▲
by
altairiumblue
8y ago
You can't "learn" machine learning (whatever that entails) in 3 months. If you want to actually build up some skills, you should look further than a youtube video with a clickbait title.
More ›