grayclhn

24 ms·

241.
by grayclhn 11y ago
I would argue that most of the scientific & statistical packages for most languages are driven by at most a handful of people, yes. Another interpretation is that R is an incredibly productive language for this sort of programming, othe
242.
by grayclhn 11y ago
Probably a decent rule of thumb is: does disclosing the source of funding make the research conclusion less credible? In TFA -- yes, to the point that it wasn't fully disclosed. For publicly funded research -- probably not.
243.
by grayclhn 11y ago
There's a pretty big difference between research funding from a private group that wants support/pr for actions it's already decided to take, and research funding that's allocated by a group of one's scientific peer
244.
by grayclhn 11y ago
Because you may want or need to change aspects of the behavior, but not while refactoring.
245.
by grayclhn 11y ago
I would rephrase that as "putting the company's mission above specific instructions that are detrimental to that mission." Intent and judgment matters a lot here: if someone has good judgment and is sincerely trying to make t
246.
by grayclhn 11y ago
> Again, what is the blame on Facebook as if they came in and registered your name and profile by force? Pretty sure you're trolling at this point, but if not google "facebook shadow user profile" and get your mind blown.
247.
by grayclhn 11y ago
First sentence of the fucking article: "Facebook claims its practice of forcing users to go by their 'real names' (or 'authentic identities' as Facebook spins it) makes the social network a safer place." Second
248.
by grayclhn 11y ago
> Obviously if someone is stalking you or something, that's a different story. Yup.
249.
by grayclhn 11y ago
That's a very different argument than "I dont get it - Facebook has some of the most robust privacy settings of any platform"
250.
by grayclhn 11y ago
Sure, but users who are legitimately concerned about privacy would be pretty foolish to rely on Facebook's settings, which was the comment I replied to. It's not like we're talking about behavior from 20 years ago. To be fair
251.
by grayclhn 11y ago
The real-name local facebook group for my small town is a cesspool, so ymmv.
252.
by grayclhn 11y ago
This is absolutely not true historically. I don't know what their privacy policies are now, but they have a track record of changing them with very little fanfare.
253.
by grayclhn 11y ago
I don't think Facebook advertises that it's a horribly bad service for domestic violence victims, etc. If they did, this policy would be more justified.
254.
by grayclhn 11y ago
Okay, for that stuff probability theory is too abstract. For basic basics, Edward Tufte has a $2 ebook that's pretty good: Data Analysis For Politics And Policy[1] and for terminology in causality, Rubin has a short open access paper[2
255.
by grayclhn 11y ago
Probability's hard to teach. You can give informal statements and kind of wave your hands at the underlying theory, or you can give a rigorous well-founded treatment that's intellectually satisfying. But the rigorous foundation us
256.
by grayclhn 11y ago
That was my reaction at the beginning of the article too (um, in part); I'm not sure exactly how I'd respond to an inspirational speech denigrating the bay bridge, but I wouldn't feel good.
257.
by grayclhn 11y ago
I think you're misunderstanding. Standard basic (measure-theoretic) probability theory is designed to handle common non-discrete real-world cases: continuous random variables like height, temperature, etc. They're not approximated
258.
by grayclhn 11y ago
It's maybe not perfect, but can you have them working in a shared dropbox directory in class? (Or something equivalent.) Then just open or refresh the file... IPython notebooks running on a central server is potentially another option
259.
by grayclhn 11y ago
Not true for everyone. Search for "Chuck Grassley twitter"
260.
by grayclhn 11y ago
All articles at wsj.com are behind a paywall. As other comments have pointed out, you can get past the paywall through google.
261.
by grayclhn 11y ago
Then tip more if you think the driver drove especially safely. :)
262.
by grayclhn 11y ago
I'm kind of amazed at the negative reactions this comment got. I'd bet the reason is, "the companies think it would cost more," but god yes it would be more convenient.
263.
by grayclhn 11y ago
It's not that complicated. Have the user toggle an "I also claim ownership of this directory" flag before letting them edit it and have it count against their storage after that. (No idea how difficult it would be to introduc
264.
by grayclhn 11y ago
"Looking at who complains" isn't evidence of that, though. One could argue that the low quality users are so turned off by the mobile experience that they can't even complain about it. That said, I wouldn't mind a b
265.
by grayclhn 11y ago
> But most of the complaints about that are coming from existing Hacker News users, for whom the experience on mobile is terrible. I'm not sure what your argument is here. Who else would complain? The people who have been driven awa
266.
by grayclhn 11y ago
Who thinks he's nice?
267.
by grayclhn 11y ago
Yes and no; base R is extremely stable, but if you need to install dplyr, maggrittr, ggplot2, Rcpp and knitr/rmarkdown to reproduce a large fraction of the work that's been done in the last few years that seems somewhat equivalent
268.
by grayclhn 11y ago
Whether Julia replaces R has more to do with how R evolves over the next 5 years than Julia IMO. As far as "production use" I think that depends on whether what you're doing is well served by the existing libraries or not.
269.
by grayclhn 11y ago
Fair enough, but it's also reasonable to approximate 9.33 × 10^-296 as 0. I felt like doing the calculations :)
270.
by grayclhn 11y ago
It turns out not to matter (since the numbers are so small) but your math implies that 10^305 people guessing would give a probability of about 9,300 that at least one of them is right on all 1000 coin flips. Since probabilities can't

More ›