aquafox

11 ms·

61.
by aquafox 2y ago
I always struggle to get a good intuition into models with interaction terms. I usually try to write down for every class of responses which terms of the model go into it and often that helps with interpretation. There's also the Explo
62.
by aquafox 2y ago
It depends. The most important assumption is independence of the observations. If that is not given, you have to either account for correlated responses using a mixed-effects model or mean-aggregate those responses (computing the mean decre
63.
by aquafox 2y ago
If you like shrinkage (I do), I highly recommend the work of Matthew Stephens, e.g. ashr [1] and vash [2] for shrinkage based on an empirically derived prior. [1] https://cran.r-project.org/web/packages/ashr/i
64.
by aquafox 2y ago
For my applications, using natural cubic splines provided by the 'ns' function in R, combined with trying out where knots should be positioned, is sufficient. Maybe have a look at the gratia package [1] for plotting lots of diagno
65.
by aquafox 2y ago
Most people don't appreciate linear regression. 1) All common statistical tests are linear models: https://lindeloev.github.io/tests-as-linear/ 2) Linear models are linear in the parameters, not the response! E.g
66.
by aquafox 2y ago
Why not use a private key based on a bunch of lottery draws that happened just before the date?
67.
by aquafox 2y ago
Every cyclist has a powermeter and the power is recorded. Why not require them to upload the power file directly after the race? Since they all ride the same course, you can correlate power output with speed (taking riders height and weight
68.
by aquafox 2y ago
There's already commandlinefu.com
69.
by aquafox 2y ago
There's a short proof of a 2D version of the paradox on page 684 of the Princeton Companion to Mathematics: https://sites.math.rutgers.edu/~zeilberg/akherim/PCM.pdf
70.
by aquafox 2y ago
It seems you are describing the Cocktail party effect: https://en.m.wikipedia.org/wiki/Cocktail_party_effect
71.
by aquafox 3y ago
Fun fact: The sequences of ceiling( 2 / (2^(1/n)-1) ) and floor( 2n / log(2) ) have identical elements for a very long time. The first time they differ is n = 777451915729368. https://oeis.org/A129935
72.
by aquafox 3y ago
Haribo is named after it's founder and location: HAns RIegel, BOnn.
73.
by aquafox 3y ago
I agree with your points and thats why it's useful to compare a MLE to an alternative model via a likelihood ratio test, in which case one sees how much better the generative model performs as compared to the wrong model. Similarly, AI
74.
Ask HN: For what purposes do you use a Raspberry Pi?
5 points by aquafox 3y ago|4 comments
75.
by aquafox 3y ago
In the original TTD, you could trigger an integer overflow by building a very long tunnel from one end of the world to the other, so that the billions of money subtracted got negative. ;) https://www.playdosgames.com/online
76.
by aquafox 3y ago
If you are into this topic, I recommend "Love and Math" by Edward Frenkel, where he describes his struggles to start an academic career as a jewish mathematician in russia.
77.
by aquafox 3y ago
Maybe not a perfect replacement, but have you tried Feedly? https://feedly.com/
78.
by aquafox 3y ago
> His is an extremely rare form, caused by a mutation in a single gene, otoferlin No. Otoferlin is the protein encoded by the OTOF gene. I wish science journalists would be more science literal.
79.
by aquafox 3y ago
I've memorized the first 130 digits of pi. Should be good enough for most tasks that require a human RNG.
80.
by aquafox 3y ago
Can't you do this with magic wormhole and it's Android client https://f-droid.org/de/packages/com.pavelsof.wormhole/ ?
81.
by aquafox 3y ago
Becaus you can't do that if you want a professional photo of your baby.
82.
by aquafox 3y ago
I took Eric Landers "Introduction to Biology" course before transitioning into bioinformatics ( https://www.edx.org/learn/biology/massachusetts-institute-of... ). Can highly recomment this course, although
83.
by aquafox 3y ago
ISDN was a true gamechanger for me. Before that, we had a 56k modem and I couldn't go online without blocking my parents' phone connection. Also, with ISDN came our first internet flatrate.
84.
by aquafox 3y ago
Samsung devices have moisture detection via USB-C for a long time. It does not work very reliable and many people have problems ( https://ifixit.com/Antworten/Ansehen/511762/How+to+stop+the+... ). I randomly
85.
by aquafox 3y ago
According to the tidyverse style guide, using magrittr's pipe assignment operator %<>% (also mentioned in the blog post), should actually be avoided. I guess because it makes code reading harder.
86.
by aquafox 3y ago
I don't understand why governments allow to privatize hospitals or other public ressources, often with the reasoning that otherwise they wouldn't be profitable. We don't expect other public services like fire departments or r
87.
by aquafox 3y ago
"Success is only rented, and the rent is due every day." -Hansi Flick, as Coach of FC Bayern Munich
88.
by aquafox 3y ago
> the constitution forbids to convict someone two times for the same crime He wasn't convicted, which means the reason he isn't convicted for the first time now is because he is not allowed to be convicted twice. That doesn&#x2
89.
by aquafox 3y ago
What helped me most is to separate the task of writing content from writing the nicely (copy-editing): https://clauswilke.com/blog/2013/09/22/silence-your-inner-cr... It is also more aligned with how on
90.
by aquafox 3y ago
All this would not be neccessary if they just deleted the data afterwards, but I didn't see this advertised. As vatys explains in his longer comment: You are just as anonymous as the least anonymous person related to you.

More ›