Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
coherentpony
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
87 ms
·
901.
▲
by
coherentpony
14y ago
Try merging a .nib. You'll see first-hand how hard it is.
902.
▲
by
coherentpony
14y ago
Unit testing a deterministic one is not too bad. Unit testing a stochastic physical system is a nightmare.
903.
▲
by
coherentpony
14y ago
Not available in the UK store.
904.
▲
by
coherentpony
14y ago
Great visualisation.
905.
▲
What keeps Light Table from being your primary editor?
(docs.google.com)
3 points
by
coherentpony
14y ago
|
2 comments
906.
▲
by
coherentpony
14y ago
From a mathematical perspective, it looks as though your statement could imply that GA may potentially be viewed as a finite dimensional analogue of GP. Interesting.
907.
▲
by
coherentpony
14y ago
>You misconstrued. Reject as little as possible (<> reject little) was the motivation with respect to more naive random walks which suffer at dimensionality. Misunderstanding on my part; apologies. Can you clarify the statement
908.
▲
by
coherentpony
14y ago
>The idea is to approximate the density of P(x) while trying to reject as little as possible. Hence less wasteful. The idea is to approximate P(x) period. Rejecting as little as possible means nothing other than one's choice of proposa
909.
▲
by
coherentpony
14y ago
Annealing is hit or miss. I often have to sample a distribution on a domain that approximates an infinite dimensional space and annealing doesn't cut it for me. There are just far too many modes. Personally, I'd like to see what GAs and
910.
▲
by
coherentpony
14y ago
>This is counter to the more targeted idea of MCMC, which generates a bunch of correlated samples to more efficiently sample the distribution. The idea of MCMC is not to more efficiently sample a distribution through to use of correlate
911.
▲
by
coherentpony
14y ago
Interesting idea! Thanks for sharing. Just a quick note, the pricing page ( http://jellifi.com/Pricing ) needs to be fixed, it still has the standard 'Lorem ipsum...' filler text.
912.
▲
by
coherentpony
14y ago
PyCon incident? You mean the incident where they gave away RaspberryPis to everyone, including females? Or the one where they raised tens of thousands of dollars for PyLadies? Or maybe something a little more indirect, like perhaps all t
913.
▲
by
coherentpony
14y ago
This is wonderfully written. Much better than a lot of other content that makes it to the front page.