1,459 reputation
934
bio website
location
age
visits member for 3 years, 4 months
seen 20 hours ago

2d
comment What is the difference between R hat and psrf?
Thanks Daniel! What is the point of splitting the chain in half in Stan? Why is it done in Stan and not in WinBUGS/JAGS?
2d
asked What is the difference between R hat and psrf?
2d
comment Calculate R-squared with JAGS and R
Thomas, 1) why do you use dgamma prior - it doesn't allow negative values of your linear coefficients, are you aware of it? It is not a standard linear regression then. 2) Why do you use inprod instead of classic full formula with + and * - is it faster in JAGS? Thanks!
Nov
20
comment How can regression trees be fit in WinBUGS/OpenBUGS/JAGS?
I am afraid this wouldn't work. Are you sure you understand how tree regression works? The method will first split using a variable with highest impact - i.e. the most important variable. The method will automatically select the order of variables by importance. There is no fixed order, like "this is variable1, this is variable2...". Random forests are even more complicated, they build plenty of such trees.
Nov
19
comment CAR spatial models in JAGS
Why don't you download the source code of GeoBUGS to see? Hardly anyone besides GeoBUGS authors will know that....
Nov
19
comment How can regression trees be fit in WinBUGS/OpenBUGS/JAGS?
@Tim, thanks. You mentioned Stan, do you think it can do it?
Nov
14
comment Zeros in non-negative data: can you multiply by 10^c, round, and then model as a count variable?
@ACD you can try it, but (intuitivelly) I would be skeptical towards treating non-count data as count data. But certainly try it and look at the residuals if it fits much better than the continuous models. But I would also advise you to try "more proper" alternatives like spline regression and regression trees and do some crossvalidation tests to see which model is actually fitting better.
Nov
14
awarded  Revival
Nov
11
comment Testing whether the coefficients of one model and another are statistically significantly different
@Maarten what I consider confusing on your answer is not the words, but the formula - your formula works, but it looks like you are treating $z$ as continuous variable instead of categorical one.
Nov
11
answered Testing whether the coefficients of one model and another are statistically significantly different
Nov
9
awarded  Popular Question
Nov
9
comment Zeros in non-negative data: can you multiply by 10^c, round, and then model as a count variable?
@ACD: "Your example doesn't really represent my data so well." - that's another argument why you shouldn't use poisson! My example shows poisson distribution, which you would assume in your data by using poisson model, and how this poisson model depends on the multiplying constant. The constant changes things a lot! It's because poisson assumes that variance = mean value.
Nov
8
comment Zeros in non-negative data: can you multiply by 10^c, round, and then model as a count variable?
@ACD, I guess the result will be less sensitive to c than to the multiplying constant you are proposing. 2) are you sure NB handles the problem completely? You wrote that it underestimates SEs.
Nov
8
answered Zeros in non-negative data: can you multiply by 10^c, round, and then model as a count variable?
Nov
6
comment It is possible to implement a regression algorithm using Random Forests?
This is what random forests are for, no? For regression! So random forests already are a regression algorithm. Or am I wrong?
Nov
6
revised Can multinomial distribution be simulated by a sequence of binomial draws?
small index correction
Nov
6
comment Can multinomial distribution be simulated by a sequence of binomial draws?
Thanks Xi'an for confirming me what I thought! Made a small index correction in your answer.
Nov
6
suggested suggested edit on Can multinomial distribution be simulated by a sequence of binomial draws?
Nov
6
comment Is multinomial the same as multiple binomial?
Thank you, now it is clear! I was confused a bit. I posted a new, less stupid question now, would be grateful if you can look at it: stats.stackexchange.com/q/122907/5509
Nov
6
asked Can multinomial distribution be simulated by a sequence of binomial draws?