Timeline for Bayesian estimation of $N$ of a binomial distribution
Current License: CC BY-SA 4.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 28 at 16:44 | history | edited | kjetil b halvorsen♦ | CC BY-SA 4.0 |
deleted 106 characters in body
|
| Sep 7, 2014 at 18:28 | comment | added | Sycorax♦ |
@COOLSerdash You may be interested in [this][1] question, where I ask which of the grid results or rstan results are more correct. [1] stats.stackexchange.com/questions/114366/…
|
|
| Sep 4, 2014 at 15:00 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 101 characters in body
|
| Sep 1, 2014 at 13:46 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 23 characters in body
|
| Sep 1, 2014 at 13:33 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 534 characters in body
|
| Sep 1, 2014 at 13:22 | comment | added | COOLSerdash |
Yes, I've tried to use round to no avail. I'll use R then. Thanks again for all your help!
|
|
| Sep 1, 2014 at 13:21 | comment | added | Sycorax♦ |
Mysteriously, the round() function in stan produces real-valued (!) outputs. Instead of using the GQ block, you'll have to gather theta[1] and N from the posterior samples and generate random deviates in R.
|
|
| Sep 1, 2014 at 7:14 | comment | added | COOLSerdash | Sorry for bothering you again, but say I wanted to predict the next $y_{i}$ from your model. Do you know how this could be done in the "generated quantities" block of Stan? | |
| Sep 1, 2014 at 6:35 | comment | added | COOLSerdash | Around 2 minutes on my Desktop. | |
| Sep 1, 2014 at 6:33 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
edited body
|
| Sep 1, 2014 at 6:27 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 114 characters in body
|
| Sep 1, 2014 at 6:27 | comment | added | COOLSerdash |
Yes! That was exactly my problem. n can't be declared as an integer and I didn't know a workaround for the problem.
|
|
| Sep 1, 2014 at 6:24 | comment | added | Sycorax♦ | The one hiccup with stan for this problem is that all parameters must be real, so that makes it a little inconvenient. But since you can penalize the log-likelihood by any arbitrary function, you just have to go through the trouble to program it... And dig out the composed functions to do so... | |
| Sep 1, 2014 at 6:21 | comment | added | COOLSerdash | +1 and accepted. I'm impressed! I also tried to use Stan for a comparison but couldn't transfer the model. My model takes about 2 minutes to estimate. | |
| Sep 1, 2014 at 6:19 | vote | accept | COOLSerdash | ||
| Sep 1, 2014 at 5:54 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
edited body
|
| Sep 1, 2014 at 5:45 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
deleted 111 characters in body
|
| Sep 1, 2014 at 5:12 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 1691 characters in body
|
| Sep 1, 2014 at 4:04 | history | edited | Sycorax♦ | CC BY-SA 3.0 |
added 6 characters in body
|
| Sep 1, 2014 at 3:58 | history | answered | Sycorax♦ | CC BY-SA 3.0 |