derived packages list on web site?
30 views
Skip to first unread message
Bob Carpenter
Mar 27, 2016, 10:37:59 PM3/27/16
to stan...@googlegroups.com
I imagine we're going to see more packages based on
RStan, etc., like this one from Young Ahn for doing
decision analysis.
Do you think I should add a spot on the web page to
list these. And if so, where? If I put them under
Citations, nobody will find them. I could create new
pages linked off of one of the interface pages, like
RStan (but this particular package also uses Loo).
I don't want to give people the impression that we're
vouching for the external packages somehow; but it'd
be nice to list the ones that are there.
- Bob
> On Mar 27, 2016, at 6:20 PM, Ben Goodrich <goodri...@gmail.com> wrote:
>
> Good to see. For any R package like this offering a fixed set of Stan programs, you would be better off going the rstanarm route and compiling them at or before installation time rather than at runtime. There is a rstan.package.skeleton() function in rstan that facilitates doing it this way. Or if you are going to IMPS, I could help you convert it then.
>
> Ben
>
> On Sunday, March 27, 2016 at 5:52:53 PM UTC-4, Young Ahn wrote:
> Hi all,
>
> Just want to announce that my lab has released an R package called hBayesDM (hierarchical Bayesian modeling of Decision-Making tasks), which is based on RStan. It is now available on CRAN (https://cran.rstudio.com/web/packages/hBayesDM/).
>
> The hBayesDM packages offers hierarchical Bayesian analysis of computational models on an array of decision-making tasks including Probabilistic Reversal Learning, Orthogonalized Go/NoGo, Two-Armed Bandit, Iowa Gambling, Risk Aversion, Delay Discounting tasks, and more to be added in the future. Users can perform hierarchical Bayesian modeling, output visualization, and Bayesian model comparisons (LOOIC and WAIC) – each with a single line of coding. Its tutorials are also available here: http://u.osu.edu/ccsl/codedata/hbayesdm/.
>
> Thanks for spreading the word and please feel free to contact me if you have any questions.
>
> Best,
> Young
>
>
> Woo-Young Ahn
>
> Assistant Professor
>
> Department of Psychology
>
> Affiliated Faculty, Translational Data Analytics
>
> Ohio State University
>
> 1835 Neil Avenue
>
> Columbus, OH 43210-1222
>
> Email: ahn...@osu.edu
>
> Office: (614) 247-7670
>
> Lab: u.osu.edu/ccsl
>
> Personal: www.ahnlab.org
>
>
>
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "Stan users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-users+...@googlegroups.com.
> To post to this group, send email to stan-...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
RStan, etc., like this one from Young Ahn for doing
decision analysis.
Do you think I should add a spot on the web page to
list these. And if so, where? If I put them under
Citations, nobody will find them. I could create new
pages linked off of one of the interface pages, like
RStan (but this particular package also uses Loo).
I don't want to give people the impression that we're
vouching for the external packages somehow; but it'd
be nice to list the ones that are there.
- Bob
> On Mar 27, 2016, at 6:20 PM, Ben Goodrich <goodri...@gmail.com> wrote:
>
> Good to see. For any R package like this offering a fixed set of Stan programs, you would be better off going the rstanarm route and compiling them at or before installation time rather than at runtime. There is a rstan.package.skeleton() function in rstan that facilitates doing it this way. Or if you are going to IMPS, I could help you convert it then.
>
> Ben
>
> On Sunday, March 27, 2016 at 5:52:53 PM UTC-4, Young Ahn wrote:
> Hi all,
>
> Just want to announce that my lab has released an R package called hBayesDM (hierarchical Bayesian modeling of Decision-Making tasks), which is based on RStan. It is now available on CRAN (https://cran.rstudio.com/web/packages/hBayesDM/).
>
> The hBayesDM packages offers hierarchical Bayesian analysis of computational models on an array of decision-making tasks including Probabilistic Reversal Learning, Orthogonalized Go/NoGo, Two-Armed Bandit, Iowa Gambling, Risk Aversion, Delay Discounting tasks, and more to be added in the future. Users can perform hierarchical Bayesian modeling, output visualization, and Bayesian model comparisons (LOOIC and WAIC) – each with a single line of coding. Its tutorials are also available here: http://u.osu.edu/ccsl/codedata/hbayesdm/.
>
> Thanks for spreading the word and please feel free to contact me if you have any questions.
>
> Best,
> Young
>
>
> Woo-Young Ahn
>
> Assistant Professor
>
> Department of Psychology
>
> Affiliated Faculty, Translational Data Analytics
>
> Ohio State University
>
> 1835 Neil Avenue
>
> Columbus, OH 43210-1222
>
> Email: ahn...@osu.edu
>
> Office: (614) 247-7670
>
> Lab: u.osu.edu/ccsl
>
> Personal: www.ahnlab.org
>
>
>
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "Stan users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-users+...@googlegroups.com.
> To post to this group, send email to stan-...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
Ben Goodrich
Mar 28, 2016, 12:22:50 AM3/28/16
to stan development mailing list
On Sunday, March 27, 2016 at 10:37:59 PM UTC-4, Bob Carpenter wrote:
Do you think I should add a spot on the web page to
list these. And if so, where?
They are listed at the bottom of
https://cran.r-project.org/web/packages/rstan/index.html
which is where anyone would go to if they were looking for R packages that are related to RStan.
Perhaps if we put something similar on mc-stan.org then people will find them who didn't know they were looking for R packages that are related to RStan.
I don't want to give people the impression that we're
vouching for the external packages somehow; but it'd
be nice to list the ones that are there.
Considering we don't even officially vouch for the parts of Stan we maintain, I don't think it is a big deal. If someone is executing binaries generated by Stan, then we have a shared interest in making them thrive.
Ben
> On Mar 27, 2016, at 6:20 PM, Ben Goodrich <goodri...@gmail.com> wrote:
>
> Good to see. For any R package like this offering a fixed set of Stan programs, you would be better off going the rstanarm route and compiling them at or before installation time rather than at runtime. There is a rstan.package.skeleton() function in rstan that facilitates doing it this way. Or if you are going to IMPS, I could help you convert it then.
>
> Ben
>
> On Sunday, March 27, 2016 at 5:52:53 PM UTC-4, Young Ahn wrote:
> Hi all,
>
> Just want to announce that my lab has released an R package called hBayesDM (hierarchical Bayesian modeling of Decision-Making tasks), which is based on RStan. It is now available on CRAN (https://cran.rstudio.com/web/packages/hBayesDM/).
>
> The hBayesDM packages offers hierarchical Bayesian analysis of computational models on an array of decision-making tasks including Probabilistic Reversal Learning, Orthogonalized Go/NoGo, Two-Armed Bandit, Iowa Gambling, Risk Aversion, Delay Discounting tasks, and more to be added in the future. Users can perform hierarchical Bayesian modeling, output visualization, and Bayesian model comparisons (LOOIC and WAIC) – each with a single line of coding. Its tutorials are also available here: http://u.osu.edu/ccsl/codedata/hbayesdm/.
>
> Thanks for spreading the word and please feel free to contact me if you have any questions.
>
> Best,
> Young
>
>
> Woo-Young Ahn
>
> Assistant Professor
>
> Department of Psychology
>
> Affiliated Faculty, Translational Data Analytics
>
> Ohio State University
>
> 1835 Neil Avenue
>
> Columbus, OH 43210-1222
>
> Email: ahn...@osu.edu
>
> Office: (614) 247-7670
>
> Lab: u.osu.edu/ccsl
>
> Personal: www.ahnlab.org
>
>
>
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "Stan users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-users+unsubscribe@googlegroups.com.
Jonah Gabry
Mar 28, 2016, 11:56:50 AM3/28/16
to stan development mailing list
What about adding a Links page to the website where can put links to Stan-related packages, blog-posts, and other stuff that doesn't fit naturally into any of the other pages?
Avraham Adler
Mar 28, 2016, 11:57:18 AM3/28/16
to stan development mailing list
On Sunday, March 27, 2016 at 10:37:59 PM UTC-4, Bob Carpenter wrote:
> I imagine we're going to see more packages based on
> RStan, etc., like this one from Young Ahn for doing
> decision analysis.
>
> Do you think I should add a spot on the web page to
> list these. And if so, where? If I put them under
> Citations, nobody will find them. I could create new
> pages linked off of one of the interface pages, like
> RStan (but this particular package also uses Loo).
>
> I don't want to give people the impression that we're
> vouching for the external packages somehow; but it'd
> be nice to list the ones that are there.
>
> - Bob
> RStan, etc., like this one from Young Ahn for doing
> decision analysis.
>
> Do you think I should add a spot on the web page to
> list these. And if so, where? If I put them under
> Citations, nobody will find them. I could create new
> pages linked off of one of the interface pages, like
> RStan (but this particular package also uses Loo).
>
> I don't want to give people the impression that we're
> vouching for the external packages somehow; but it'd
> be nice to list the ones that are there.
>
> - Bob
If many of the packages will end up on CRAN, perhaps we can petition CRAN about having a "task view" type page such as <https://cran.r-project.org/web/views/NumericalMathematics.html> or <https://cran.r-project.org/web/views/HighPerformanceComputing.html>?
Avi
Bob Carpenter
Mar 28, 2016, 12:39:00 PM3/28/16
to stan...@googlegroups.com
I think there may be packages in other languages besides R.
The view on CRAN doesn't tell you anything about the packages
without following links to them.
The downside of having a list ourselves is that we have to
maintain it.
In terms of vouching for, I just meant that we actively support
our packages on stan-dev and at least at the math and Stan level,
try to actively test them. And we use them ourselves. External
packages won't have any of this.
Maybe we should just announce ones that look interesting on
Andrew's blog.
- Bob
> > To unsubscribe from this group and stop receiving emails from it, send an email to stan-users+...@googlegroups.com.
> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.
The view on CRAN doesn't tell you anything about the packages
without following links to them.
The downside of having a list ourselves is that we have to
maintain it.
In terms of vouching for, I just meant that we actively support
our packages on stan-dev and at least at the math and Stan level,
try to actively test them. And we use them ourselves. External
packages won't have any of this.
Maybe we should just announce ones that look interesting on
Andrew's blog.
- Bob
> > To post to this group, send email to stan-...@googlegroups.com.
> > For more options, visit https://groups.google.com/d/optout.
>
>
> --
> > For more options, visit https://groups.google.com/d/optout.
>
>
> You received this message because you are subscribed to the Google Groups "stan development mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-dev+u...@googlegroups.com.
Avraham Adler
Mar 28, 2016, 1:21:02 PM3/28/16
to stan development mailing list
On Monday, March 28, 2016 at 12:39:00 PM UTC-4, Bob Carpenter wrote:
> I think there may be packages in other languages besides R.
>
> I think there may be packages in other languages besides R.
>
> In terms of vouching for, I just meant that we actively support
> our packages on stan-dev and at least at the math and Stan level,
> try to actively test them. And we use them ourselves. External
> packages won't have any of this.
>
> Maybe we should just announce ones that look interesting on
> Andrew's blog.
>
> - Bob
> our packages on stan-dev and at least at the math and Stan level,
> try to actively test them. And we use them ourselves. External
> packages won't have any of this.
>
> Maybe we should just announce ones that look interesting on
> Andrew's blog.
>
> - Bob
As Ben said, any R package that explicitly Imports, Links, or Depends on rstan will be listed in the reverse sections at the bottom of the CRAN page for the package. Otherwise, it sounds like the easiest option is to have a page on the wiki, divided by Purpose & Language (let y'all decide which goes first).
This way, package/code creators can either e-mail you, create issues, pull request, or straight edit themselves, and you can decide which to feature on the e-mail list or Andrew's blog. It does mean that someone has to keep an eye on the wiki, but that is the case anyway, no?
Avi
Reply all
Reply to author
Forward
0 new messages