stan-dev / rstan

RStan, the R interface to Stan
https://mc-stan.org
1.04k stars 268 forks source link

print treepepth, etc. in output of print() #45

Open bob-carpenter opened 10 years ago

bob-carpenter commented 10 years ago

We want to include the same diagnostic output "parameters" as in CmdStan, which means adding

So that means adding:

CmdStan's output looks like:

                 Mean     MCSE   StdDev        5%   50%   95%  N_Eff  N_Eff/s  R_hat
lp__             -7.3  3.5e-02  6.9e-01  -8.7e+00  -7.0  -6.7    390    34020   1.00
accept_stat__    0.64  1.2e-02  3.6e-01   5.1e-03  0.74   1.0    882    76898   1.00
stepsize__        1.8  7.8e-15  5.6e-15   1.8e+00   1.8   1.8   0.50       44   1.00
treedepth__     0.076  8.6e-03  2.7e-01   0.0e+00  0.00   1.0    942    82167   1.00
n_leapfrog__     2.7  4.9e-02  1.3e+00        1.0   3.0   3.0    716    65090  1.0e+00
n_divergent__    0.00  0.0e+00  0.0e+00   0.0e+00  0.00  0.00   1000    90909   1.00
theta            0.25  4.2e-03  1.2e-01   9.0e-02  0.23  0.47    827    72146   1.00
jgabry commented 8 years ago

@bob-carpenter I've been closing outdated RStan issues but not sure if this old one should be closed. Do you still think we should be printing the sampler parameters in the output of print()?

bob-carpenter commented 8 years ago

Up to you and Ben. It'd be nice if they could be extracted in the same way, but I also see the point of keeping them separate as they're not parameters per se.

betanalpha commented 8 years ago

I think we need at least an equivalent to extract. Right now you have to go through waaaaaaay too many hoops to get at the sampler parameters.

On Aug 7, 2016, at 1:50 AM, Bob Carpenter notifications@github.com wrote:

Up to you and Ben. It'd be nice if they could be extracted in the same way, but I also see the point of keeping them separate as they're not parameters per se.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

bob-carpenter commented 8 years ago

Yes, please to something that works just like extract to pull out sampler params. Even better if I can just use extract to extract them.

I can never remember how to pull them out so I have Jonah's instructions pinned to my desktop.

And I would very much like at least an option to see them in the printed output by default along with other diagnostics. Then we'll tell every user to turn on that output if they have a question.

On Aug 7, 2016, at 4:48 AM, Michael Betancourt notifications@github.com wrote:

I think we need at least an equivalent to extract. Right now you have to go through waaaaaaay too many hoops to get at the sampler parameters.

On Aug 7, 2016, at 1:50 AM, Bob Carpenter notifications@github.com wrote:

Up to you and Ben. It'd be nice if they could be extracted in the same way, but I also see the point of keeping them separate as they're not parameters per se.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.