Table 7 (Posterior and MPD estimates of spawning biomass...) about csas-latex HOT 17 CLOSED

Thanks Chris!

from csas-latex.

I'm working on this

from csas-latex.

I have updated my figures and captions to show 90% CIs -- those are both set by the variable ciLevel in "ResearchDocumentV2.R".
I tried to update the CIs for the tables in the csas-latex template part by updating the variable mcmc.ci in "custom-knitr-variables.r", and confidence.vals in "model-setup.r". This didn't seem to do as I expected. Or at least, the column headers in the tables were not updated. But I see now that you're working on this, so maybe you've already updated your CIs?

from csas-latex.

I've also updated my code to use 90% CIs. That is set in model-setup.r confidence.vals <- c(0.05, 0.95) The amount of things that I'm changing means you need to delete.rdata.files();build() each time you fetch my stuff and that's when those calculations are done.

from csas-latex.

We are not outputting sbo in the mcmc files. So we'd have to do that and re-run all models -mceval step.
So I'll go back to iscam-pbs and see if I can add that.

from csas-latex.

I added sbo to iscam-pbs mcmc outputs. You can merge it and try it out. I am seeing the same values though
i.e. bo = sbo
so that's not the issue.

from csas-latex.

My bad! I was dividing bt/bo for the MPD depletion instead of sbt/sbo. Fixed it in commit b2a425c. So I'm going to revert iscam-pbs and you don't need to do anything there. Whew, we don't need to re-run any models again.

from csas-latex.

great!

b0 and sb0 in the rep file and mcmc outputs should be different values though because sbo is calculated using longterm average weight-age and maturity-age, whereas bo uses the 'years' specified in the pfc file - which for herring is the 5-year recent average weight-age and maturity-age.
Can you confirm this is what you are seeing in the outputs?

from csas-latex.

Yes, they are different in the REP file. sbo seems to be roughly 3 times bo for HG AM2 for example. I have divided by bo for the MPD and the outputs match the mcmc. The mcmc calculation is sbt/bo because I saw that sbo = bo in the mcmc output. Perhaps I didn't run it out long enough to see a change?

I have to head out now but will investigate further in the morning.

from csas-latex.

Am I correct that I need to "decide" whether I want to include a plot or table of SB0 values, and that if I do, that we need to re-run the mcmc's? (not a desirable option...)

from csas-latex.

To do that you would just need to re-run the mceval step for each model, so it wouldn't take too long.

from csas-latex.

I just ran out the mcmc to 50,000 and bo=sbo exactly for every sample. So we need to figure out asap how to fix this.

from csas-latex.

Sarah- Chris and I realized that the sbo and bo values captured in the mcmc file ARE the same because the bo comes from the slowMSY calc (Robyn's code) so they are meant to match. Chris is going to remove the bo output from the mcmc file so that this doesn't confuse us next year.

from csas-latex.

Should I rerun the mceval when this is updated?

from csas-latex.

Yes, because the output will be called sbo instead of bo. I'll let you know in a bit

from csas-latex.

Or you could just go through and manually change the column name "bo" in each iscam_mcmc.csv file to "sbo", whichever is easier. I think the latter..

from csas-latex.

I have modified iscam-pbs to output sbo in the mcmc, so I'll close this. I will still need to modify the csas-latex code to make use of the new column name "sbo".

from csas-latex.