Forums

Back

forecast issues with recent SS executables

RSS (Opens New Window)(Opens New Window)
YL
Yevhen Leonchyk, modified 5 Years ago.

forecast issues with recent SS executables

Youngling Posts: 36 Join Date: 8/1/17 Recent Posts

I'm trying to split this thread (which long ago moved away from the stated subject) but not doing it correctly.

The message below is the second indication that there is some issue with forecasting in some of bug-fix 3.30.13.xx executables. Rick has made some fixes to a few forecast issues in 3.30.13.07 posted yesterday to the "Fixes to 3.30.13" folder, so it's probably a good idea to start there and see if that sorts out the problem. If not, we can use this new thread to track remaining problems.

Hi, Chantel

It does not depend on r4ss. I use "forecast=T" in both cases. It depends on ss.exe version. If I use the latest fixed ss3, I have not got forecasting. But if I use the last stable ss3, I have got one with the same r4ss. I consider the fixed ss3 has a problem.

IT
Ian Taylor, modified 6 Years ago.

RE: forecast issues with recent SS executables

Youngling Posts: 117 Join Date: 12/8/14 Recent Posts

Sorry, that last message was from me (Ian Taylor). Editing the post from Yevhen Leonchyk to split the thread and change the subject line made it look like my reply was from him.

CW
Chantel Wetzel, modified 6 Years ago.

RE: forecast issues with recent SS executables

Youngling Posts: 32 Join Date: 10/7/15 Recent Posts
Actually, this is not an ss executable issue.  Looking at the Report file across model versions I see that the forecast quantities have not been printed to the Derived Quantities section due to a specification in the starter file.  However, the forecast quantities have been calculated and printed in the TIME_SERIES section.  If you change this line in your starter file, you will always get the forecast years in the Derived Quantities section: 

-1   #_min yr for sdreport outputs (-1 for styr)
-2   #_max yr for sdreport outputs (-1 for endyr; -2 for endyr+Nforecastyrs)

In the plots, r4ss uses the time-series section from the Report file so these values can be plotted but only if you set the forecastplot = TRUE in the SS_plot function.

Using the code I sent I was able to create the time-series plots with and without the forecast period using your model files and the corrected executable.

Let me know if this doesn't work for you.

On Fri, Jun 14, 2019 at 10:49 AM Yevhen Leonchyk <VLab.Notifications@noaa.gov> wrote:

Hi, Chantel

It does not depend on r4ss. I use "forecast=T" in both cases. It depends on ss.exe version. If I use the latest fixed ss3, I have not got forecasting. But if I use the last stable ss3, I have got one with the same r4ss. I consider the fixed ss3 has a problem.


--
Yevhen Leonchyk Stock Synthesis Virtual Lab Forum https://vlab.noaa.gov/group/stock-synthesis/discussions-forums-/-/message_boards/view_message/6852011 VLab.Notifications@noaa.gov


--
--------------------------------------------------------------------
Chantel Wetzel, PhD
Fishery Resource Analysis and Monitoring Division
Northwest Fisheries Science Center
National Marine Fisheries Service
National Oceanic and Atmospheric Administration
Phone: 206-302-1753
YL
Yevhen Leonchyk, modified 6 Years ago.

RE: forecast issues with recent SS executables

Youngling Posts: 36 Join Date: 8/1/17 Recent Posts

1) The first case:

bug-fix 3.30.13

-1   #_min yr for sdreport outputs (-1 for styr)
-2   #_max yr for sdreport outputs (-1 for endyr; -2 for endyr+Nforecastyrs)

SS_plots(tmp, uncertainty=F, forecastplot = TRUE)

Please, sea fig. 1. There is no forecasing. It is not shown in the figure.

2) The second case:

stable 3.30.13

-1   #_min yr for sdreport outputs (-1 for styr)
-2   #_max yr for sdreport outputs (-1 for endyr; -2 for endyr+Nforecastyrs)

SS_plots(tmp, uncertainty=F, forecastplot = TRUE)

Please, sea fig. 2. There is forecasing for the next years.

 

I use the same setting in both cases changing only ss3.exe versions.

 

Attachments:

Attachment

Attachment

YL
Yevhen Leonchyk, modified 6 Years ago.

RE: forecast issues with recent SS executables

Youngling Posts: 36 Join Date: 8/1/17 Recent Posts

Just in case, here are my files.

Attachments: