Issue when simulating paa and age-specific selectivity == 0

[GiancarloMCorrea]

Let's assume we have age-specific selectivity for a given fleet, and selectivity is = 0 for ages 1 and 2. The set_osa_obs.R script will remove the paa observations for those ages in the 'obs' data frame.

Now, if we want to simulate paa data for that fleet, an issue arises here (same for surveys). The length of the vector 't_pred_paa' is = n_ages, but the length of the vector 'ages_obs_y' is = n_ages - 2 (ages 1 and 2 were removed). Therefore, these lines will be problematic: n_ages simulated values will be produced and sorted in a vector whose length = n_ages - 2.

I fixed this issue here, but want to confirm if this would be the best way to do it.

[timjmiller]

Thanks Giancarlo. Did you get this error using multionial, Dirichlet-multinomial, or MVTweedie? It looks like it might only be an issue with these likelihoods. The other likelihoods have an "ages" argument that can reduce the size of the returned observation vector.

…

On Mon, May 15, 2023 at 12:16 PM Giancarlo M. Correa < ***@***.***> wrote: Let's assume we have age-specific selectivity for a given fleet, and selectivity is = 0 for ages 1 and 2. The set_osa_obs.R script will remove the paa observations for those ages in the 'obs' data frame. Now, if we want to simulate paa data for that fleet, an issue arises here <https://github.com/timjmiller/wham/blob/de2e89fd329e4dc1c5e83c8ab99294ce01ee0ef6/src/wham_v0.cpp#LL1044C20-L1044C20> (same for surveys). The length of the vector 't_pred_paa' is = n_ages, but the length of the vector 'ages_obs_y' is = n_ages - 2 (ages 1 and 2 were removed). Therefore, these lines <https://github.com/timjmiller/wham/blob/de2e89fd329e4dc1c5e83c8ab99294ce01ee0ef6/src/age_comp_sim.hpp#L313C11-L315> will be problematic: n_ages simulated values will be produced and sorted in a vector whose length = n_ages - 2. I fixed this issue here <https://github.com/GiancarloMCorrea/wham/blob/74f4a007d8656ec0fe74fd85598ad3f3c1a7808a/src/age_comp_sim.hpp#L330-L395>, but want to confirm if this would be the best way to do it. — Reply to this email directly, view it on GitHub <#80>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEIGN7HP6QOUM44YFGZ5UY3XGJJFJANCNFSM6AAAAAAYCNRHLY> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Timothy J. Miller, PhD (he, him, his) Research Fishery Biologist NOAA, Northeast Fisheries Science Center Woods Hole, MA 508-495-2365

[GiancarloMCorrea]

Hi Tim,

I used multinom. I agree, that issue will happen only with those options.

[timjmiller]

I have verified that it is also an issue for evaluating the likelihood when fitting this configuration. Testing a fix now.

…

On Mon, May 22, 2023 at 11:46 AM Giancarlo M. Correa < ***@***.***> wrote: Hi Tim, I used multinom. I agree, that issue will happen only with those options. — Reply to this email directly, view it on GitHub <#80 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEIGN7HVFZW33AYNNITZAA3XHOC6XANCNFSM6AAAAAAYCNRHLY> . You are receiving this because you commented.Message ID: ***@***.***>

-- Timothy J. Miller, PhD (he, him, his) Research Fishery Biologist NOAA, Northeast Fisheries Science Center Woods Hole, MA 508-495-2365

[Cole-Monnahan-NOAA]

When I do this the data are always. I've only experimented with this setup using the multinomial but it seemed to estimate fine.

[timjmiller]

This is now corrected on devel branch.