pacific-hake / hake-assessment

:zap: :fish: Build the assessment document using latex and knitr
MIT License
13 stars 6 forks source link

Bridging model final relative biomass plot does not agree with base model plot #1105

Closed andrew-edwards closed 10 months ago

andrew-edwards commented 10 months ago

Figure 16 we have image

for which green is final bridging step (use modeled spatio-temporal maturity). But Figure 27 for base model we have image

showing the stock being healthier all the way through (mostly way above B40). The bridging plot closely matches last year's plot (none of the bridging steps show a large change in relative biomass trajectory).

Am I missing something obvious?

andrew-edwards commented 10 months ago

Maybe someone first check a build on the server, though I downloaded all model results yesterday after Chris re-uploaded them, and we checked they were the same. So I don't think it's that.

aaronmberger-nwfsc commented 10 months ago

There is clearly an error in the rel spawning biomass figure with scale. This may have arisen with recent edits to that plot when trying to add the dynamic B0 stuff (???).

The bridging figure looks correct.

andrew-edwards commented 10 months ago

Yes, that's it. I have Kelli's version she saved the other day and the relative figure looks correct. Phew.

andrew-edwards commented 10 months ago

Was comparing with last year's figure and wondered why it looked so different! Thanks. Chris can presumably easily fix it - I want to do some writing (and I'd get bogged down in debugging if I start digging into it!).

andrew-edwards commented 10 months ago

Think it's actually a bit more than that, as curr_depl_median has the value 143%. The other day it was 80%. So likely not the figure change. And the numbers are wrong in the text pages I printed out yesterday, and that was before the new figure was born.

cgrandin commented 10 months ago

Looks to be fixed as of 67486bbc9fa7e0b5072dd86102a0104efc8bf1a1.

Figure 27 was the incorrect one, they now both match and curr_depl_median is 76% which makes sense given that the model changed a bit