To include hook competition

andrew-edwards commented 5 years ago

So basically this involves just scaling all the raw catch rates by a set-specific number. I'll do those and save the .rds results in a new folder gfsynopsis/data-cache/iphc-competition/. Everything should be the same structure, just the numbers will be different.

@seananderson - with two sets of results will be it 'easy' to show two time series, like you do for the other surveys? I think it will be useful to see if/how hook competition matters. I'll end up with results in the same format.

Steps for Andy:

[x] lunch (trying to stay sane!)
[x] check can get hooks with bait easily returned for all the different IPHC data types (prob have to add to my IPHC spp names).
[ ] for now just save once at the set level, then other species can easily use it
[ ] explore how the adjustment factor varies between years (maybe add histograms into write up)
[ ] build gfsynopsis using competition results (instead of default), to check nothing breaks. Create a competition archive also (of figures and data) and a switch as to whether to include competition or not
[ ] look into showing both, as per my question above
[ ] send any files to Sean since he's at home

andrew-edwards commented 5 years ago

@elisekeppel - for the GFbio data I can't just use

x <- get_iphc_hooks("bait on hook")

as that's not a real species. Wondering if we have to just create a new .sql from get-iphc-hook-level.sql. I'll first check those other queries you made recently.

andrew-edwards commented 5 years ago

@seananderson - for the hook competition results I was going to create all the figures in a compeition/ folders, but looking at make.R in gfsynopsis it will be simpler (hmm....) just to overlay the results on the existing ones in the IPHC plot. Can I easily see how to do that - could you just point me to the code where you did it for your surveys. Thanks. I haven't done the calculations yet but have the necessary data extracted, just have to scale the catch rates.

seananderson commented 5 years ago

Oh, it's super ugly because I'm just trying to get this thing done! https://github.com/pbs-assess/gfsynopsis/blob/master/R/make-pages.R#L348-L373

andrew-edwards commented 5 years ago

Won't get done in time for working paper. Will discuss it and show some data by year, but the calculations will get fiddly. A lot more hook-with-baits are zero, which I hadn't anticipated (and which would blow up the competition adjustment factor).

andrew-edwards commented 3 years ago

We included some data on this in Anderson et al. (2019). Is now a full-on project with Joe Watson. Much more than an Issue.

pbs-assess / gfiphc

To include hook competition #2