search

brianlangseth-NOAA / Spatial-Workshop-SPASAM

Spatially stratified simulation-estimation framework incorporating multistage stock-recruit relationships and incorporating larval IBM outputs
1 stars 0 forks source link

TO DO: Diagnostics #20

Closed brianlangseth-NOAA closed 2 years ago

brianlangseth-NOAA commented 2 years ago

For testing purposes develop a set of diagnostics for testing model fit. Similar to SIM_TIM.R but geared towards fitting observed vs. estimated as opposed to EM versus True values. It shouldn't take too much to switch over from TIM_Diagnostics_master.R

JDeroba commented 2 years ago

@brianlangseth-NOAA @AmySchueller-NOAA Just added YFT_Diagnostics.R to the main directory. It creates Model_Diagnostics.pdf in the user specified EM folder. The code still needs work to handle multiple fleets and surveys, but it sure beats Excel. I already find this useful; e.g., initial abundance (most likely recruitment) is way off, one of the fleets has zero observed or estimated catches so something is way off, selectivity is way off...

brianlangseth-NOAA commented 2 years ago

Awesone @JDeroba. BTW, using @ work really well, always go to emails, though Im still unsure whether typical comments result in an email or require someone having to check (likely depending on individuals "watch" status). Regardless, I will try the diagnostics on the new data. Just commited data_loading.R changes with cleaned up script for the newest data, as well as the .dat file for YFT_1area in commit d205729

brianlangseth-NOAA commented 2 years ago

@JDeroba updated diagnostics to be fleet specific with commit 12ff732