Implement processing of all-age data

OSOceanAcoustics / echopop

Generating biological estimates for animal "pop"ulation from echosounder data

https://echopop.readthedocs.io/

Apache License 2.0

1 stars 4 forks source link

Implement processing of all-age data #129

Closed leewujung closed 5 months ago

leewujung commented 11 months ago

currently only the option exclude_age1=True is implemented
did Chu implement the option to include all ages?
consider changing the option name to allow selecting adult and/or age-1

See #91 for some previous info from @b-reyes.

@emiliom : Do you know if all-age processing was implemented in Matlab? (this is why I "assigned" this to you, not to ask you to implement the functions).

rebeccathomas-NOAA commented 11 months ago

I know we absolutely need processing of age-1 data, but I'm not exactly sure what kind of processing Chu was doing with the age-1 in his latest version. I know his code produces different numbers with age-1 =TRUE and not true. I also know Chu was working on a few different implementations of age-1, and I'm not sure what is in his latest code.

leewujung commented 9 months ago

Notes from 2024/01/29 meeting:

Age binning:

include age-1 in strata, but kick them out before kriging
should not include age-0
current estimate is age-2 only, trying to move to include age-1
age-1 has very different net selectivity

Temp decision:

age-0 and age-1 should be in their own bins
so have age-0 / age-1 / age-2+
But what do we include for the total number of fish? TBD

@brandynlucca -- could you comment on what the implemented handling is at the moment and point to the code section? to help us get back to this context quickly when we circle back to this, in the hopefully not very distant future!

leewujung commented 9 months ago

I accidentally closed the wrong issue! Re-opening this.

brandynlucca commented 5 months ago

The argument exclude_age1 has been added to transect_analysis(...) that selects the correct NASC data source depending on if exclude_age1 = True (i.e. adult-only while accounting for leaked age-1 contributions) and exclude_age1 = False (i.e. everything gets accounted for). This argument then propagates into other parts of the analysis without requiring any additional input (i.e. stratified_analysis(...) and kriging_analysis(...)).

The apportionment step is independent of this decision and is determined based on the original biological data. The primary change induced by exclude_age1 is 1) which NASC data to use and 2) whether kriged population estimates (e.g. biomass) apportioned by age and length need to be further reapportioned to remove contributions of age-1 fish.

This can be closed once PR #241 is merged.