Skip to contents

Calculate variance adjustments for discard or mean body weight data

Source: R/helper_fxns.R
calc_var_adjust.Rd

Function developed for U.S. west coast Sablefish assessment in 2019 to tune discard data or mean body weight data which are common inputs for U.S. west coast groundfish assessments but as of 2023 have not often had any data weighting method applied to them.

Usage

calc_var_adjust(data, type = c("CV", "sd"))

Arguments

data

Either the "discard" or "mnwgt" elements of the list returned by SS_output(). Other data types might work here but haven't been tested.

type

Either "CV" or "sd" specifying the type of control file variance adjustment, where the SS3 options are 2=add_to_discard_stddev`` and 3=add_to_bodywt_CV, so if datais discard data, type should be "CV" and ifdata` is mean body weight, type should be "sd".

Value

A table of input and estimated uncertainty values in units of both CV and sd including the following:

  • fleet is the fleet number

  • mean_out is the mean of the expected values

  • mean_in is the mean of the observed values

  • CV_in is the mean input CV

  • sd_in is the mean input SD values (which may include variance adjustments already)

  • sd_out is the SD of the observed relative to the expected values, calculated as described above

  • CV_out is the CV of the observed relative to the expected, calculated as described above

  • added is the value that could be added to any existing value in the "Input variance adjustments factors" section of the control file.

  • type is the data type code used in "Input variance adjustments factors"

Details

The calculation is based on sd_out = sqrt(mean(Obs - Exp)^2)). Added sd is calculated as sd_out - sd_in where sd_in is the mean of the input standard deviations (possibly including existing variance adjustments). When a CV adjustment is required, the sd_out is converted to CV_out by dividing by the mean of the expected values and with the added CV calculated as CV_out - CV_in.

Author

Kelli F. Johnson