Estimation

Using LIME for estimation requires three vital steps:

1. Setup data correctly
2. Run the LIME model using TMB
3. Check convergence and plot output
4. Run sensitivities

Data setup

It is very important that the data are set up correctly for the LIME estimation to run.

Length data

A matrix with the time step (e.g. year) along the rows and length bins along the columns.

• Rows should be named with the year names. Only years with length data should be included. Years can be in true years (1998-2017) or indexed years (1-20) as long as there is consistency between items in the data list (see "Input List" below).
• Columns should be named with the upper length bins. Columns should range from the first possible length bin (e.g. 1 for 1-cm length bins) to the last observed length bin. If there are no observations until 30cm, add a matrix of zeros for lengths 1,2,3,...29cm. If the last observed length bin is less than 1.5 times the asymptotic length (linf), LIME will internally add a matrix of zeros to be able to account for potentially larger fish than were observed.

## years with length data -- rename with your own years with length data
length_years <- rownames(true$LF)

## length bins -- rename with your upper length bins
length_bins <- colnames(true$LF)

## matrix with length frequency data
LF <- true$LF

## set rownames as years
rownames(LF) <- length_years

## set colnames as length bins
colnames(LF) <- length_bins

## first six years of length data
head(LF)