Expected profit model with normal catch function

Calculate the negative log-likelihood of the expected profit model (EPM) with a normal catch function. For more information on the EPM normal model see section 8.4.3 in the FishSET user manual. https://docs.google.com/document/d/1p8mK65uG8yp-HbzCeBgtO0q6DSpKV1Zyk_ucNskt5ug/edit#heading=h.mrt9b1ee2yb8

Usage

epm_normal(starts3, dat, otherdat, alts, project, expname, mod.name)

starts3: Starting values as a numeric vector. The order of parameters in the vector is:
c([catch-function params], [travel-dist params], [stdev], [common scale param]),
where the length of catch-function parameters is the # of alternatives * # of catch-function variables, length of travel-distance parameters is the # of travel-distance variables, length of standard deviation defaults to 1 but alternative- specific standard deviation values can be specified (length = # of alternatives), and the common scale parameter is a single value.
dat: Data matrix, see output from shift_sort_x, alternatives with distance.
otherdat: List that contains other data used in the model, see section 8.4.3 in the FishSET user manual for more details (link in the description above): (1) 'griddat': catch-function variables that interact with alternative-specific catch-function parameters and do not vary across alternatives (e.g., vessel gross tonnage). (2) 'intdat': travel-distance variables that interact with travel-distance parameters and the distance matrix and do not vary across alternatives. (3) 'prices': price in terms of $/landings units. This is typically a vector with prices for each observation, but can be a single value representing price for the entire dataset.
alts: Number of alternative choices in model
project: Name of project
expname: Expected catch table (optional)
mod.name: Name of model run for model result output table

ld: negative log likelihood

This function is called in discretefish_subroutine when running an EPM model with a normal catch function.