Package of 'doRLP'

Change Log:

• 5th Oct 2013. Upload the package

Download and Install:
[source code] doRLP_1.0.tar.gz

• Install the 'tar.gz' ball by R command such as

   
  command line> R CMD INSTALL glm.tree_1.0.tar.gz
  

Required additional package: snow, snowfall, ( sstools )

Main functions

• get.data ; conduct simulation of age-structured population dynamics and get fisheries data
• get.ssfile ; create data file of Stock Synthesis (v2.23b)
• do.all ; Wrapper function above the two functions and do estimation & likelihood profile by each replication. Need additional files for configuration for SS estimation and other miscellaneous files.

Procedures

• Install 'wine' if conducting the simulation in Linux platform.
• Required packages, snow, snowfall (if using multi-core) and sstools (can be downloaded from here ), and this package
• Create empty folder (e.g. 'white') for simulation
• Move to the 'white' folder, and create new folder named by 'orgrun_tmp'
• Put needed files under 'orgrun_tmp'. An example folder can be downloaded from here, except for executable file for SS. Executable file for SS can be downloaded from the official HP.
  • control file for SS named by "BLUE.CTL" for normal estimation, and "BLUE_LP.CTL" for likelihood profile. The names of the control files should be as the name.
  • starter.ss and other miscellaneous files
  • profilevalues.ss ; specifying fixed values when conducting likelihood profiles
  • executable file for SS (ss3.exe)

# "do.all" does all procedures of create data, write dat file for SS, condurct SS and likelihood profiles.
tmpres <- do.all(nsim=5, # number of replications
                 Rsigma=0.6, # CV of recruitment deviations assumed int the true population
                 cpue.error=rep(0.2,2), # CV of observation errors in abunance indices for each fishery (in the true population)
                 obs.cpue=rep(0.2,2), # CV of observation errors inputted to SS dat file (in SS)
                 Nlen=rep(400,2), # Number of sample size for multinomial observational errors in the length composition data (in the true population)
	         Nlen.ss=rep(400,2), # Number of sample size inputted to SS dat file (in SS)
                 Fall=0.1, # Apex fishing mortality for the 1st fleet
                 mc=2, # number of threds used for the calculations
                 is.likeprof=T) # if TRUE, do likelihood profiles
# some arguments are shared with the arguments specified to get.data. 
# time-varying assumptions are controlled by options of get.data