RNL_new trans model_with DEB_1.6.2.R

# RNL_new trans model_with DEB_1.6.2

# 30-8-18
# updates
# R 3.5.0
# RStudio 1.1.453
# JGR 3.4.1
# Netlogo 5.3.1
# fixed rJava installation error for Mac OSX El Capitan 10.11.+ 
# fixed JGR installation and loading error from rJava error

#19-6-18
# time stamp for updated file RNL_new trans model_with DEB_1.6.2

# 5-1-17 RNL_new trans model_with DEB_1.6.2
# moved gutfull from 'auxiliary params' to 'update animal and env traits'
# added Random and Clumped shade density input (shadedens) for "Sleepy IBM_v.6.1.1_two strategies_shadedens.nlogo"  
# updated 'Sleepy IBM_v.6.1.1_two strategies.nlogo' NL model so food patch input matches output (Food-patches / 10)
# updated movement strategy input

# 1-12-16 RNL_new trans model_with DEB_1.6.2
# updated zenith angle code

# 30-11-16 RNL_new trans model_with DEB_1.6.1_olve_movement cost
# added new DEB estimations (Mike)
# updated soil moisture input (Mike)
# updated mass input from DEB calcs

#24-10-16 RNL_new trans model_with DEB_1.6.1_olve_movement cost
# added movement cost (loco), spatial movement, spdf, and homerange plot outputs to export results loop
# set movement cost to miniscule amount when not searching to fix feeding behav on green food patches (NLCommand("set Movement-cost", 1e-09) 
# added new direct movement cost (loco)
# changed V_pres to debout[2]

#20-10-16
# updated movement cost loop to reset to 0 in NL when not searching
# added NL_strategy variable to NL setup ... in progress

#28-7-16
# added direct movement cost (loco) estimated from ml 02 g-km (John-Adler 1986) and converted to J 

#16-5-16
# added direct movement cost (NL_move) to DEB loop (if (actstate == "S") {NLCommand("set Movement-cost", NL_move)})
# added direct movement cost initial par values
# updated NL_move var
# uses DEB model that includes direct movement cost (cost of locomotion)  ('DEB_movecost.R')
# changed all instances of Min-T_b and Max-T_b to T_opt_lower and T_opt_upper to relfect changes in Netlogo (v6.1.1_two strategies)
# changed VTMIN and VTMAX to activity range Tb limits

#29-1-16
# created sim count (sc) and function to replace results with most recent results of sim run. use to have access to all previous sim runs. Used for getting mean results from multiple sims.
# added results, spdf, homerange, and NL plots to write results loop 

#14-1-15
# added days to turtle homerange spdf to plot movement paths by day
# added wetmass (V_pres, wetfood, wetgonad, wetstorage) NL plotting function in DEB loop 
# updated DEB loop to run on 2-min time steps
# updated initial debout variables in loop to user friendly values
# added gutfull and gutthresh pars within DEB loop to update feeding behav (decision-making part in NL)

#6-1-15
# defined gutfull and X_food pars prior to sim loop (also fixes error of not recognising gutfull in sim loop)
# added V_pres, wetgonad, wetstorage, and wetfood from debout to results output

#30-12-15 / RNL_new trans model_with DEB_1.6.1_olve
# added plot function to compare gutfill 75% and 100% 
# updated reserve level in loop with 'X_food' and 'actfeed' vars
# updated mass var in NL plot output file name 

#24-12-15 modularised with NicheMapR functions

# 15-6-15
# Added hr<-4*emis*sigma*((Tc+Trad)/2+273)^3 # radiation resistance (Line 270)

#22-5-15
# updated with new DEB model parameters ... still needs Mike to check discrepancies against his params

#11-5-15 (v.1.6)
# Added ctminthresh parameter to keep turtle alive when under VTMIN (12 hours)
# Added plot export function from NL.

#30-4-15
# Updated movement path plot functions to accomodate for new NL spatial plot output in NL model.
 # Movement path plots accept dynamic month input-- sep, nov, dec, etc.

#25-4-15
# Fixed NLGetAgentSet error and homerange plot error (changed 'die' in Soft foraging model (NL))

# 30-3-15
# Added home range polygon plotting function
# Added plotting code for new monthly simulations

#23-3-15
# Added lower and upper activity ranges (NL_T_opt_l, NL_T_opt_u)
# Added procedure for plotting per simulation month

#2-12-14
# new analytical transient model (one lump trans)

# TO DO
# Confirm if new VTMAX of 51 C works for month of November
# Feedcount in NL doesn't match 'NLfoodcount',   ........ under progress
# Need to change q to something other than 0?

# 27-11-14
# WORKING VERSION
# X_food variable changed to X_food<-NLReport("[energy-gain] of turtle 0]")
# DEB loop now runs on one NL hour

# 3-10-14
# Uses simulation part from RNL_new trans model_with DEB.R, but with the following:
#   - NLCommand("set T_b precision", Tb, "2") # Updating Tb
#   - acthr<-NLGetAgentSet("in-food?","turtles", as.data.frame=T) ; acthr<-as.numeric(acthr) # Reports true if turtle is in food. For cum. feeding bouts: acthr<- NLReport("feedcount") (# hours of activity. This would change with input from IBM)
#   - X_food<-NLCommand("ask turtle 0 [set energy-gain Low-food-gain]") # sets energy intake from feeding. (changes with input from IBM)
#   - Reports reserve-level


# 1-10-14
# Now opens NL version 5.1.0
# Removed f==1 from 'Regulates X_food dynamics' loop (Line 1116)

# 16-9-14
 # Tb updated after DEB simulation
 # acthr reports when turtle in food patch
 # Feeding dynamics updates with movement (: 1098)

 #12-9-14
 # NL reserve level updated in NL setup procedure (:634)
 # X_food updates with NL "[handle-food]"" procedure  (:695)

# ---------------------------------------------------------------------------
# ------------------- initial Mac OS and R config ---------------------------
# ---------------------------------------------------------------------------
# if already loaded, uninstall RNetlogo and rJava
p<-c("rJava", "RNetLogo")
remove.packages(p)

# if using Mac OSX El Capitan+ and not already in JQR, download and open JGR 

# for a rJava error, run the following in terminal (src: https://stackoverflow.com/questions/30738974/rjava-load-error-in-rstudio-r-after-upgrading-to-osx-yosemite)
# sudo ln -s $(/usr/libexec/java_home)/jre/lib/server/libjvm.dylib /usr/local/lib
# then install rJava from source
install.packages("rJava", repos = "https://cran.r-project.org/", type="source")
library(rJava)

# load JGR after downloading 
Sys.setenv(NOAWT=1)
install.packages("JGR")
library(JGR)
Sys.unsetenv("NOAWT")
JGR() # open JGR

# ------------------------- JGR onwards ----------------------------
# install RNetlogo from source if haven't already
install.packages("RNetLogo", repos = "https://cran.r-project.org/", type="source")
library(RNetLogo)

# ------------------- for PC and working Mac OSX ---------------------------
# ------------------- model setup ---------------------------
# get packages
packages <- c("NicheMapR","adehabitatHR","rgeos","sp", "maptools", "raster","rworldmap","rgdal","dplyr")
install.packages(packages,dependencies = T)
lapply(packages,library,character.only=T)

#source DEB and heat budget models from https://github.com/darwinanddavis/MalishevBullKearney
source('DEB.R')
source('onelump_varenv.R')

# set dirs
setwd("<your working dir>") # set wd
results.path<- "<dir path to store result outputs>" # set results path

#----------- read in microclimate data ---------------
tzone<-paste("Etc/GMT-",10,sep="")
metout<-read.csv('metout.csv')
soil<-read.csv('soil.csv')
shadmet<-read.csv('shadmet.csv')
shadsoil<-read.csv('shadsoil.csv')
micro_sun_all<-cbind(metout[,2:5],metout[,9],soil[,6],metout[,14:16])
colnames(micro_sun_all)<-c('dates','JULDAY','TIME','TALOC','VLOC','TS','ZEN','SOLR','TSKYC')
micro_shd_all<-cbind(shadmet[,2:5],shadmet[,9],shadsoil[,6],shadmet[,14:16])
colnames(micro_shd_all)<-c('dates','JULDAY','TIME','TALOC','VLOC','TS','ZEN','SOLR','TSKYC')

# choose a day(s) to simulate
daystart<-paste('09/09/05',sep="") # yy/mm/dd
dayfin<-paste('10/12/31',sep="") # yy/mm/dd
micro_sun<-subset(micro_sun_all, format(as.POSIXlt(micro_sun_all$dates), "%y/%m/%d")>=daystart & format(as.POSIXlt(micro_sun_all$dates), "%y/%m/%d")<=dayfin)
micro_shd<-subset(micro_shd_all, format(as.POSIXlt(micro_shd_all$dates), "%y/%m/%d")>=daystart & format(as.POSIXlt(micro_shd_all$dates), "%y/%m/%d")<=dayfin)
days<-as.numeric(as.POSIXlt(dayfin)-as.POSIXlt(daystart))

# create time vectors
time<-seq(0,(days+1)*60*24,60) #60 minute intervals from microclimate output
time<-time[-1]
times2<-seq(0,(days+1)*60*24,2) #two minute intervals for prediction
time<-time*60 # minutes to seconds
times2<-times2*60 # minutes to seconds

# apply interpolation functions
velfun<- approxfun(time, micro_sun[,5], rule = 2)
Zenfun<- approxfun(time, micro_sun[,7], rule = 2)
Qsolfun_sun<- approxfun(time, micro_sun[,8], rule = 2)
Tradfun_sun<- approxfun(time, rowMeans(cbind(micro_sun[,6],micro_sun[,9])), rule = 2)
Tairfun_sun<- approxfun(time, micro_sun[,4], rule = 2)
Qsolfun_shd<- approxfun(time, micro_shd[,8]*.1, rule = 2)
Tradfun_shd<- approxfun(time, rowMeans(cbind(micro_shd[,6],micro_shd[,9])), rule = 2)
Tairfun_shd<- approxfun(time, micro_shd[,4], rule = 2)

# upper and lower activity thermal limits
VTMIN<- 26 
VTMAX<- 35  

# ****************************************** read in DEB parameters *******************************

debpars=as.data.frame(read.csv('DEB_pars_Tiliqua_rugosa.csv',header=FALSE))$V1 # read in DEB pars

# set core parameters
z=debpars[8] # zoom factor (cm)
F_m = 13290 # max spec searching rate (l/h.cm^2)
kap_X=debpars[11] # digestion efficiency of food to reserve (-)
v=debpars[13] # energy conductance (cm/h)
kap=debpars[14] # kappa, fraction of mobilised reserve to growth/maintenance (-)
kap_R=debpars[15] # reproduction efficiency (-)
p_M=debpars[16] # specific somatic maintenance (J/cm3)
k_J=debpars[18] # maturity maint rate coefficient (1/h)
E_G=debpars[19] # specific cost for growth (J/cm3)
E_Hb=debpars[20] # maturity at birth (J)
E_Hp=debpars[21] # maturity at puberty (J)
h_a=debpars[22]*10^-1 # Weibull aging acceleration (1/h^2)
s_G=debpars[23] # Gompertz stress coefficient (-)

# set thermal respose curve paramters
T_REF = debpars[1]-273
TA = debpars[2] # Arrhenius temperature (K)
TAL = debpars[5] # low Arrhenius temperature (K)
TAH = debpars[6] # high Arrhenius temperature (K)
TL = debpars[3] # low temp boundary (K)
TH = debpars[4] # hight temp boundary (K)

# set auxiliary parameters
del_M=debpars[9] # shape coefficient (-) 
E_0=debpars[24] # energy of an egg (J)
mh = 1 # survivorship of hatchling in first year
mu_E = 585000 # molar Gibbs energy (chemical potential) of reserve (J/mol)
E_sm=186.03*6

# set initial state
E_pres_init = (debpars[16]*debpars[8]/debpars[14])/(debpars[13]) # initial reserve
E_m <- E_pres_init
E_H_init = debpars[21] + 5

#### change inital size here by multiplying by < 0.85 ####
V_pres_init = (debpars[26] ^ 3) * 0.85 
d_V<-0.3
mass <- V_pres_init + V_pres_init*E_pres_init/mu_E/d_V*23.9

# ***************** end TRANSIENT MODEL SETUP ***************
# ***********************************************************


# initial setup complete, now run the model for a particular 2 min interval


# **************************************************************************************************************
# ***************************************** start NETLOGO SIMULATION  ******************************************

# ****************************************** open NETLOGO *****************************************

nl.path<- "<dir path to Netlogo program>" 
ver <-"<version number of Netlogo>" # type in version of Netlogo e.g. "6.0.1"
# if error, try adding "/app" to end of dir path for running in Windows and El Capitan for OSX
# nl.path<-"<dir path to Netlogo program>/app" 
NLStart(nl.path)
NLStart(nl.path, nl.jarname = paste0("netlogo-",ver,".jar"))
model.path<- "<dir path to Netlogo model>"
NLLoadModel(model.path)

# ****************************************** setup NETLOGO MODEL **********************************

# 1. update animal and env traits
month<-"sep"
NL_days<-117       # No. of days simulated
NL_gutthresh<-0.75
gutfull<-0.8

# set resource density
if(density=="high"){
	NL_shade<-100000L       # Shade patches
	NL_food<-100000L         # Food patches
	}else{ 
	NL_shade<-1000L 	# Shade patches
	NL_food<-1000L	# Food patches
}

# 2. update initial conditions for DEB model 
Es_pres_init<-(E_sm*gutfull)*V_pres_init
acthr<-1
Tb_init<-20
step = 1/24
debout<-DEB(step = step, z = z, del_M = del_M, F_m = F_m * 
    step, kap_X = kap_X, v = v * step, kap = kap, p_M = p_M * 
    step, E_G = E_G, kap_R = kap_R, k_J = k_J * step, E_Hb = E_Hb, 
    E_Hj = E_Hb, E_Hp = E_Hp, h_a = h_a/(step^2), s_G = s_G, 
    T_REF = T_REF, TA = TA, TAL = TAL, TAH = TAH, TL = TL, 
    TH = TH, E_0 = E_0, E_pres=E_pres_init, V_pres=V_pres_init, E_H_pres=E_H_init, acthr = acthr, breeding = 1, Es_pres = Es_pres_init, E_sm = E_sm)

# 3. calc direct movement cost
V_pres<-debout[2]
step<-1/24 #hourly

p_M2<-p_M*step #J/h
p_M2<-p_M2*V_pres # loco cost * structure
names(p_M2)<-NULL # remove V_pres name attribute from p_M

# movement cost for time period
VO2<-0.45 # O2/g/h JohnAdler etal 1986

# multiple p_M by structure = movement cost (diff between p_M with loco cost and structure for movement period)
# p_M with loco cost 
loco<-VO2*mass*20.1 # convert ml O2 to J = J/h 
loco<-loco+p_M2 # add to p_M = J/h
loco<-loco/30/V_pres ; loco #J/cm3/2min

Es_pres_init<-(E_sm*gutfull)*V_pres_init
X_food<-3000
V_pres<-debout[2]
wetgonad<-debout[19]
wetstorage<-debout[20]
wetfood<-debout[21]
ctminthresh<-120000
Tairfun<-Tairfun_shd
Tc_init<-Tairfun(1)+0.1 # Initial core temperature

NL_T_b<-Tc_init       # Initial T_b
NL_T_b_min<-VTMIN         # Min foraging T_b
NL_T_b_max<-VTMAX        # Max foraging T_b
NL_ctminthresh<-ctminthresh # No. of consecutive hours below CTmin that leads to death
NL_reserve<-E_m        # Initial reserve density
NL_max_reserve<-E_m    # Maximum reserve level
NL_maint<-round(p_M, 3)               # Maintenance cost
NL_move<-round(loco, 3) 		      # Movement cost
NL_zen<-Zenfun(1*60*60)     # Zenith angle

strategy<-function(strategy){ # set movement strategy 
  if (strategy == "O"){
    NLCommand("set strategy \"Optimising\" ") 
    }else{
    NLCommand("set strategy \"Satisficing\" ") 
    }
  }
strategy("O") # "S"

shadedens<-function(shadedens){ # set shade density to clumped (to match food) or sparse 
  if (shadedens == "Random"){
    NLCommand("set Shade-density \"Random\" ") 
    }else{
    NLCommand("set Shade-density \"Clumped\" ") 
    }
  }
shadedens("Clumped") # set clumped resources

sc<-1 # set no. of desired simualations---for automating writing of each sim results to file. N = N runs
for (i in 1:sc){ # start sc sim loop

NLCommand("set Shade-patches",NL_shade,"set Food-patches",NL_food,"set No.-of-days",NL_days,"set T_b precision",
NL_T_b, "2","set T_opt_lower precision", NL_T_b_min, "2","set T_opt_upper precision", NL_T_b_max, "2",
"set reserve-level", NL_reserve, "set Maximum-reserve", NL_max_reserve, "set Maintenance-cost", NL_maint,
"set Movement-cost precision", NL_move, "3", "set zenith", NL_zen, "set ctminthresh", NL_ctminthresh, 
"set gutthresh", NL_gutthresh, 'set gutfull', gutfull, 'set V_pres precision', V_pres, "5", 'set wetstorage precision', wetstorage, "5", 
'set wetfood precision', wetfood, "5", 'set wetgonad precision', wetgonad, "5","setup")

#NLCommand("inspect turtle 0")

NL_ticks<-NL_days / (2 / 60 / 24) # No. of NL ticks (measurement of days)
NL_T_opt_l<-NLReport("[T_opt_lower] of turtle 0")
NL_T_opt_u<-NLReport("[T_opt_upper] of turtle 0")

# data frame setup for homerange polygon
turtles<-data.frame() # make an empty data frame
NLReport("[X] of turtle 0"); NLReport("[Y] of turtle 0")
who<-NLReport("[who] of turtle 0")

# **********************************************************
# ******************** start NETLOGO SIMULATION  ***********

debcall<-0 # check for first call to DEB
stepcount<-0 # DEB model step count

for (i in 1:NL_ticks){
stepcount<-stepcount+1
NLDoCommand(1, "go")

######### Reporting presence of shade
shade<-NLGetAgentSet("in-shade?","turtles", as.data.frame=T); shade<-as.numeric(shade) # returns an agentset of whether turtle is currently on shade patch

# choose sun or shade
tick<-i
times3<-c(times2[tick],times2[tick+1])

if(shade==0){
  Qsolfun<-Qsolfun_sun
  Tradfun<-Tradfun_sun
  Tairfun<-Tairfun_sun
}else{
  Qsolfun<-Qsolfun_shd
  Tradfun<-Tradfun_shd
  Tairfun<-Tairfun_shd
}
if(i==1){
Tc_init<-Tairfun(1)+0.1 #initial core temperature
}

# one_lump_trans params
Qsol<-Qsolfun(mean(times3)); Qsol
vel<-velfun(mean(times3)) ;vel
Tair<-Tairfun(mean(times3));Tair
Trad<-Tradfun(mean(times3)); Trad
Zen<-Zenfun(mean(times3)); Zen

# calc Tb params at 2 mins interval
Tbs<-onelump_varenv(t=120,time=times3[2],Tc_init=Tc_init,thresh = 30, AMASS = mass, lometry = 3, Tairf=Tairfun,Tradf=Tradfun,velf=velfun,Qsolf=Qsolfun,Zenf=Zenfun)
Tb<-Tbs$Tc
rate<-Tbs$dTc
Tc_init<-Tb

NLCommand("set T_b precision", Tb, "2") # Updating Tb
NLCommand("set zenith", Zenfun(times3[2])) # Updating zenith

# time spent below VTMIN
ctminhours<-NLReport("[ctmincount] of turtle 0") * 2/60 # ticks to hours
if (ctminhours == NL_ctminthresh) {NLCommand("ask turtle 0 [stop]")}

# ******************** start DEB SIMULATION  ***************

if(stepcount==1) { # run DEB loop every time step (2 mins)
stepcount<-0

# report activity state
actstate<-NLReport("[activity-state] of turtle 0")
 # Reports true if turtle is in food 
actfeed<-NLGetAgentSet("in-food?","turtles", as.data.frame=T); actfeed<-as.numeric(actfeed)
 
n<-1 # time steps
step<-2/1440 # step size (2 mins). For hourly: 1/24
# update direct movement cost
if(actstate == "S"){
	NLCommand("set Movement-cost", NL_move)
	}else{
		NLCommand("set Movement-cost", 1e-09)
		} 
# if within activity range, it's daytime, and gut below threshold 
if(Tbs$Tc>=VTMIN & Tbs$Tc<=VTMAX & Zen!=90 & gutfull<=NL_gutthresh){ 
  acthr=1 # activity state = 1 
if(actfeed==1){ # if in food patch
	X_food<-NLReport("[energy-gain] of turtle 0") # report joules intake
	}
	}else{
		X_food = 0 
		acthr=0
		}

# calculate DEB output 
if(debcall==0){
	# initialise DEB
	debout<-matrix(data = 0, nrow = n, ncol = 26)
	deb.names<-c("E_pres","V_pres","E_H_pres","q_pres","hs_pres","surviv_pres","Es_pres","cumrepro","cumbatch","p_B_past","O2FLUX","CO2FLUX","MLO2","GH2OMET","DEBQMET","DRYFOOD","FAECES","NWASTE","wetgonad","wetstorage","wetfood","wetmass","gutfreemass","gutfull","fecundity","clutches")
	colnames(debout)<-deb.names
	# initial conditions
	debout<-DEB(E_pres=E_pres_init, V_pres=V_pres_init, E_H_pres=E_H_init, acthr = acthr, Tb = Tb_init, breeding = 1, Es_pres = Es_pres_init, E_sm = E_sm, step = step, z, del_M = del_M, F_m = F_m * 
    step, kap_X = kap_X, v = v * step, kap = kap, p_M = p_M * 
    step, E_G = E_G, kap_R = kap_R, k_J = k_J * step, E_Hb = E_Hb, 
    E_Hj = E_Hb, E_Hp = E_Hp, h_a = h_a/(step^2), s_G = s_G, 
    T_REF = T_REF, TA = TA, TAL = TAL, TAH = TAH, TL = TL, 
    TH = TH, E_0 = E_0)
	debcall<-1
	}else{
		debout<-DEB(step = step, z = z, del_M = del_M, F_m = F_m * 
    step, kap_X = kap_X, v = v * step, kap = kap, p_M = p_M * 
    step, E_G = E_G, kap_R = kap_R, k_J = k_J * step, E_Hb = E_Hb, 
    E_Hj = E_Hb, E_Hp = E_Hp, h_a = h_a/(step^2), s_G = s_G, 
    T_REF = T_REF, TA = TA, TAL = TAL, TAH = TAH, TL = TL, 
    TH = TH, E_0 = E_0, 
		  X=X_food,acthr = acthr, Tb = Tbs$Tc, breeding = 1, E_sm = E_sm, E_pres=debout[1],V_pres=debout[2],E_H_pres=debout[3],q_pres=debout[4],hs_pres=debout[5],surviv_pres=debout[6],Es_pres=debout[7],cumrepro=debout[8],cumbatch=debout[9],p_B_past=debout[10])
		}
mass<-debout[22]
gutfull<-debout[24]
NL_reserve<-debout[1]
V_pres<-debout[2]
wetgonad<-debout[19]
wetstorage<-debout[20]
wetfood<-debout[21] 

#update NL wetmass properties 
NLCommand("set V_pres precision", V_pres, "5")
NLDoCommand("plot xcor ycor")
NLCommand("set wetgonad precision", wetgonad, "5")
NLDoCommand("plot xcor ycor")
NLCommand("set wetstorage precision", wetstorage, "5")
NLDoCommand("plot xcor ycor")
NLCommand("set wetfood precision", wetfood, "5")
NLDoCommand("plot xcor ycor") 
 

} #--- end DEB loop

NLCommand("set reserve-level", NL_reserve) # update reserve
NLCommand("set gutfull", debout[24])# update gut level

# ******************** end DEB SIMULATION ******************

# generate results, with V_pres, wetgonad, wetstorage, and wetfood from debout
if(i==1){
	results<-cbind(tick,Tb,rate,shade,V_pres,wetgonad,wetstorage,wetfood,NL_reserve) 
	}else{
		results<-rbind(results,c(tick,Tb,rate,shade,V_pres,wetgonad,wetstorage,wetfood,NL_reserve))
		}
results<-as.data.frame(results)

# generate data frames for homerange polygon
if (tick == NL_ticks - 1){
	X<-NLReport("[X] of turtle 0"); head(X)
	Y<-NLReport("[Y] of turtle 0"); head(Y)
	turtles<-data.frame(X,Y)
	who1<-rep(who,NL_ticks); who # who1<-rep(who,NL_ticks - 1); who 
	turtledays<-rep(1:NL_days,length.out=NL_ticks,each=720) 
	turtle<-data.frame(ID = who1,days=turtledays)
	turtles<-cbind(turtles,turtle)
	}

} # *************** end NL loop **************************

# get hr data
spdf<-SpatialPointsDataFrame(turtles[1:2], turtles[3]) # creates a spatial points data frame (adehabitatHR package)
homerange<-mcp(spdf,percent=95)

# writing new results
if (exists("results")){  #if results exist
	sc<-sc-1 
	nam <- paste("results", sc, sep = "") # generate new name with added sc count
	rass<-assign(nam,results) #assign new name to results. call 'results1, results2 ... resultsN'
	namh <- paste("turtles", sc, sep = "")  #generate new name with added sc count
	rassh<-assign(namh,turtles) #assign new name to results. call 'results1, results2 ... resultsN'
	nams <- paste("spdf", sc, sep = "") 
	rasss<-assign(nams,spdf) 
	namhr <- paste("homerange", sc, sep = "")  
	rasshr<-assign(namhr,homerange) 

	fh<-results.path; fh
	for (i in rass){
		# export all results
		write.table(results,file=paste(fh,nam,".R",sep=""))
		}
	for (i in rassh){
		# export turtle location data
		write.table(turtles,file=paste(fh,namh,".R",sep=""))
		}
		#export NL plots
		month<-"sep"
		#spatial plot
		sfh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_move","",sep="");sfh
		NLCommand(paste("export-plot \"Spatial coordinates of transition between activity states\" \"",results.path,sfh,".csv\"",sep=""))
		#temp plot 
		tfh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_temp",sep="")
		NLCommand(paste("export-plot \"Body temperature (T_b)\" \"",results.path,tfh,".csv\"",sep=""))
		#activity budget
		afh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_act","",sep="");afh
		NLCommand(paste("export-plot \"Global time budget\" \"",results.path,afh,".csv\"",sep=""))
		#text output
		xfh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_txt",sep="");xfh
		NLCommand(paste("export-output \"",results.path,xfh,".csv\"",sep=""))
		#gut level
		gfh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_gut","",sep="");gfh
		NLCommand(paste("export-plot \"Gutfull\" \"",results.path,gfh,".csv\"",sep=""))
		#wet mass 
		mfh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_wetmass","",sep="");mfh
		NLCommand(paste("export-plot \"Total wetmass plot\" \"",results.path,mfh,".csv\"",sep=""))
		#movement cost (loco) 
		lfh<-paste(month,NL_days,round(mass,0),NL_shade,as.integer(NL_food*10),"_",sc,"_loco","",sep="");lfh
		NLCommand(paste("export-plot \"Movement costs\" \"",results.path,lfh,".csv\"",sep=""))
	}
} # ********************** end sc sim loop *****************************

# example files of results output in results.path
list.files(results.path)

#*********************** end NETLOGO SIMULATION ****************************
#***************************************************************************