R/ParallelModel.R

In SoilR: Models of Soil Organic Matter Decomposition

#
# vim:set ff=unix expandtab ts=2 sw=2:
ParallelModel=structure(function
### This function creates a (linear) numerical model for n independent (parallel) pools that can be queried afterwards. 
### It is used by the convinient wrapper functions \code{\link{TwopParallelModel}} and \code{\link{ThreepParallelModel}}
### but can also be used independently.
(times,		##<< A vector containing the points in time where the solution is sought.
 coeffs_tm,	##<< A TimeMap object consisting of a vector valued function containing the decay rates for the n pools as function of time and the time range where this function is valid. The length of the vector is equal to the number of pools.
 startvalues,	##<< A vector containing the initial amount of carbon for the n pools. 
 ##<<The length of this vector is equal to the number of pools and thus equal to the length of k. This is checked by the function.
 inputrates, ##<< An  object consisting of a vector valued function describing the inputs to the pools as funtions of time \code{\link{TimeMap.new}}
 solverfunc =deSolve.lsoda.wrapper,    ##<< The function used to actually solve the ODE system. This can be \code{\link{deSolve.lsoda.wrapper}} or any other user provided function with the same interface. 
  pass=FALSE  ##<< if TRUE forces the constructor to create the model even if it is invalid 
 ){
    coeffs=getFunctionDefinition(coeffs_tm)
    ns=length(startvalues)
    nk=length(coeffs(1))
    if (nk!=ns){
       print("The vectors startvalues and coeffs are not of the same length")
    }
    A=function(t){diag(x=coeffs(t))}
    tstart=getTimeRange(coeffs_tm)[[1]]
    tend=getTimeRange(coeffs_tm)[[2]]
    A_tm=BoundLinDecompOp(A,tstart,tend)
    obj=Model(times,A_tm,startvalues,inputrates,solverfunc,pass)
### a model object
}
,ex=function(){
      t_start=0 
      t_end=10 
      tn=50
      timestep=(t_end-t_start)/tn 
      t=seq(t_start,t_end,timestep) 
      k=TimeMap(
        function(times){c(-0.5,-0.2,-0.3)},
        t_start,
        t_end
      )
      c0=c(1, 2, 3)
      #constant inputrates
      inputrates=BoundInFlux(
          function(t){matrix(nrow=3,ncol=1,c(1,1,1))},
          t_start,
          t_end
      ) 
      mod=ParallelModel(t,k,c0,inputrates)
      Y=getC(mod)
      lt1=1 ;lt2=2 ;lt3=3 
      col1=1; col2=2; col3=3
      plot(t,Y[,1],type="l",lty=lt1,col=col1,
	   ylab="C stocks",xlab="Time") 
      lines(t,Y[,2],type="l",lty=lt2,col=col2) 
      lines(t,Y[,3],type="l",lty=lt3,col=col3) 
      legend(
	 "topleft",
	 c("C in pool 1",
	   "C in 2",
	   "C in pool 3"
	 ),
	 lty=c(lt1,lt2,lt3),
	 col=c(col1,col2,col3)
      )
      Y=getAccumulatedRelease(mod)
      plot(t,Y[,1],type="l",lty=lt1,col=col1,ylab="C release",xlab="Time") 
      lines(t,Y[,2],lt2,type="l",lty=lt2,col=col2) 
      lines(t,Y[,3],type="l",lty=lt3,col=col3) 
      legend("topright",c("R1","R2","R3"),lty=c(lt1,lt2,lt3),col=c(col1,col2,col3))
 
}       
)