Number of Derivatives =/= Number of Initial Parameters Error in R

Question

I am trying to do an ODE solver but I am having issues with the vectors between my equations in the lotka() function and the initial parameters I set in the TLfit() function. I don't understand why it says I have a length 1 vector when I have 3 parameters listed. Could anyone offer some insight?

require(deSolve); # for ODE solver

# define the right-hand side of the ODE
lotka <- function(t,y,parms) {
  S=y[1]
  I=y[2]
  R=y[3]
  B=parms[1]; G=parms[2]; # parameters, like before
  dN=numeric(3) # vector to hold the derivatives
  dN[1]=-B*S*I
  dN[2]=B*S*I-(G*I)
  dN[3]=G*I
  return(list(dN))
}

# data 
mydata <- read.csv('PlagueBombay.csv')
tvals <- mydata$Week
xvals <-mydata$WeeklyDeaths

# Least squares objective function 
TLfit=function(logp) {
  p=exp(logp);
  parms=p[1:2]; I0=p[3]; times=1:32; 
  out=ode(I0,times,lotka,parms);
  mse=mean((out[,2]-xvals)^2);
  return(mse);
}
p0=c(B=.5,G=.5,I0=.5); # initial guess at parameters
logp0=log(p0); 
fit=optim(logp0,TLfit);

I get the error:

Error in checkFunc(Func2, times, y, rho) : 
  The number of derivatives returned by func() (3) must equal the length of the initial conditions vector (1)

Answer 1

You did exactly as the error says, your initial value is I0=p[3] which is a single scalar. You probably want something like y0=[1-I0,I0,0] as initial value.