Distributing and accumulating dispersed seeds

Here is what I think will be an efficient way of putting the dispersed seeds in the right place. Let relDseeds be a \((2 D_{max}+1)\times n_{pots}\) matrix giving dispersal relative to the home pot (with row Dmax + 1 being retention and rows above that being backwards dispersal). Then we can create an absolute dispersal matrix:

# Lowest index of each column of relDseeds to include in Dseeds
lower <- pmax(1, Dmax - (1:npots) + 2) 

# Add an index row, available in the applied function as x[1]
relDseeds<-rbind(1:npots, relDseeds)

# Function to construct a column of Dseeds from a column of relDseeds
padCol <- function(x, lower, Dmax, npots) {
  leading_zeros <- rep(0, max(0, x[1] - Dmax - 1))
  dispersed_seeds <- x[-(1:lower[x[1]])]
  trailing_zeros <- rep(0, npots-x[1])
  c(leading_zeros, dispersed_seeds, trailing_zeros)
}

# Make a list of vectors for each column of Dseeds
Dseeds <- apply(relDseeds, 2, padCol, lower=lower, Dmax=Dmax, npots=npots)

# Convert the list to a matrix
Dseeds <- matrix(unlist(Dseeds), npots + Dmax, npots)

Finally, we can get the total number of seeds dispersed to each pot using apply:

apply(Dseeds, 1, sum)

I think that, if relDseeds is actually the 3rd and 4th dimension of an array (with genotypes and reps being the first two), then we can apply all of the above to efficiently do this calculation for all g and r.