| Title: | Models for Simulating Populations |
|---|---|
| Description: | Run population simulations using an Individual-Based Model (IBM) compiled in C. |
| Authors: | Guillaume Chapron |
| Maintainer: | Guillaume Chapron <[email protected]> |
| License: | GPL-3 |
| Version: | 0.3 |
| Built: | 2024-10-25 03:30:34 UTC |
| Source: | https://github.com/gchapron/population |
A package to run population simulations using an Individual-Based Model compiled in C. The population model is a discrete, age-structured model and follows the formalizing of a post-breeding Leslie matrix model.
Version 0.1 proposes functions to run and plot population projections and includes demographic and environmental stochasticities. There is also the option to parallelize simulations (except on Windows).
Version 0.2 fixes a bug that generated wrong results at very large population sizes.
| Package: | population |
| Type: | Package |
| Version: | 0.2 |
| Date: | 2018-02-05 |
| License: | GPL-3 |
Guillaume Chapron <[email protected]>
# Initial number of individuals n0 <- 10 n1 <- 20 n2 <- 15 n3 <- 10 n4 <- 5 # Age-specific survival rates s0 <- 0.5 s1 <- 0.6 s2 <- 0.7 s3 <- 0.8 s4 <- 0.9 # Age-specific number of offspring b1 <- 0.5 b2 <- 0.8 b3 <- 1.8 b4 <- 1.8 b5 <- 1.1 # Project 10 years ahead repeated 10000 times years <- 10 runs <- 10000 results <- project( years = years, runs = runs, initial_population = c(n0, n1, n2, n3, n4), survival = cbind(c(s0, s1, s2, s3, s4), 0.0), # no environmental stochasticity litter = cbind(c(b1, b2, b3, b4, b5), 0.0) # no environmental stochasticity ) # Plot projection plot_projection(results, "mean") # Equivalent model with a post-breeding Leslie matrix postM <- matrix(nrow=5, ncol=5, byrow=TRUE, data = c( s0*b1, s1*b2, s2*b3, s3*b4, s4*b5, s0, 0, 0, 0, 0, 0, s1, 0, 0, 0, 0, 0, s2, 0, 0, 0, 0, 0, s3, 0 )) popvector <- c(n0, n1, n2, n3, n4) N <- numeric(years) N[1] <- sum(popvector) for (i in 2:years) { popvector <- postM N[i] <- sum(popvector) } # Check we get similar results lines(1:years, N, col="blue", lwd=2)# Initial number of individuals n0 <- 10 n1 <- 20 n2 <- 15 n3 <- 10 n4 <- 5 # Age-specific survival rates s0 <- 0.5 s1 <- 0.6 s2 <- 0.7 s3 <- 0.8 s4 <- 0.9 # Age-specific number of offspring b1 <- 0.5 b2 <- 0.8 b3 <- 1.8 b4 <- 1.8 b5 <- 1.1 # Project 10 years ahead repeated 10000 times years <- 10 runs <- 10000 results <- project( years = years, runs = runs, initial_population = c(n0, n1, n2, n3, n4), survival = cbind(c(s0, s1, s2, s3, s4), 0.0), # no environmental stochasticity litter = cbind(c(b1, b2, b3, b4, b5), 0.0) # no environmental stochasticity ) # Plot projection plot_projection(results, "mean") # Equivalent model with a post-breeding Leslie matrix postM <- matrix(nrow=5, ncol=5, byrow=TRUE, data = c( s0*b1, s1*b2, s2*b3, s3*b4, s4*b5, s0, 0, 0, 0, 0, 0, s1, 0, 0, 0, 0, 0, s2, 0, 0, 0, 0, 0, s3, 0 )) popvector <- c(n0, n1, n2, n3, n4) N <- numeric(years) N[1] <- sum(popvector) for (i in 2:years) { popvector <- postM N[i] <- sum(popvector) } # Check we get similar results lines(1:years, N, col="blue", lwd=2)
Get number of available cores for parallel simulations. Non-Windows systems only.
get_cores(runs)get_cores(runs)
runs |
Number of times (or Monte Carlo runs) to repeat the simulation. |
This function detects the number of cores (see 'detectCores' in package 'parallel') available and returns the largest possible number of cores being an integer divider of the number of runs. On multi-core machines at least one core is not used for the simulation.
get_cores()
get_cores(2) get_cores(1000)get_cores(2) get_cores(1000)
Plot population projections.
plot_projection(projection, kind)plot_projection(projection, kind)
projection |
A list obtained after running functions 'project' or 'project_cores'. |
kind |
(optional) A string indicating which quantity should be plotted ("median" or "mean"). If missing, all projections are shown. |
Plot all population projections or the median or mean with 95% confidence interval. Only total population sizes are displayed.
No returned value, plot created
years <- 10 runs <- 100 init.pop <- c(30, 20, 15, 12, 10, 9, 8, 7, 6, 5) surv.md <- c(0.5, 0.7, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9) surv.sd <- c(0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) surv.msd <- cbind(surv.md, surv.sd) litter.md <- c(0.2, 1.1, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 1.8, 0.2) litter.sd <- c(0.1, 0.2, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) litter.msd <- cbind(litter.md, litter.sd) nclass <- 4 # vary number of classes projection <- project( years = years, runs = runs, initial_population = init.pop[1:nclass], survival = surv.msd[1:nclass,], litter = litter.msd[1:nclass,] ) plot_projection(projection) plot_projection(projection, kind="median")years <- 10 runs <- 100 init.pop <- c(30, 20, 15, 12, 10, 9, 8, 7, 6, 5) surv.md <- c(0.5, 0.7, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9) surv.sd <- c(0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) surv.msd <- cbind(surv.md, surv.sd) litter.md <- c(0.2, 1.1, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 1.8, 0.2) litter.sd <- c(0.1, 0.2, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) litter.msd <- cbind(litter.md, litter.sd) nclass <- 4 # vary number of classes projection <- project( years = years, runs = runs, initial_population = init.pop[1:nclass], survival = surv.msd[1:nclass,], litter = litter.msd[1:nclass,] ) plot_projection(projection) plot_projection(projection, kind="median")
Run stochastic population projections.
project(years, runs, initial_population, survival, litter, seed)project(years, runs, initial_population, survival, litter, seed)
years |
Number of years to project the population. |
runs |
Number of times (or Monte Carlo runs) to project the population. |
initial_population |
Vector of initial number of individuals for each class. This vector must contain only positive integers. |
survival |
Matrix of survival for each class, with nrow = number of classes and ncol = 2. The first column is the median value of the survival of each class. The second column is the standard deviation of the survival of each class. |
litter |
Matrix of litter size for each class, with nrow = number of classes and ncol = 2. The first column is the median value of the litter size of each class. The second column is the standard deviation of the litter size of each class. |
seed |
(optional) seed for the R random number generator. If missing, the seed is set to 1. |
Run stochastic population projections with an Individual-Based Model (IBM) compiled in C.
runs |
a 3-dimensional array of numbers of individuals with dimension c(years, classes, runs) |
years <- 10 runs <- 100 init.pop <- c(30, 20, 15, 12, 10, 9, 8, 7, 6, 5) surv.md <- c(0.5, 0.7, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9) surv.sd <- c(0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) surv.msd <- cbind(surv.md, surv.sd) litter.md <- c(0.2, 1.1, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 1.8, 0.2) litter.sd <- c(0.1, 0.2, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) litter.msd <- cbind(litter.md, litter.sd) nclass <- 4 # vary number of classes projection <- project( years = years, runs = runs, initial_population = init.pop[1:nclass], survival = surv.msd[1:nclass,], litter = litter.msd[1:nclass,] )years <- 10 runs <- 100 init.pop <- c(30, 20, 15, 12, 10, 9, 8, 7, 6, 5) surv.md <- c(0.5, 0.7, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9) surv.sd <- c(0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) surv.msd <- cbind(surv.md, surv.sd) litter.md <- c(0.2, 1.1, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 1.8, 0.2) litter.sd <- c(0.1, 0.2, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) litter.msd <- cbind(litter.md, litter.sd) nclass <- 4 # vary number of classes projection <- project( years = years, runs = runs, initial_population = init.pop[1:nclass], survival = surv.msd[1:nclass,], litter = litter.msd[1:nclass,] )
Run parallel stochastic population projections. Non-Windows systems only.
project_parallel(years, runs, initial_population, survival, litter, cores)project_parallel(years, runs, initial_population, survival, litter, cores)
years |
Number of years to project the population. |
runs |
Number of times (or Monte Carlo runs) to project the population. |
initial_population |
Vector of initial number of individuals for each class. This vector must contain only positive integers. |
survival |
Matrix of survival for each class, with nrow = number of classes and ncol = 2. The first column is the median value of the survival of each class. The second column is the standard deviation of the survival of each class. |
litter |
Matrix of litter size for each class, with nrow = number of classes and ncol = 2. The first column is the median value of the litter size of each class. The second column is the standard deviation of the litter size of each class. |
cores |
(optional) number of cores to use for the parallel projections. If missing, it is set to the value returned by get_cores(). |
Run parallel stochastic population projections with an Individual-Based Model (IBM) compiled in C.
runs |
a 3-dimensional array of numbers of individuals with dimension c(years, classes, runs) |
years <- 10 runs <- 100 init.pop <- c(30, 20, 15, 12, 10, 9, 8, 7, 6, 5) surv.md <- c(0.5, 0.7, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9) surv.sd <- c(0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) surv.msd <- cbind(surv.md, surv.sd) litter.md <- c(0.2, 1.1, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 1.8, 0.2) litter.sd <- c(0.1, 0.2, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) litter.msd <- cbind(litter.md, litter.sd) nclass <- 4 # vary number of classes # with 2 cores projection <- project_parallel( years = years, runs = runs, initial_population = init.pop[1:nclass], survival = surv.msd[1:nclass,], litter = litter.msd[1:nclass,], cores = 2 ) # with all possible cores (not run) # projection <- project_parallel( # years = years, # runs = runs, # initial_population = init.pop[1:nclass], # survival = surv.msd[1:nclass,], # litter = litter.msd[1:nclass,] # )years <- 10 runs <- 100 init.pop <- c(30, 20, 15, 12, 10, 9, 8, 7, 6, 5) surv.md <- c(0.5, 0.7, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9, 0.9) surv.sd <- c(0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) surv.msd <- cbind(surv.md, surv.sd) litter.md <- c(0.2, 1.1, 2.8, 2.8, 2.8, 2.8, 2.8, 2.8, 1.8, 0.2) litter.sd <- c(0.1, 0.2, 0.15, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1) litter.msd <- cbind(litter.md, litter.sd) nclass <- 4 # vary number of classes # with 2 cores projection <- project_parallel( years = years, runs = runs, initial_population = init.pop[1:nclass], survival = surv.msd[1:nclass,], litter = litter.msd[1:nclass,], cores = 2 ) # with all possible cores (not run) # projection <- project_parallel( # years = years, # runs = runs, # initial_population = init.pop[1:nclass], # survival = surv.msd[1:nclass,], # litter = litter.msd[1:nclass,] # )