make_MC_sample: Makes a Monte Carlo sample using Sobol' sequences
In jkeirstead/jkr: Functions and packages frequently used by James Keirstead

Description Usage Arguments Value See Also Examples

Generates a Monte Carlo sample using Sobol' low-discrepancy quasi-random sequences. Random sampling with R's standard methods is inefficient for Monte Carlo analysis as the sampled values do not cover the parameter space evenly. This function allows users to create parameter samples using quasi-random low-discrepancy Sobol' sequences to get around this problem.

1	make_MC_sample(n, vals)

n

number of samples to draw

vals

a list describing distributions. Each list entry should consist of the following named fields:

var a variable name
dist the distribution name, e.g. 'unif' for Uniform (see R's distribution names)
params a list of distribution parameters. The list names will vary by distribution.

a data frame with an index column "n" and each distribution in a named column

https://gist.github.com/jkeirstead/1730440

{
## Define the distributions
vals <- list(list(var="Uniform",
              dist="unif",
              params=list(min=0,max=1)),
         list(var="Normal",
              dist="norm",
              params=list(mean=0,sd=1)),
         list(var="Weibull",
              dist="weibull",
              params=list(shape=2,scale=1)))

## Generate 100 samples
samp <- make_MC_sample(100, vals)
}