Sampling distribution functions can be used on the PSA Inputs page to specify how parameters are sampled in probabilisitic sensitivity analyses. A list of sampling distribution functions that you may use is provided below, along with relevant syntax.
Type |
Function |
Description |
Example |
Normal |
normal(mean, sd) |
Defines sampling for a normally-distributed variable |
normal(mean = 24.1, sd = 8.3) |
Log-normal |
lognormal(mean, sd, meanlog, sdlog) |
Defines sampling for a lognormally-distributed variable. Mean and SD can be provided on a log scale (using the ‘meanlog’ and ‘sdlog’ arguments) |
lognormal(mean = 24.1, sd = 8.3) |
Gamma |
gamma(mean, sd) |
Defines sampling for a gamma-distributed variable |
gamma(mean = 1.2, sd = 0.3) |
Binomial |
binomial(prob, size) |
Defines sampling for a binomially-distributed variable |
binomial(prob = 0.2, size = 50) |
Logit-normal |
logitnormal(mu, sigma) |
Defines sampling for a variable distributed normally on the logit scale |
logitnormal(mu = 2.3, sigma = 0.7) |
Poisson |
poisson(mean) |
Defines sampling for a poisson-distributed variable |
poisson(mean = 4.31) |
Beta |
beta(shape1, shape2) |
Defines sampling for a beta-distributed variable |
beta(shape1 = 3.3, shape2 = 5.4) |
Triangle |
triangle(lower, upper, peak = (lower + upper)/2) |
Defines sampling for a triangular-distributed variable |
triangle(upper = 5.4, lower = 2.3, peak = 4.1) |