combine {bayesAB}

Combine two bayesAB objects given a binary function.

Description

Combine two (or any number, in succession) bayesTest objects into a new arbitrary posterior distribution. The resulting object is of the same class.

Usage

combine(bT1, bT2, f = `+`, params, newName)

Arguments

bT1

a bayesTest object

bT2

a bayesTest object

f

a binary function (f(x, y)) used to combine posteriors from bT1 to bT2

params

a character vector of length 2, corresponding to names of the posterior parameters you want to combine; defaults to first posterior parameter if not supplied

newName

a string indicating the name of the new 'posterior' in the resulting object; defaults to string representation of f(params[1], params[2])

Value

a bayesTest object with the newly combined posterior samples.

Note

The generics '+.bayesTest', '*.bayesTest', '-.bayesTest', and '/.bayesTest' are shorthand for combine(f = '+'), combine(f = '*'), combine(f = '-'), and combine(f = '/').

See Also

grab

Examples

A_binom <- rbinom(100, 1, .5)
B_binom <- rbinom(100, 1, .6)

A_norm <- rnorm(100, 6, 1.5)
B_norm <- rnorm(100, 5, 2.5)

AB1 <- bayesTest(A_binom, B_binom,
                 priors = c('alpha' = 1, 'beta' = 1),
                 distribution = 'bernoulli')

AB2 <- bayesTest(A_norm, B_norm,
                 priors = c('mu' = 5, 'lambda' = 1, 'alpha' = 3, 'beta' = 1),
                 distribution = 'normal')

AB3 <- combine(AB1, AB2, f = `*`, params = c('Probability', 'Mu'), newName = 'Expectation')
# Equivalent to
AB3 <- AB1 * grab(AB2, 'Mu')

# To get the same posterior name as well
AB3 <- rename(AB3, 'Expectation')

# Dummy example
weirdVariable <- (AB1 + AB2) * (AB2 / AB1)
weirdVariable <- rename(weirdVariable, 'confusingParam')

print(AB3)
summary(AB3)
plot(AB3)