PoissonBinomial (version 1.2.7)

GenPoissonBinomial-Distribution: The Generalized Poisson Binomial Distribution

Description

Density, distribution function, quantile function and random generation for the generalized Poisson binomial distribution with probability vector probs.

Usage

dgpbinom(x, probs, val_p, val_q, wts = NULL, method = "DivideFFT", log = FALSE)
pgpbinom(
  x,
  probs,
  val_p,
  val_q,
  wts = NULL,
  method = "DivideFFT",
  lower.tail = TRUE,
  log.p = FALSE
)
qgpbinom(
  p,
  probs,
  val_p,
  val_q,
  wts = NULL,
  method = "DivideFFT",
  lower.tail = TRUE,
  log.p = FALSE
)
rgpbinom(
  n,
  probs,
  val_p,
  val_q,
  wts = NULL,
  method = "DivideFFT",
  generator = "Sample"
)

Value

dgpbinom gives the density, pgpbinom computes the distribution function, qgpbinom gives the quantile function and rgpbinom

generates random deviates.

For rgpbinom, the length of the result is determined by n, and is the lengths of the numerical arguments for the other functions.

Arguments

x

Either a vector of observed sums or NULL. If NULL, probabilities of all possible observations are returned.

probs

Vector of probabilities of success of each Bernoulli trial.

val_p

Vector of values that each trial produces with probability in probs.

val_q

Vector of values that each trial produces with probability in 1 - probs.

wts

Vector of non-negative integer weights for the input probabilities.

method

Character string that specifies the method of computation and must be one of "DivideFFT", "Convolve", "Characteristic", "Normal" or "RefinedNormal" (abbreviations are allowed).

log, log.p

Logical value indicating if results are given as logarithms.

lower.tail

Logical value indicating if results are \(P[X \leq x]\) (if TRUE; default) or \(P[X > x]\) (if FALSE).

p

Vector of probabilities for computation of quantiles.

n

Number of observations. If length(n) > 1, the length is taken to be the number required.

generator

Character string that specifies the random number generator and must either be "Sample" or "Bernoulli" (abbreviations are allowed).

Details

See the references for computational details. The Divide and Conquer ("DivideFFT") and Direct Convolution ("Convolve") algorithms are derived and described in Biscarri, Zhao & Brunner (2018). They have been modified for use with the generalized Poisson binomial distribution. The Discrete Fourier Transformation of the Characteristic Function ("Characteristic") is derived in Zhang, Hong & Balakrishnan (2018), the Normal Approach ("Normal") and the Refined Normal Approach ("RefinedNormal") are described in Hong (2013). They were slightly adapted for the generalized Poisson binomial distribution.

In some special cases regarding the values of probs, the method parameter is ignored (see Introduction vignette).

Random numbers can be generated in two ways. The "Sample" method uses R's sample function to draw random values according to their probabilities that are calculated by dgpbinom. The "Bernoulli" procedure ignores the method parameter and simulates Bernoulli-distributed random numbers according to the probabilities in probs and sums them up. It is a bit slower than the "Sample" generator, but may yield better results, as it allows to obtain observations that cannot be generated by the "Sample" procedure, because dgpbinom may compute 0-probabilities, due to rounding, if the length of probs is large and/or its values contain a lot of very small values.

References

Hong, Y. (2018). On computing the distribution function for the Poisson binomial distribution. Computational Statistics & Data Analysis, 59, pp. 41-51. tools:::Rd_expr_doi("10.1016/j.csda.2012.10.006")

Biscarri, W., Zhao, S. D. and Brunner, R. J. (2018) A simple and fast method for computing the Poisson binomial distribution. Computational Statistics and Data Analysis, 31, pp. 216–222. tools:::Rd_expr_doi("10.1016/j.csda.2018.01.007")

Zhang, M., Hong, Y. and Balakrishnan, N. (2018). The generalized Poisson-binomial distribution and the computation of its distribution function. Journal of Statistical Computational and Simulation, 88(8), pp. 1515-1527. tools:::Rd_expr_doi("10.1080/00949655.2018.1440294")

Examples

set.seed(1)
pp <- c(1, 0, runif(10), 1, 0, 1)
qq <- seq(0, 1, 0.01)
va <- rep(5, length(pp))
vb <- 1:length(pp)

dgpbinom(NULL, pp, va, vb, method = "DivideFFT")
pgpbinom(75:100, pp, va, vb, method = "DivideFFT")
qgpbinom(qq, pp, va, vb, method = "DivideFFT")
rgpbinom(100, pp, va, vb, method = "DivideFFT")

dgpbinom(NULL, pp, va, vb, method = "Convolve")
pgpbinom(75:100, pp, va, vb, method = "Convolve")
qgpbinom(qq, pp, va, vb, method = "Convolve")
rgpbinom(100, pp, va, vb, method = "Convolve")

dgpbinom(NULL, pp, va, vb, method = "Characteristic")
pgpbinom(75:100, pp, va, vb, method = "Characteristic")
qgpbinom(qq, pp, va, vb, method = "Characteristic")
rgpbinom(100, pp, va, vb, method = "Characteristic")

dgpbinom(NULL, pp, va, vb, method = "Normal")
pgpbinom(75:100, pp, va, vb, method = "Normal")
qgpbinom(qq, pp, va, vb, method = "Normal")
rgpbinom(100, pp, va, vb, method = "Normal")

dgpbinom(NULL, pp, va, vb, method = "RefinedNormal")
pgpbinom(75:100, pp, va, vb, method = "RefinedNormal")
qgpbinom(qq, pp, va, vb, method = "RefinedNormal")
rgpbinom(100, pp, va, vb, method = "RefinedNormal")