Inverse transform sampling
Inverse transform sampling (also known as inversion sampling, the inverse probability integral transform, the inverse transformation method, Smirnov transform, golden rule, etc.) is a basic method for pseudo-random number sampling, i.e. for generating sample numbers at random from any probability distribution given its cumulative distribution function (cdf).
The basic idea is to uniformly sample a number between 0 and 1, interpreted as a probability, and then return the largest number from the domain of the distribution such that . For example, imagine that is the standard normal distribution (i.e. with mean 0, standard deviation 1). Then if we choose , we would return 0, because 50% of the probability of a normal distribution occurs in the region where . Similarly, if we choose , we would return 1.95996...; if we choose , we would return 2.5758...; if we choose , we would return 4.891638...; etc. Essentially, we are randomly choosing a proportion of the area under the curve and returning the number in the domain such that exactly this proportion of the area occurs to the left of that number. Intuitively, we are unlikely to choose a number in the tails because there is very little area in them: We'd have to pick a number very close to 0 or 1.
Computationally, this method involves computing the quantile function of the distribution — in other words, computing the cumulative distribution function (CDF) of the distribution (which maps a number in the domain to a probability between 0 and 1) and then inverting that function. This is the source of the term "inverse" or "inversion" in most of the names for this method. Note that for a discrete distribution, computing the CDF is not in general too difficult: We simply add up the individual probabilities for the various points of the distribution. For a continuous distribution, however, we need to integrate the probability density function (PDF) of the distribution, which is impossible to do analytically for most distributions (including the normal distribution). As a result, this method may be computationally inefficient for many distributions and other methods are preferred; however, it is a useful method for building more generally applicable samplers such as those based on rejection sampling.
For the normal distribution, the lack of an analytical expression for the corresponding quantile function means that other methods (e.g. the Box–Muller transform) may be preferred computationally. It is often the case that, even for simple distributions, the inverse transform sampling method can be improved on: see, for example, the ziggurat algorithm and rejection sampling. On the other hand, it is possible to approximate the quantile function of the normal distribution extremely accurately using moderate-degree polynomials, and in fact the method of doing this is fast enough that inversion sampling is now the default method for sampling from a normal distribution in the statistical package R.
The probability integral transform states that if is a continuous random variable with cumulative distribution function , then the random variable has a uniform distribution on [0, 1]. The inverse probability integral transform is just the inverse of this: specifically, if has a uniform distribution on [0, 1] and if has a cumulative distribution , then the cumulative distribution function of the random variable is .
The problem that the inverse transform sampling method solves is as follows:
- Let X be a random variable whose distribution can be described by the cumulative distribution function F.
- We want to generate values of X which are distributed according to this distribution.
The inverse transform sampling method works as follows:
- Generate a random number u from the standard uniform distribution in the interval [0,1].
- Compute the value x such that F(x) = u.
- Take x to be the random number drawn from the distribution described by F.
Expressed differently, given a continuous uniform variable U in [0, 1] and an invertible cumulative distribution function F, the random variable X = F −1(U) has distribution F (or, X is distributed F).
A treatment of such inverse functions as objects satisfying differential equations can be given. Some such differential equations admit explicit power series solutions, despite their non-linearity.
Proof of correctness
Claim: If U is a uniform random variable on (0, 1) then follows the distribution F.
- Probability integral transform
- Copula, defined by means of probability integral transform.
- Quantile function, for the explicit construction of inverse CDFs.
- Inverse distribution function for a precise mathematical definition for distributions with discrete components.
- Aalto University, N. Hyvönen, Computational methods in inverse problems. Twelfth lecture https://noppa.tkk.fi/noppa/kurssi/mat-1.3626/luennot/Mat-1_3626_lecture12.pdf
- Luc Devroye (1986). Non-Uniform Random Variate Generation. New York: Springer-Verlag.
- Steinbrecher, G., Shaw, W.T. (2008). Quantile mechanics. European Journal of Applied Mathematics 19 (2): 87–112.
- Luc Devroye (1986). "Section 2.2. Inversion by numerical solution of F(X) = U". Non-Uniform Random Variate Generation. New York: Springer-Verlag.