Inverse Transform Sampling - Proof of Correctness

Proof of Correctness

Let F be a continuous cumulative distribution function, and let F−1 be its inverse function (using the infimum because CDFs are weakly monotonic and right-continuous):

Claim: If U is a uniform random variable on (0, 1) then follows the distribution F.

Proof:

\begin{align} & \Pr(F^{-1}(U) \leq x) \\ & {} = \Pr(\inf\;\{y \mid F(y)=U\} \leq x)\quad &\text{(by definition of }F^{-1}) \\ & {} = \Pr(U \leq F(x)) \quad &\text{(applying }F,\text{ which is monotonic, to both sides)} \\ & {} = F(x)\quad &\text{(because }\Pr(U \leq y) = y,\text{ since }U\text{ is uniform on the unit interval)}. \end{align}

Read more about this topic:  Inverse Transform Sampling

Famous quotes containing the words proof of, proof and/or correctness:

    “In the reproof of chance
    Lies the true proof of men.”
    —William Shakespeare (1564–1616)

    “Right and proof are two crutches for everything bent and crooked that limps along.”
    —Franz Grillparzer (1791–1872)

    “Rather would I have the love songs of romantic ages, rather Don Juan and Madame Venus, rather an elopement by ladder and rope on a moonlight night, followed by the father’s curse, mother’s moans, and the moral comments of neighbors, than correctness and propriety measured by yardsticks.”
    —Emma Goldman (1869–1940)