Notes on the runtime of A* sampling

The challenge of simulating random variables is a central problem in Statistics and Machine Learning. Given a tractable proposal distribution $P$, from which we can draw exact samples, and a target distribution $Q$ which is absolutely continuous with respect to $P$, that is $P \gg Q$, the A* sampling algorithm can be used to draw exact samples from $Q$, so long as we can evaluate the Radon-Nikodym derivative of $Q$ with respect to $P$. Maddison et al. originally showed that for a target distribution $Q$ and proposal distribution $P$, the runtime of A* sampling is upper bounded by $\mathcal{O}(\exp(D_{\infty}[Q||P]))$ where $D_{\infty}[Q||P]$ is the Renyi divergence from $Q$ to $P$. This runtime can be prohibitively long for many cases of practical interest. Here, we show that with additional restrictive assumptions on $Q$ and $P$, we can achieve much faster runtimes. Specifically, we show that if $Q$ and $P$ are distributions on $\mathbb{R}$ and their Radon-Nikodym derivative is unimodal, the runtime of A* sampling is $\mathcal{O}(D_{\infty}[Q||P])$, which is exponentially faster than A* sampling without further assumptions.