Index: The Book of Statistical ProofsProbability Distributions ▷ Univariate continuous distributions ▷ Log-normal distribution ▷ Median

Theorem: Let $X$ be a random variable following a log-normal distribution:

$\label{eq:lognorm} X \sim \ln \mathcal{N}(\mu, \sigma^2) \; .$

Then, the median of $X$ is

$\label{eq:lognorm-med} \mathrm{median}(X) = e^\mu \; .$

Proof: The median is the value at which the cumulative distribution function is $1/2$:

$\label{eq:median} F_X(\mathrm{median}(X)) = \frac{1}{2} \; .$ $\label{eq:lognorm-cdf} F_X(x) = \frac{1}{2} \left[ 1 + \mathrm{erf} \left( \frac{\ln (x)-\mu}{\sigma \sqrt{2}} \right) \right]$

where $\mathrm{erf}(x)$ is the error function defined as

$\label{eq:erf} \mathrm{erf}(x) = \frac{2}{\sqrt{\pi}} \int_{0}^{x} \exp(-t^2) \, \mathrm{d}t \; .$

Thus, the inverse CDF is

$\label{eq:lognorm-cdf-inv} \begin{split} \ln(x) &= \sigma \sqrt{2} \cdot \mathrm{erf}^{-1}(2p-1) + \mu \\ x &= \mathrm{exp} \left[ \sigma \sqrt{2} \cdot \mathrm{erf}^{-1}(2p-1) + \mu \right] \end{split}$

where $\mathrm{erf}^{-1}(x)$ is the inverse error function. Setting $p = 1/2$, we obtain:

$\label{eq:lognorm-med-qed} \begin{split} \ln \left[ \mathrm{median}(X) \right] &= \sigma \sqrt{2} \cdot \mathrm{erf}^{-1}(0) + \mu \\ \mathrm{median}(X) &= e^\mu \; . \end{split}$
Sources:

Metadata: ID: P306 | shortcut: lognorm-med | author: majapavlo | date: 2022-02-07, 22:33.