1
$\begingroup$

How can i recover μ and σ for the lognormal distribution (income) knowing mean GDP per-capita (which should be my μ) and the Gini coefficient with data on 5 quintile income shares? Thanks to all.

σ is a number. In the formula for σ = 2/ erf (g), erf is a function defined for values of x that I do not know.

$\endgroup$
  • 1
    $\begingroup$ The parameter $\mu$ of a log-normal distribution is usually not its mean but the mean of its logarithm $\endgroup$ – Henry Jun 19 '17 at 16:03
  • $\begingroup$ By the way, $\sigma = 2 \, {\rm erf}^{-1}(g)$ in my answer below does not mean $\sigma = 2 / {\rm erf}(g)$. In this context ${\rm erf}^{-1}$ means the inverse function, not one over the value of the function. Just like $\sin^{-1}$ means arcsine, not one over sine. $\endgroup$ – David Wright Jun 20 '17 at 20:47
2
$\begingroup$

So you know the mean $m$ and Gini $g$ of a distribution, and, assuming it is a lognormal distribution, you want to get the lognormal parameters $\mu$ and $\sigma$? That is pretty straightforward.

The mean and Gini of a lognormal are: $$m = \exp(\mu + \sigma^2 /2) \qquad g = {\rm erf}(\sigma / 2)$$ Just invert these equations to obtain: $$\sigma = 2 \, {\rm erf}^{-1}(g) \qquad \mu = \ln(m) - \sigma^2 / 2$$

$\endgroup$
  • $\begingroup$ yes but the value of erf? It is a function defined for values of x that I do not know! I know only the data that I listed above $\endgroup$ – userIT8991 Jun 20 '17 at 18:41
  • 1
    $\begingroup$ @a.russolT: ${\rm erf}$ and ${\rm erf}^{-1}$ are just special functions like log or exp; you can find calculators for them online (keisan.casio.com/exec/system/1180573448). You said you know the Gini. Suppose the Gini is 0.40; ${\rm erf}^{-1}(0.40) = 0.37$, so $\sigma = 2 \times 0.37 = 0.74$. $\endgroup$ – David Wright Jun 20 '17 at 20:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.