Advertisement

Lognormal distribution calculator

Use a lognormal model when the natural logarithm of a positive measurement is normally distributed. Enter log-space parameters, not the raw-data mean and SD.

Calculator

Value
1.000000
Density or probability mass
0.398942
P(X ≤ x)
50.000000%
P(X > x)
50.000000%
Selected probability
50.000000%
Mean
1.648721
Variance
4.670774
Standard deviation
2.161197

Advertisement

How this calculator works

Take natural logarithms before calculating the supplied mean and SD. The raw measurement is positive and generally right-skewed. The median is exp(μ), while the mean is larger because of the upper tail. Log parameters depend on the measurement unit, so keep that unit fixed throughout the calculation.

Formula

ln(X) ~ N(μ,σ²); F(x) = Φ((ln(x)-μ)/σ); E[X] = exp(μ+σ²/2); Var[X] = (exp(σ²)-1)exp(2μ+σ²).

Worked example

With μ = 0 and σ = 1, the median is 1, the mean is 1.6487213 and the probability below 1 is 50%.

Advertisement

Frequently asked questions

Which logarithm is used?

The natural logarithm, base e.

Can X equal zero?

The continuous support is strictly positive. Evaluation at zero returns zero cumulative probability.

Can I enter the arithmetic mean as μ?

No. μ is the mean after taking natural logarithms.

Sources

Advertisement