loglinv
statistics: x = loglinv (p, mu, sigma)
Inverse of the log-logistic cumulative distribution function (iCDF).
For each element of p, compute the quantile (the inverse of the CDF) of the log-logistic distribution with mean parameter mu and scale parameter sigma. The size of x is the common size of p, mu, and sigma. A scalar input functions as a constant matrix of the same size as the other inputs.
Mean of logarithmic values mu must be a non-negative real value, scale
parameter of logarithmic values sigma must be a positive real value and
p is supported in the range , otherwise NaN is
returned.
Further information about the loglogistic distribution can be found at https://en.wikipedia.org/wiki/Log-logistic_distribution
OCTAVE/MATLAB use an alternative parameterization given by the pair , i.e. mu and sigma, in analogy with the logistic distribution. Their relation to the and parameters used in Wikipedia are given below:
mu = log (a)
sigma = 1 / a
See also: loglcdf, loglpdf, loglrnd, loglfit, logllike, loglstat
Source Code: loglinv
Plot various iCDFs from the log-logistic distribution
p = 0.001:0.001:0.999;
x1 = loglinv (p, log (1), 1/0.5);
x2 = loglinv (p, log (1), 1);
x3 = loglinv (p, log (1), 1/2);
x4 = loglinv (p, log (1), 1/4);
x5 = loglinv (p, log (1), 1/8);
plot (p, x1, '-b', p, x2, '-g', p, x3, '-r', p, x4, '-c', p, x5, '-m')
ylim ([0, 20])
grid on
legend ({'σ = 2 (β = 0.5)', 'σ = 1 (β = 1)', 'σ = 0.5 (β = 2)', ...
'σ = 0.25 (β = 4)', 'σ = 0.125 (β = 8)'}, 'location', 'northwest')
title ('Log-logistic iCDF')
xlabel ('probability')
ylabel ('x')
text (0.03, 12.5, 'μ = 0 (α = 1), values of σ (β) as shown in legend')