Shifted log-logistic distribution

The shifted log-logistic distribution is a probability distribution also known as the generalized log-logistic or the three-parameter log-logistic distribution.[1][2] It has also been called the generalized logistic distribution,[3] but this conflicts with other uses of the term: see generalized logistic distribution.

Shifted log-logistic
Probability density function
values of as shown in legend
Cumulative distribution function
values of as shown in legend
Parameters

location (real)
scale (real)

shape (real)
Support



PDF


where
CDF


where
Mean


where
Median
Mode
Variance


where

Definition

edit

The shifted log-logistic distribution can be obtained from the log-logistic distribution by addition of a shift parameter  . Thus if   has a log-logistic distribution then   has a shifted log-logistic distribution. So   has a shifted log-logistic distribution if   has a logistic distribution. The shift parameter adds a location parameter to the scale and shape parameters of the (unshifted) log-logistic.

The properties of this distribution are straightforward to derive from those of the log-logistic distribution. However, an alternative parameterisation, similar to that used for the generalized Pareto distribution and the generalized extreme value distribution, gives more interpretable parameters and also aids their estimation.

In this parameterisation, the cumulative distribution function (CDF) of the shifted log-logistic distribution is

 

for  , where   is the location parameter,   the scale parameter and   the shape parameter. Note that some references use   to parameterise the shape.[3][4]

The probability density function (PDF) is

 

again, for  

The shape parameter   is often restricted to lie in [-1,1], when the probability density function is bounded. When  , it has an asymptote at  . Reversing the sign of   reflects the pdf and the cdf about  .

edit
  • When   the shifted log-logistic reduces to the log-logistic distribution.
  • When   → 0, the shifted log-logistic reduces to the logistic distribution.
  • The shifted log-logistic with shape parameter   is the same as the generalized Pareto distribution with shape parameter  

Applications

edit

The three-parameter log-logistic distribution is used in hydrology for modelling flood frequency.[3][4][5]

Alternate parameterization

edit

An alternate parameterization with simpler expressions for the PDF and CDF is as follows. For the shape parameter  , scale parameter   and location parameter  , the PDF is given by [6][7]

 

The CDF is given by

 

The mean is   and the variance is  , where  .[7]

References

edit
  1. ^ Venter, Gary G. (Spring 1994), "Introduction to selected papers from the variability in reserves prize program" (PDF), Casualty Actuarial Society Forum, 1: 91–101
  2. ^ Geskus, Ronald B. (2001), "Methods for estimating the AIDS incubation time distribution when date of seroconversion is censored", Statistics in Medicine, 20 (5): 795–812, doi:10.1002/sim.700, PMID 11241577
  3. ^ a b c Hosking, Jonathan R. M.; Wallis, James R (1997), Regional Frequency Analysis: An Approach Based on L-Moments, Cambridge University Press, ISBN 0-521-43045-3
  4. ^ a b Robson, A.; Reed, D. (1999), Flood Estimation Handbook, vol. 3: "Statistical Procedures for Flood Frequency Estimation", Wallingford, UK: Institute of Hydrology, ISBN 0-948540-89-3
  5. ^ Ahmad, M. I.; Sinclair, C. D.; Werritty, A. (1988), "Log-logistic flood frequency analysis", Journal of Hydrology, 98 (3–4): 205–224, doi:10.1016/0022-1694(88)90015-7
  6. ^ "EasyFit - Log-Logistic Distribution". Retrieved 1 October 2016.
  7. ^ a b "Guide to Using - RISK7_EN.pdf" (PDF). Retrieved 1 October 2016.