Re: Log-Logistic denisty function



Martin Fravi wrote:
In Matlab, the log-logistic distribution/denisty function has only 2
parameters (a location parameter and a scale parameter). But the
log-logistic function has also a shape parameter (see for example
www.causascientia.org/math_stat/Dists/Compendium.pdf). So I am
confused!

Martin, in addition to what Tom mentioned in his post, it's also possible that you are referring to a generalization of the logistic/log-logistic distributions in which the exponent in the denominator of the density is a positive shape parameter, rather than being fixed at the value 2. While the Statistics Toolbox does not explicitly support that generalization, it may be possible to use the "custom distribution" fitting feature of the MLE function to fit such a distribution to data, though I am not familiar enough with the distribution to know whether maximum likelihood estimation works well. It's also fairly straight-forward to create an M file that allows you to add this as a custom distribution to the DIstribution Fitting tool DFITTOOL.

Hope this helps.

- Peter Perkins
The MathWorks, Inc.
.



Relevant Pages

  • Re: Weibull Trouble
    ... distribution after running a life test with no failures. ... It seems as if you want to do statistical Power analysis ... "problem" is that of running a life-test with no failures. ... tested with failures and should have a similar shape parameter. ...
    (sci.stat.math)
  • Re: fitting a generalized pareto
    ... Similar is in for the extreme value distribution, I made a new logpdf and logsurvival for the generalized pareto.: ... Limiting the search area for the shape parameter only above zero, gives a non-optimal convergences and therefor a bad fit. ...
    (comp.soft-sys.matlab)
  • random numbers from beta distribution
    ... % The generalized beta distribution, ... first/right shape parameter ... right/upper boundary of support (real) ...
    (comp.soft-sys.matlab)
  • Re: Fitting exponential and Erlang Distribution
    ... distribution. ... What is the bes way to estimate the shape parameter k? ... How about a maximum likelihood estimate? ... MLE for the parameters of the gamma distribution in the first part of ...
    (sci.stat.math)
  • Re: Statistical Distribution Analysis
    ... Martin Fravi wrote: ... that a given distribution ... Martin, if you have access to the Statistics Toolbox, there is a GUI known as the Distribution Fitting Tool that more or less does what you are looking for. ... However, it is not simply a black box that spits out what it thinks is the best fitting distribution, nor would you want that. ...
    (comp.soft-sys.matlab)