+ Reply to Thread
Results 1 to 3 of 3

Thread: 3 parameter distributions. pdfs?

  1. #1
    Points: 3,614, Level: 37
    Level completed: 76%, Points required for next Level: 36

    Thanked 0 Times in 0 Posts

    3 parameter distributions. pdfs?

    Hi, this is my first post and this looks like the most appropriate place for it. I've used minitab to get the most appropriate distributions for various sets of data. This data is then to be used to fuel inputs for a supply chain simulation model. Rather than look at the specifics of what came out I want to use the parameter estimates that come with it for the respective pdfs' and get plots so i can compare and contrast them visually.

    Anyway, thats the problem area. Specifically, 3-parameter distributions keep cropping up. I cannot seem to find the pdfs for the 3-parameter gamma and 3parameter log logistic distributions. Im also wondering, if im just looking at shape, does it matter? Because the third parameter generally seems to just be a location parameter and so will just cause the curve to "shift" up and down. Am I right in assuming that. Any info would be great.


  2. #2
    TS Contributor
    Points: 5,172, Level: 46
    Level completed: 11%, Points required for next Level: 178

    Athens , Greece
    Thanked 6 Times in 2 Posts
    Not very familiar with reliability issues but I think it matters whether using 2 or 3 parameters. So check this out and let us now what you came up with

    > http://www.weibull.com/LifeDataWeb/g...stribution.htm

  3. #3
    TS Contributor
    Points: 4,575, Level: 43
    Level completed: 13%, Points required for next Level: 175

    Nashville, TN
    Thanked 0 Times in 0 Posts

    In the Gaussian (Normal) PDF, there are two parameters, one for location (mean) and one for scale (variance). The gamma distribution also has two parameters, however, they do not have the same meaning as that for the normal. The mean and variance of the gamma are both functions of the two parameters. Hence, It is not possible to say which parameter has a greater effect of the location or scale of the distribution.

    As for your problem, I imagine that you are giving Minitab a set of data and it tells you what distribution provides the best fit and estimates of its parameters. In this case, it has suggested the generalized gamma, which has three parameters. This probably means that the two parameter gamma (and other two parameter distributions) were insufficient to capture the variation in the empirical distribution (your data). Given this, perhaps you should stick with the three parameter gamma.

    For plotting the PDF, you could use the function that is displayed at the HTML address given by mp83. However, you should make sure that the parameters in this function corrospond to those given by Minitab, as there are several parametrizations of the gamma PDF.


+ Reply to Thread


Similar Threads

  1. Replies: 8
    Last Post: 03-15-2013, 02:02 AM
  2. what to do for unknown PDFs.
    By sibtain125 in forum Statistical Research
    Replies: 0
    Last Post: 04-12-2011, 10:59 PM
  3. Replies: 0
    Last Post: 04-10-2009, 03:52 PM
  4. Determining Joint PDFs, and Marginal PDFs
    By woody198403 in forum Probability
    Replies: 13
    Last Post: 05-15-2008, 01:04 AM
  5. Exponential PDFS
    By Taka in forum Statistics
    Replies: 8
    Last Post: 04-23-2008, 06:10 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats