I need help to determine the distribution type for my set of data

7 views (last 30 days)
I have an array of data that needs to be fitted with the pd function. But the pd function needs me to specify what distribution my data is in (distname). But that's the things that I don't know about my data. How can I know what my distribution my data is in?

Accepted Answer

John D'Errico
John D'Errico on 6 Feb 2023
You CANNOT know what distribution any set of nummbers were generated from. That is provably imposssible. For example:
x = 0.432562446545
What distribution does it come from? Surely, if it is possible, if there were some scheme to do so, then you can tell me?
Oh wait. Now you say that you need more data? Is this enough?
x = [0.432562446545 , 1.2357657767]
Maybe a third point? Sorry, but no. No matter how much data I give you, there will never be a way to know in advance which distribution was used to generate some set of numbers.
You can use statistics, to decide that with some degree of probability, that some distribution was not involved, but even then the statistics could conclude falsely that some specific distribution may have been or was not the source of any given set of data.
So all you can do is compare the probabilities that a given set of data may have resulted from some distribution. And there are various tests one can employ, and various ways one can estimate the distibution parameters. These are not questions about MATLAB of course, and since I'm certainly not a statistician, I would suggest reading a text on statistics for this. Or, perhaps, ask the question on an appropriate forum, about statistics.

More Answers (0)




Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by