0

Good afternoon, I'm finding myself in the need to generate a column of 8760 numbers following a Weibull probability density distribution with specific parametres, which has been an easy task so far. Nevertheless, I also need another property to be respected: the differences between any number and the previous one in the column must also follow a probability density distribution, which I haven't chosen yet, amongst gamma, beta or rayleigh. I'm working with Excel and MatLab, and I have no idea how to achieve that, especially considering that I couldn't find any case like this on the internet.

How can I solve this problem?

Thanks

M. Barbaro
  • 1
  • 1
  • 1
    Well, the usual way to construct a list with a constraint on successive differences would be to generate values from the specified distribution for the differences, and then add them up. But I think that's inconsistent with the other requirement -- given a distribution for the differences, there will be just one distribution for their cumulative sums. I don't see a way to get a specific distribution for the cumulative sums. – Robert Dodier Jul 17 '18 at 00:10
  • This seems confused. What you are saying sort of sounds like "I want to generate circles. I know how to do this, but I also need another property to be respected: the ratio of the circumference to the diameter must be some other number which I haven't chosen yet, maybe `pi`, maybe `e`, maybe something else." You don't have as many degrees of freedom as you think you have. – John Coleman Jul 17 '18 at 12:14
  • This seems like a modeling problem rather than a programming problem. Perhaps you can post a question to [statistics.se] which explains what you are trying to do. They might help you to formulate your goal in a way which doesn't seem contradictory. – John Coleman Jul 17 '18 at 13:11

0 Answers0