0

I have a relatively big dataset for which I want to make a simulation like what are the consequences of some scenario.

I can make a monte carlo simulation but there are lots of variables so it would be impossible for the user to specify a probability distribution function for each variable.

What other ways you suggest to define scenarios which would be possible (user can't specify input for each variable ) and to make the simulation.

I am thinking about a way for example to generate a new dataset from the previous one based on some scenario and which will be scientifically correct.

Notes :
- all of this is going to be inside an application.
- the dataset variables are dependent on time

Atmo101
  • 31
  • 3
  • This is way too broad for SO. You realize that there are dozens of books and tens of thousands (see, e.g., https://informs-sim.org) of papers written about how to build and analyze simulation models? – pjs Mar 31 '19 at 19:15
  • Thanks for your comment. Yes I knew the subject is broad,I could use Monte Carlo but as I said there are lots of variables and in Monte Carlo the user has to input PDF for each variable, I was just hoping that maybe there is some simple idea or method that could be used to avoid making the user input PDF for each variable. – Atmo101 Mar 31 '19 at 21:14
  • If you look through the help section regarding how to ask a good question, you'll find that very specifically targeted questions tend to work best around here. It helps if you provide what you've done so far, and where you're having trouble with it. Broad or open-ended questions tend to get shut down as off-topic, and often downvoted. I'd strongly recommend that you read the guidelines. – pjs Mar 31 '19 at 22:32

0 Answers0