Copulas

Discussion in 'SP9' started by BeckyBoo, Feb 15, 2012.

  1. BeckyBoo

    BeckyBoo Member

    Am I the only one fidning copulas hard to get my head around? I am studying part 3 of the course and have now missed the deadline for assignment 3 as getting to grips with the idea of copulas is taking me a long time!
     
  2. Simon James

    Simon James ActEd Tutor Staff Member

    Hi BeckyBoo

    You won't be alone - this can be a tough concept when you first come across it. The basic idea is that a copula can be used to "join up" multiple marginal distributions to form a joint distribution.

    So if we have some real-world data … we may have an idea of the underlying marginal distributions. A copula is a way of combining these marginal distributions to form a joint distribution … which preserves any correlation/ dependency in the original data.

    Don't let this stop you progressing in the course. You may prefer to park copulas for now and move on. There is little in the remainder of the course which is dependent on a thorough understanding of the theory!
     

Share This Page