|
An -dimensional rectangle is a subset of
of the form
, where each is an interval, with end points
, where
is the set of extended real numbers (so that
itself may be considered as an interval).
Groundedness. A function
is said to be grounded if for each
, and each
where , the function
defined by
is right-continuous at , the lower end point of .
Margin. Note that defined above may or may not exist as each
, the upper end point of ( ). If the limit exists, then we call this limiting function, also written , a (one-dimensional) margin of :
Given an -dimensional rectangle
, let's call each a side of . A vertex of is a point
such that each of its coordinates is an end point. Clearly is a convex set and the sides and vertices lie on the boundary of .
-volume. Suppose we have a function
, with defined as above. Let be a closed -dimensional rectangle in (
), with sides
,
. The -volume of is the sum
where is a vertex of , is the number of lower end points that occur in the coordinate representation of , and the sum is taken over all vertices of .
The name volume is derived from the fact that if
, then for each closed rectangle ,
is the volume of in the traditional sense.
Note, however, depending on the function ,
may be 0 or even negative. For example, if is a linear function, then the -volume is identically 0 for every closed rectangle , whenever is even. An example where
is negative is given by the function
, and is the unit square.
-increasing. A function
where is an open -dimensional rectange is said to be -increasing if
is non-negative evaluated at each closed rectangle
.
Any multivariate distribution function is both grounded and -increasing.
A copula, introduced by Sklar, is both a variant and a generalization of a multivariate distribution function.
Formally, a copula is a function from the -dimensional unit cube ( ) to
satisfying the following conditions:
is -increasing,
is grounded,
- every margin
of is the identity function.
If we replace the domain by any -dimensional rectangle , then the resulting function is called a subcopula.
For example, the functions
,
, and
defined on the unit cube are all copulas.
(This entry is in the process of being expanded, more to come shortly).
- 1
- B. Schweizer and A. Sklar, Probabilistic Metric Spaces, Dover Publications, (2005).
|