numerical vector classes

  CDSVectors are container objects which behaves much like Python
sequences: the [] operator behaves as expected, returning an
element or slice and the len() function returns the size of the
list. Note that a proper Python list is generated using the slice [:].

Linear algebraic functions (+*-) are also defined on CDSVector objects. 

constructors:
  CDSVector_type(size)
    where the vector size is specified.
  CDSVector_type(python_object)
    where python_object is a sequence whose elements are compatible with type.
  CDSVector_type(other_CDSVector_type)
    this is a copy constructor.

classes:

CDSVector_double
  vector of floating point values. CDSVector_doubles may be multiplied
by a <m cdsMatrix>.RMat of the appropriate size.

CDSVector_Vec3
  vector of <m vec3>.Vec3 objects.

CDSVector_int
  vector of integer values.

the following functions will act on CDSVector objects:

  sqrt(vec) - return a new CDSVector object each of whose elements is
              the sqrt of the correspondig argument of vec.
  log(vec)  - return a CDSVector whose elements are logrithms of the 
	      corresponding elements of vec.
  vec_norm  - return a CDSVector_double whose elements are the norm of the 
	      corresponding elements of vec.
  norm(vec) - return the norm of the entire vector.
  sum(vec)  - return the sum of the entire vector.
  __pow__
  
miscellaneous members
  fromList(obj) - read a compatible Python object into the CDSVector.
  scale(x)      - scale each element by the value x.
  __pow__(x)    - raise each member to the x power (accessed via the ** 
                  operator). 

Note that mathematical operations which take element members out of
range are not supported. For instance, sqrt(CDSVector_int) is not
supported.