calcbits:

--------------------------------------------------------------------------
   Determine the number of bits to represent a float number to the 
   precision provided.

   Since version 7.
--------------------------------------------------------------------------
   Form:
   [bits]=calcbits(bounds,precision)
--------------------------------------------------------------------------

   ------
   Inputs
   ------  
   bounds            the bounds on the variables 
   precision         the least difference to distinguish two numbers
  
   -------
   Outputs
   -------
   bits              the number of bits required per variable

--------------------------------------------------------------------------