MathGroup Archive 1995

[Date Index] [Thread Index] [Author Index]

Search the Archive

Teaching the UnitStep[x] function to be smarter

  • To: mathgroup at christensen.cybernetics.net
  • Subject: [mg1589] Teaching the UnitStep[x] function to be smarter
  • From: danny at cherrypit.Princeton.EDU (Mark D. Rintoul III)
  • Date: Mon, 3 Jul 1995 02:46:02 -0400
  • Organization: Princeton University

	I often use the DiracDelta.m package to do integrals, although
most because of it's UnitStep[x] function.  It's usually very handy,
but I'm not quite satisfied with the way that it checks to see if it's
arguments are positive and negative.  In many cases, it should be
fairly clear that it is either specifically 0 or 1, given that all of
it's arguments are defined to be positive (or negative).  Does anyone
have any experience dealing with this (or related) function?
Also, as a general reference, can anyone point me to some sample documents
that deal with establishing a relationship between set's of variables
that a function will understand.  For example, I'd like my functions to
know that not only are A and B positive, but A > B so A-B is also
positive.  Thanks for any help.

					Danny Rintoul
					danny at cherrypit.princeton.edu
					http://cherrypit.princeton.edu/~danny


  • Prev by Date: Embedded Data Within PS Graphics
  • Next by Date: META-Question: Object-oriented graphics in MMA?
  • Previous by thread: Embedded Data Within PS Graphics
  • Next by thread: projected gradient