Survey about convergence of iterations « Next Oldest | Next Newest »

 ▼ Namir Posting Freak Posts: 2,247 Threads: 200 Joined: Jun 2005 03-21-2009, 10:10 AM Hello All, When performing iterative processes you often compare one of more values between the current and last iterations. This comparison involves a tolerance value used to determine whether or not the current iteration has generated good enough results. My question to all out there is "Do you prefer comparing the absolute difference between iteration values or a percentage absolute difference with a tolerance value? Keep in mind that the tolerance value or % can be specified with the arguments to the procedure performing the iterations. Fixing the tolerance value or % is a bit more tricky since it may not be able to handle a wide variety of problems. Thanks Namir Edited: 21 Mar 2009, 10:12 a.m. ▼ Egan Ford Posting Freak Posts: 1,619 Threads: 147 Joined: May 2006 03-21-2009, 01:28 PM Quote: Do you prefer comparing the absolute difference between iteration values or a percentage absolute difference with a tolerance value? The former. Usually something like: ```if |diff| < 1E-desired precision then end loop ``` Mad Dog ebaycalcnut Member Posts: 124 Threads: 17 Joined: Dec 2006 03-22-2009, 03:25 PM Actually, I concur. If you pick a reasonably small absolute value, and I do mean reasonable compared to the solution values, you should not need to mess with the headache-y percentage tolerance. It would be interesting to try a coupled algorithm where you first try an absolute value tolerance, get a solution, then switch to a percentage tolerance using as initial values solutions from the absolute value tolerance problem. I guess this would essentially be a double convergence. Karl Schneider Posting Freak Posts: 1,792 Threads: 62 Joined: Jan 2005 03-22-2009, 04:05 PM Namir -- If the calculated value(s) that establish convergence are large in magnitude, they can be normalized by scaling. One application is power-flow programs for complicated AC networks. Sums of calculated values of power establish network convergence when they match known injection values. These values are large -- typically in megawatts (MW) and megavars (MVAr) -- along with voltage in kilovolts (kV) and current in hundreds of amperes (A). The powerflow calculations are done utilizing normalized "per unit" values, typically ranging from zero to ten. Mismatches are then small in magnitude, so absolute tolerances are employed. -- KS

 Possibly Related Threads... Thread Author Replies Views Last Post Convergence Dream Machine anetzer 11 2,509 04-12-2013, 07:16 AM Last Post: David Hayden HP 41CX Survey Pac Keith Burr 7 2,015 03-29-2013, 10:09 AM Last Post: Keith Burr HP-41C Custom Survey Module Identificaiton Help Dan Grelinger 0 748 01-27-2013, 01:51 PM Last Post: Dan Grelinger Error Message/Procedure HP-48SX with TDS Survey SX card bporter101 6 1,496 12-10-2012, 10:18 AM Last Post: bporter101 Survey: Best programming language for next-gen HP? Oliver Unter Ecker 32 6,397 06-06-2012, 03:22 PM Last Post: David Hayden Survey for Special Math Problem Namir 7 1,855 06-03-2012, 09:46 PM Last Post: Namir [WP34S] Numeric Integrator Has Max Iterations Set? Les Wright 6 1,676 11-23-2011, 05:35 AM Last Post: Paul Dale HP15C LE Keyboard Survey Results M. Joury 23 4,211 10-02-2011, 07:52 PM Last Post: M. Joury HP15C LE Keyboard Survey Take 3 (last round) M. Joury 33 5,624 09-28-2011, 02:28 PM Last Post: Michael de Estrada HP 15C LE Keyboard Survey Redux M. Joury 57 9,531 09-25-2011, 04:03 PM Last Post: db (martinez, ca.)

Forum Jump: