Hp50g MES - Printable Version +- HP Forums (https://www.hpmuseum.org/forum) +-- Forum: HP Calculators (and very old HP Computers) (/forum-3.html) +--- Forum: General Forum (/forum-4.html) +--- Thread: Hp50g MES (/thread-15326.html) Hp50g MES - Tugdual - 07-09-2020 04:03 PM Does anybody know what algorithm is used in the Hp50g Multiple Equation Solver? RE: Hp50g MES - Joe Horn - 07-10-2020 03:27 AM The HP Solve Equation Library Application Card Owner's Manual says on page 169: "When you solve problems using the Multiple-Equation Solver, the application uses the same numeric root finder that's used by the HP Solve application, which is built into the HP 48." I assume that the same is true for the HP 50g. RE: Hp50g MES - Tugdual - 07-10-2020 10:44 AM (07-10-2020 03:27 AM)Joe Horn Wrote:  The HP Solve Equation Library Application Card Owner's Manual says on page 169: "When you solve problems using the Multiple-Equation Solver, the application uses the same numeric root finder that's used by the HP Solve application, which is built into the HP 48." I assume that the same is true for the HP 50g. Thanks Joe. Would that then be a non linear equation solver? Like Levenberg-Marquardt method? RE: Hp50g MES - Joe Horn - 07-10-2020 12:13 PM (07-10-2020 10:44 AM)Tugdual Wrote:  Would that then be a non linear equation solver? Like Levenberg-Marquardt method? Anybody know? If it's the same as the solver that HP first introduced in the HP-34C, then it's a "modified secant iteration method" as explained here by Dr Kahan himself in 1979: https://people.eecs.berkeley.edu/~wkahan/Math128/SOLVEkey.pdf RE: Hp50g MES - JurgenRo - 07-10-2020 06:35 PM (07-10-2020 12:13 PM)Joe Horn Wrote:   (07-10-2020 10:44 AM)Tugdual Wrote:  Would that then be a non linear equation solver? Like Levenberg-Marquardt method? Anybody know? If it's the same as the solver that HP first introduced in the HP-34C, then it's a "modified secant iteration method" as explained here by Dr Kahan himself in 1979: https://people.eecs.berkeley.edu/~wkahan/Math128/SOLVEkey.pdf Thanks for the reading, Joe! RE: Hp50g MES - ndzied1 - 07-10-2020 09:07 PM Not sure if this is what is used in the 50g but here is an HP patent for a multiple equation solver. It seems to be more about the multiple equation part than the solving of each equation. https://patents.google.com/patent/US5175700 Method and apparatus for solving multiple equations It also references this patent: https://patents.google.com/patent/US4845652A/en Generic equation solver interface for solving mathematical equations Very much wish the Prime will eventually incorporate this type of multiple equation solving. RE: Hp50g MES - Thomas Okken - 07-10-2020 09:41 PM It sounds like this older topic might be relevant... not my area of expertise but hoping it may be useful: Multiple equation solver for the HP-28C (program) RE: Hp50g MES - Tugdual - 07-13-2020 04:18 PM Thanks all for the replies. So to summarize it would be a sort of combination of an algo to try to solve obvious solutions, when only one variable remains and/or a secant method applied to multiple variables, hence something resembling https://en.wikipedia.org/wiki/Broyden%27s_method to expand on mutliple variables.