09-13-2017 07:12 PM
I used an SAS platformm 9 years ago that used the Levenberg-Marquardt algorithm (SAS software (SAS 9.1.3 on a Service Pack 2 on a XP Pro Platform) that allowed to solve 12 simultanous equations for which there were more variables than equations. Is this platfrom still available?
DVK = [Dk3 + b*(Dk5 + a)/(1 + b(1+a) = 1.5
DKac = [Dk5 + c + Keq3*b*c]/[1 + c + Keq3 + bc + Keq5(1+d)] = 1.4
Kac = k7Keq3Keq5/[1 + c + Keq3 + bc + Keq5(1+d)] = 400
09-13-2017 08:09 PM
Do you happen to know the PROC?
I suspect its in SAS/OR so it's likely available but the question is usually do you have a license for it.
Not sure what you're doing but you can explore the lastest versions of SAS as an independent learner, ie no commercial usage at SAS Analytics U.
09-13-2017 08:19 PM
This is not really my field but it looks like Proc Calis might be worth checking out https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_calis_sect0...
09-13-2017 08:18 PM
Any code using SAS procedures that was written 10 (or 20 or more) years ago will still work in the latest versions.
Unlike most programmimng languages.
09-13-2017 08:29 PM
Tell me about it @ChrisNZ - earlier this year I had to fix some C# code which had broken simply by virtue of a new version .NET being installed; I eventually managed it but not before tearing out most of what little hair I have left...
09-13-2017 08:43 PM
Anything from Microsoft is temporary and on borrowed time it seems.
Ctrl-F in outlook is no longer shortcut for Find for goodness's sake! They deserve to die now.
Even Python deprecates things in bulk. Sigh...
09-13-2017 11:27 PM
From the SAS/ETS documentation for proc model:
PROC MODEL features include the following: SAS programming statements to define simultaneous systems of nonlinear equations tools to analyze the structure of the simultaneous equation system ARIMA, PDL, and other dynamic modeling capabilities tools to specify and estimate the error covariance structure tools to estimate and solve ordinary differential equations the following methods of parameter estimation: ordinary least squares (OLS) two-stage least squares (2SLS) seemingly unrelated regression (SUR) and iterative SUR (ITSUR) three-stage least squares (3SLS) and iterative 3SLS (IT3SLS) generalized method of moments (GMM) simulated method of moments (SMM) full information maximum likelihood (FIML) general log-likelihood maximization simulation and forecasting capabilities Monte Carlo simulation goal-seeking solutions
and yes, proc model supports the Marquardt-Levenberg minimization method.
The SAS Forecasting and Econometrics forum community is devoted to SAS/ETS tools.
09-14-2017 10:06 AM
Be aware that "solve" in this instance is somewhat ambiguous. For example, if you have 15 variables for those 12 equations then you have a three-dimensional space of feasible solutions. If you are content to get any feasible solution, you can do an optimization: input the equations as nonlinear constraints and use a trivial objective function (such as the identity function).