BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
Vane08
Fluorite | Level 6
/* II-  PROGRAME OPTMODEL 															*/ 
	/***********************************************************************************/
					proc optmodel; 
						
						/* 0 Tableau des chemins & Marges								*/ 
						/************************************************************/
								/* declaration de paramètre 	*/
						
						Set <string> Vecteur_&Typej._&Heure._&Annee.; 
						Set <string> Marge_&Typej._&Heure._&Annee.; 
						num Tij{Vecteur_&Typej._&Heure._&Annee.};

					    num marges{Marge_&Typej._&Heure._&Annee.};
						num STDMarges{Marge_&Typej._&Heure._&Annee.};
						num Nobs{Marge_&Typej._&Heure._&Annee.};
						num chemin{Vecteur_&Typej._&Heure._&Annee., Marge_&Typej._&Heure._&Annee.};
								
						/* Reading data 													*/ 
						/******************************************************************/

					 	read data Vecteur_&Typej._&Heure._&Annee. into 
							Vecteur_&Typej._&Heure._&Annee. = [Vecteur] Tij; 
						print Tij;

					    read data Marge_&Typej._&Heure._&Annee. into 
							Marge_&Typej._&Heure._&Annee. = [N] 
							marges = marges
							STDMarges = STDMarges
							NObs = Nobs 
							{p in Vecteur_&Typej._&Heure._&Annee.} <chemin[p,N]= col(p)>;
						print  marges STDMarges Nobs chemin;
							

						/* III Resolution  												*/ 
						/*********************************************************************/
							/* V-1) Déclaration de variables */ 
							
							var  X{Vecteur_&Typej._&Heure._&Annee.} >= 0;
							Var  Y{Marge_&Typej._&Heure._&Annee.} >= 0 ; 
							

							/* V-2) La fonction objective : minimisation de la variance */ 

							minimize f = sum{p in Vecteur_&Typej._&Heure._&Annee.} (X[p]-Tij[p])**2/(Tij[p]/3)**2 + sum{r in Marge_&Typej._&Heure._&Annee.} ((Y[r]- Marges[r])**2/(STDMarges[r]**2/Nobs[r]));



							/* V-3) subject to the following constraints */

							Con Bornes{r in Marge_&Typej._&Heure._&Annee.}: sum{p in Vecteur_&Typej._&Heure._&Annee.} chemin[p, r]*X[p] = Y[r];

/* Con Condtion1{r in Marge_&Typej._&Heure._&Annee. :}: If r = 17 then sum{p in Vecteur_&Typej._&Heure._&Annee.} chemin[p, r]*X[p] > 0;
Con Condtion1{r in Marge_&Typej._&Heure._&Annee. :}: If r = 16 then sum{p in Vecteur_&Typej._&Heure._&Annee.} chemin[p, r]*X[p] > 0;*/
							
						
							
						 solve ;
						print X;
						Print Y;
						
						
						create data Vecteur_est_&Typej._&Heure._&Annee. from [p] Tij&Annee. = X ;
						create data Marge_est_&Typej._&Heure._&Annee. from [r]  Marges&Annee. = Y ;
						

Good morning, 

 

I want to make conditions in my constraint because in the solutions of X and Y, I need some sum in the line of Y like : 

 

1) In the first solution Y: (Marge_est_&Typel.&Heure._&Annee. ) with Y[r] 

                 ==> I need that  Y[16] = Y[17] +Y[18]   with Y[17]#0 and Y[18] #0 

                       (there is some case like those lines but not all the line in the table, it is a transportation problem..) 

                                                 and in the same time

2) I have the constraint that the sum in line of  X[1...52]*chemin[52, 31]  is equal at the sum of Y[1..31]  

                      for example : for the first line of Y   : [1...52]*chemin[1..52, 1] = Y[1] ; I have no problem to write this second constraint "con" like in the code I sent here in Pj. 

 

==>The origin of this problem is that in my  real data set in the table   Marge_&Typel.&Heure._&Annee.  :

                                               the line 17 = 0 , the line 18 = 0 so the QP affect s100% of Y[16] in just one of Y[18] or Y[17]

 

==>So I thought that the solutions were : making  some condition contraint to estimates line in this case to have in the estimation not Y[16] =  Y[17]  +0        with Y[18] = 0

                   but Y [16] = Y[17] +Y[16]   with Y[18] #0 and Y[17] #0

 

Thank you in advance for your response 

@RobPratt

1 ACCEPTED SOLUTION

Accepted Solutions
RobPratt
SAS Super FREQ

If I understand correctly, you should include the following two additional statements:

   con YSum: Y['N16'] = Y['N17'] + Y['N18'];
   for {r in /N17 N18/} Y[r].lb = 1e-3;

 

The first statement declares your new constraint, and the second statement changes the lower bounds on Y['N17'] and Y['N18'] from 0 to 1e-3.  Here, 1e-3 is just an example.  You have to decide for yourself how large Y[r] must be for you to consider it to be positive.

View solution in original post

2 REPLIES 2
RobPratt
SAS Super FREQ

If I understand correctly, you should include the following two additional statements:

   con YSum: Y['N16'] = Y['N17'] + Y['N18'];
   for {r in /N17 N18/} Y[r].lb = 1e-3;

 

The first statement declares your new constraint, and the second statement changes the lower bounds on Y['N17'] and Y['N18'] from 0 to 1e-3.  Here, 1e-3 is just an example.  You have to decide for yourself how large Y[r] must be for you to consider it to be positive.

Vane08
Fluorite | Level 6

Thank you for your response ! It gives me the syntax for a condition but I think that I need changing something in my resolution programm. 

 

sas-innovate-2024.png

Don't miss out on SAS Innovate - Register now for the FREE Livestream!

Can't make it to Vegas? No problem! Watch our general sessions LIVE or on-demand starting April 17th. Hear from SAS execs, best-selling author Adam Grant, Hot Ones host Sean Evans, top tech journalist Kara Swisher, AI expert Cassie Kozyrkov, and the mind-blowing dance crew iLuminate! Plus, get access to over 20 breakout sessions.

 

Register now!

Multiple Linear Regression in SAS

Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 2 replies
  • 951 views
  • 0 likes
  • 2 in conversation