understanding GAMS syntax in a mathematical way

Problems with syntax of GAMS
yanzhiping
User
User
Posts: 85
Joined: 9 months ago

understanding GAMS syntax in a mathematical way

Postby yanzhiping » 5 months ago

mathematically, if there are two equations

x1=9;
x1>8;

then you will get the result x1=9.

but in GAMS. if you write x1.fx=9;
then x1.lo=8.

then it will treat x1 as 8<x1<9.

this is very confusing since it is not the same as the mathematical problem. and it could change the problem in a wrong way. since we want x1=9 rather than 8<x1<9.

Is there any way to avoid this?

cladelpino
User
User
Posts: 55
Joined: 10 months ago

Re: understanding GAMS syntax in a mathematical way

Postby cladelpino » 5 months ago

Hi, as you can read in https://www.gams.com/latest/docs/usergu ... arl/fx.htm, when doing x1.fx=9 actually:
... GAMS sets the lower and upper bound to that value.
This means x1.fx=9 is exactly doing x1.lo=9 and x1.up=9 in one statement.

Additionaly, this kind of statements (setting variable attributes) are executed in order. So when you do x1.lo=8 afterwards, you are "overwriting" the previous set of the lower bound.

Best regards!
Claudio

yanzhiping
User
User
Posts: 85
Joined: 9 months ago

Re: understanding GAMS syntax in a mathematical way

Postby yanzhiping » 4 months ago

how about if I have two lines as below.

q<some value;
q.up=inf;

will the second line determine the upper bound of q which makes the first line useless?

cladelpino
User
User
Posts: 55
Joined: 10 months ago

Re: understanding GAMS syntax in a mathematical way

Postby cladelpino » 4 months ago

you mean a constraint defined ? ie

c1.. q=l=someValue;

q.up=someValue;

This would lead to models which are identical on the mathematical level, but NOT on the "software" (GAMS+solver) level.


Who is online

Users browsing this forum: No registered users and 2 guests