invalid subscripts discarded

[D.T. Fokum]

Hello,

I have a model that uses a parameter y. In my data file I have the
following statements:

param y default 0 :=
[*,*,1]:
0 1 2 :=
2 1 . .
3 . . 1
4 . 1 .
[*,*,2]:
0 1 2 :=
5 1 . .
7 . 1 .
[*,*,3]:
0 1 2 :=
1 1 . .
6 . 1 . ;

However, when I try to solve my model I get a message that states:
Error executing "solve" command:
error processing param y:
2 invalid subscripts discarded:
y[5,0,2]
y[7,1,2]

From earlier posts on this group I found this link:
http://www.ampl.com/FAQ/#WhydoIgetaninvalidsubscriptdiscardedmessagewhenIdisplayanAMPLparameter,
unfortunately the information there doesn't appear to be related to my
problem. What am I doing incorrectly?

Thanks,
Daniel

[Robert Fourer]

The discarded invalid subscripts [5,0,2] and [7,1,2] correspond to the two
values in the parameter y data that have a third index equal to 2. So most
likely the problem is that 2 is not a member of the third indexing set for
y.

Bob Fourer
4...@ampl.com

[Daniel Fokum]

Bob,

Thanks for your response; my problem is resolved. The third indexing set for y was UNIT, which I had defined as:
set UNIT:= 0 .. 3 by 1;

I dropped " by 1" from the set declaration and the model was solved correctly without any subscripts discarded.

Thanks,
Daniel

[Paul]

Are MOSD, MOLD etc. supposed to belong to set DEST? If so, you omitted the final "D" when declaring values for DEST.

Paul

[shed...@gmail.com]

I have a similar problem. the invalid subscript problem. I've tried the default empty set someone mentioned however, I still have the following issues:

Is there a way to suppress the errors noted below and loop it so I don't have to type each let statement? AMPL  solves my problem, but the invalid subscripts causes problems when try to put it in a  for loop ?  I understand the error comes from the size of the set changing from 8 down to 3 and now the parameters linked to the dropped sets are producing invalid subscripts. please help ? 

here are my sets 

set pairs[1] := 1 2 3;

set pairs[2] := 1 2 4;

set pairs[3] := 1 2 5;

set pairs[4] := 1 2 6;

set pairs[5] := 1 2 7;

set pairs[6] := 1 2 8;

set pairs[7] := 1 3 4;

set pairs[8] := 1 3 5;

set pairs[9] := 1 3 6;

ampl: update data;

ampl: data CR3_14_14ver1.dat;

ampl: solve;

ampl: let AVAIL := pairs[1];

ampl: display PAT;

Error executing "display" command:

error processing param PAT:

90 invalid subscripts discarded:

PAT[5,1,1]

PAT[5,2,1]

PAT[5,3,1]

and 87 more.

PAT [1,*,*] (tr)

:   1   2   3   4   5   6    :=

1   1   1   0   0   1   1

2   0   1   1   1   1   1

3   1   1   1   1   1   1

 [2,*,*] (tr)

:   1   2   3   4   5   6    :=

1   1   1   0   0   1   1

2   0   0   1   1   1   1

3   1   1   1   1   1   1

 [3,*,*] (tr)

:   1   2   3   4   5   6    :=

1   1   1   0   0   1   1

2   0   1   1   1   1   1

3   1   1   1   0   1   1

;

ampl: 

How can i structure a loop without getting the subscript errors ?

[charles shedrick]

Just a quick update. I figured out a work around. Using the option eexit command I can suppress the bailout and allow my loop to run. I still get the discarded subscript message but I think I will have to totally rework the model to fix those.

--
You received this message because you are subscribed to a topic in the Google Groups "AMPL Modeling Language" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ampl/Wr_KxvKWBpM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ampl+uns...@googlegroups.com.
To post to this group, send email to am...@googlegroups.com.
Visit this group at http://groups.google.com/group/ampl.
For more options, visit https://groups.google.com/d/optout.

[Robert Fourer]

Instead of defining "set AVAIL;", define

set AVAILall;
set AVAIL within AVAILall;

Index your data (sets and parameters) over AVAILall, but index your model (variables, objectives, constraints) over AVAIL. Then you can make all the changes you want like "let AVAIL := pairs[1];" and you won't get any error messages about data being discarded.

Bob Fourer
am...@googlegroups.com

=======

[charles shedrick]

thanks Bob!! that was an even better solution. very clean.  initially I was getting no data in set AVAIL so I defined it similar to set AVAILall in my data file and it appears to work nicely !!.