ORIE 3300/5300
Fall 2011
Prof. Bland
Problem Set 10
Due 11/22/2011, 11:40 a.m.
Submit hardcopy in Hollister B14 at the beginning of lecture on 11/22
. Only 1 sub-
mission from each group.
Problem 1 is a continuation of Recitation Exercise 10. Problem 2 will be continued
as Recitation Exercise 11.
1. Finish recitation Exercise 10. For each of the two approaches, submit a summary
of the optimal values of the decision variables, yield and risk as functions of the
relevant parameter.
Show your analysis beginning from the point where the
recitation solutions ended.
2. Twenty-five students will each be taking one of ten different tutorial courses
next semester. Each tutorial will have at most three students. The file
Prefer-
ences.txt
shows for each student a ranking of her first through fifth choices. An
entry of 6 in row i and column j indicates that course i was not ranked in the
top 5 by student j; it is not really student j’s sixth choice, it is one of her 20
unranked courses.
As explained briefly in class on 11/17, this looks like a transportation problem,
except we have not yet established an objective function. For now, we are going
to assign a cost (penalty) of
k
for assigning student j to her kth most preferred
course i and minimize the total cost. (This is NOT a good choice of a cost func-
tion, but we will use it for now, and return to the choice of objective function
coefficients later.)
CPlex solver.
You will want to use the CPlex solver (on all parts of this problem and future
variations on this problem)– use the
option solver cplex
command.
The print command and conditional indexing
.
See AMPL book pp.238-239. The AMPL command
display
is somewhat lim-
ited.
The
print
command functions like the display command.
Among the
features that make the print command useful is it takes conditional indexing –
e.g., the command
This
preview
has intentionally blurred sections.
Sign up to view the full version.
print
{
i in ORIG, j in DEST: cost[i,j] = 0
}
:
i, j;
would print a line of output only for those choices of
i
and
j
in the appropri-
ate sets which satisfy the condition after the colon – i.e., the cost of assigning
destination
j
to origin
i
is zero. Conditioning on the decision variables could
be pretty handy in reducing and organizing your output. In each part, use the
print command to display a list of all courses and the students assigned to them.
If you would like to have tab separated entries on each line (recommended), see
the end of the next-to-last paragraph on p. 238 of the AMPL book. This will
make your output a little prettier and a little easier for the eye to scan, and will
also make it easier to import into a spreadsheet, if you wish to do that.
This is the end of the preview.
Sign up
to
access the rest of the document.
