Open
Description
Summary
I'm having trouble with Piecewise constraints and the GLPK solver.
I've reproduced the problem with a simple example from Pyomo's github. See below for error message.
Steps to reproduce the issue
# Steps to reproduce
$ glpsol --version
# GLPSOL--GLPK LP/MIP Solver 5.0
$ python piecewise.py
# Error...# piecewise.py
"""Simple Piecewise example.
Adapted from: https://github.com/Pyomo/pyomo/blob/main/examples/pyomo/piecewise/points.py
"""
import pyomo.environ as pyo
solver = "glpk"
x = [0.0, 1.5, 3.0, 5.0]
y = [1.1, -1.1, 2.0, 1.1]
model = pyo.ConcreteModel()
model.x = pyo.Var(bounds=(min(x), max(x)))
model.y = pyo.Var()
model.fx = pyo.Piecewise(model.y, model.x, pw_pts=x, pw_constr_type="EQ", f_rule=y)
model.o = pyo.Objective(expr=model.y)
if __name__ == "__main__":
solver = pyo.SolverFactory(solver)
solver.solve(model).write()Error Message
ERROR: Solver (glpk) returned non-zero return code (1)
ERROR: Solver log: GLPSOL--GLPK LP/MIP Solver 5.0 Parameter(s) specified in
the command line:
--write /tmp/tmp9rrz7ols.glpk.raw --wglp /tmp/tmpeg1_vla_.glpk.glp
--cpxlp /tmp/tmp2n5b9j40.pyomo.lp
Reading problem data from '/tmp/tmp2n5b9j40.pyomo.lp'...
/tmp/tmp2n5b9j40.pyomo.lp:40: invalid bound definition CPLEX LP file
processing error
Traceback (most recent call last):
File "/home/ogaday/Code/Other/pyomo-piecewise-issue/pyomo_piecewise_issue.py", line 36, in <module>
solver.solve(model).write()
File "/home/ogaday/Code/Other/pyomo-piecewise-issue/.venv/lib/python3.10/site-packages/pyomo/opt/base/solvers.py", line 628, in solve
raise ApplicationError("Solver (%s) did not exit normally" % self.name)
pyomo.common.errors.ApplicationError: Solver (glpk) did not exit normallytmp2n5b9j40.pyomo.lp:
\* Source Pyomo model name=unknown *\
min
x1:
+1 x2
s.t.
c_e_x3_:
+1 x4
-1.5 x5
-3.0 x6
-5.0 x7
= 0
c_e_x8_:
+1 x2
-1.1 x9
+1.1 x5
-2.0 x6
-1.1 x7
= 0
c_e_x10_:
+1 x9
+1 x5
+1 x6
+1 x7
= 1
bounds
-inf <= x2 <= +inf
0.0 <= x4 <= 5.0
0 <= x9 <= +inf
0 <= x5 <= +inf
0 <= x6 <= +inf
0 <= x7 <= +inf
SOS
x11: S2::
x9:1
x5:2
x6:3
x7:4
end
Information on your system
Pyomo version: pyomo 6.7.1
Python version: Python 3.10.8
Operating system: Ubuntu 22.04.4 LTS on Windows 10 x86_64 (WSL) / Kernel: 5.15.146.1-microsoft-standard-WSL2
How Pyomo was installed (PyPI, conda, source): PyPI/Poetry
Solver (if applicable): GLPK 5.0
Solver installed via apt:
sudo apt install glpk-doc glpk-utils libglpk-dev python3-swiglpk
# glpk-doc is already the newest version (5.0-1).
# glpk-utils is already the newest version (5.0-1).
# libglpk-dev is already the newest version (5.0-1).
# python3-swiglpk is already the newest version (4.65.1-1build2).Additional information
Changing the solver to cbc, or setting pw_repn="DCC" both seem to solve the issue. Is this behaviour expected? My optimisation knowledge is fairly shallow. If there are some combinations of solvers and arguments that aren't feasible could this be documented?
Thanks in advance!