solve_flat_ocp and scalar timepts #1110
Description
flowing from gh-1101, which complains about the code in question,
solve_flat_ocp argument timepts is advertised in the docstring as being allowed to be scalar:
timepts : float or 1D array_like
The list of points for evaluating cost and constraints, as well as
the time horizon. If given as a float, indicates the final time for
the trajectory (corresponding to xf)
but this doesn't work:
import numpy as np
import control as ct
import control.flatsys as fs
f = fs.LinearFlatSystem(ct.ss(ct.tf([1],[1,1])))
def terminal_cost(x, u):
return (x-5).dot(x-5)+u.dot(u)
traj1 = fs.solve_flat_ocp(f, [0, 1], x0=[23],
terminal_cost=terminal_cost)
traj2 = fs.solve_flat_ocp(f, 1, x0=[23],
terminal_cost=terminal_cost)
teval = np.linspace(0,1,101)
r1 = traj1.response(teval)
r2 = traj2.response(teval)
assert np.max(abs(r1.x-r2.x)) == 0
assert np.max(abs(r1.u-r2.u)) == 0
assert np.max(abs(r1.y-r2.y)) == 0produces with master @ ebff125
Traceback (most recent call last):
File "/home/rory/projects/pycontrol/pyflakes/flatbug2.py", line 13, in <module>
traj2 = fs.solve_flat_ocp(f, 1, x0=[23],
terminal_cost=terminal_cost)
File "/home/rory/src/python-control/control/flatsys/flatsys.py", line 725, in solve_flat_ocp
T0 = timepts[0] if len(timepts) > 1 else T0
^^
UnboundLocalError: cannot access local variable 'T0' where it is not associated with a value
I think the fix is as below. Diff also removes Tf, which is not used in the function. With this diff the test script runs with error.
@@ -721,8 +721,7 @@ def solve_flat_ocp(
# Process final time
timepts = np.atleast_1d(timepts)
- Tf = timepts[-1]
- T0 = timepts[0] if len(timepts) > 1 else T0
+ T0 = timepts[0] if len(timepts) > 1 else 0
# Process keyword arguments
if trajectory_constraints is None: