Describe the bug
This should not work (to be updated), and would not work as the final time passed, t1, is not equal to the actual variable, tf. (Works here since the variable is actually not used in the dynamics.)
https://control-toolbox.org/OptimalControl.jl/stable/example-double-integrator-time.html#Indirect-method
To Reproduce
See above
Expected behavior
- passing time(s) must be compulsory and explicit to avoid confusing time(s) used to evaluate that can be different from the actual variable(s)
- new syntax: when there is a variable, the flow generated by
Flow must include a named argument (without default so that omitting it raises an error):
xf, pf = f(t0, x0, p0, t1; variable=tf)
Remark It is OK (= not an error) to allow t1 to be different from tf
See also
Also update this part ot the doc:
https://control-toolbox.org/OptimalControl.jl/stable/manual-flow-ocp.html
Describe the bug
This should not work (to be updated), and would not work as the final time passed,
t1, is not equal to the actual variable,tf. (Works here since the variable is actually not used in the dynamics.)https://control-toolbox.org/OptimalControl.jl/stable/example-double-integrator-time.html#Indirect-method
To Reproduce
See above
Expected behavior
Flowmust include a named argument (without default so that omitting it raises an error):Remark It is OK (= not an error) to allow
t1to be different fromtfSee also
Also update this part ot the doc:
https://control-toolbox.org/OptimalControl.jl/stable/manual-flow-ocp.html