Subprocess.check_call() about task-models HOT 3 CLOSED

Hi @SimonNgj. By default, the software expects the pomdp-solve binary to be located in your $PATH.

If you used the commands from the README you need to make sure that ~/.local is in your $PATH.

Alternatively it's also possible to use the binary in the sample directory but you need to add the sample directory to your $PATH,

export PATH="$HOME/MyTest/task-models/samples/:$PATH"

In addition, I'll add a more meaningful error message when the solver binary cannot be found.

[EDIT] My bad, actually, there is already a check for the executable being found, so it seems that the issue is something else. Can you add the following line after line 16 in icra_scenario2pomdp.py, run it again, and post the output?

print('Solver path: {}'.format(p._solver_path))

from task-models.

My problem is solved now. Thank you so much @omangin. You're right. I just add
export PATH="$HOME/MyTest/task-models/samples/:$PATH"
before running: python icra_scenario2pomdp.py

Output: ....

Epoch: 498...7 vectors in 0.00 secs. (1.27 total) (err=1.00e-01)
Epoch: 499...7 vectors in 0.01 secs. (1.28 total) (err=1.00e-01)
Epoch: 500...7 vectors in 0.00 secs. (1.28 total) (err=1.00e-01)
++++++++++++++++++++++++++++++++++++++++
Solution found.  See file:
	tosolve.alpha
	tosolve.pg
++++++++++++++++++++++++++++++++++++++++
User time = 0 hrs., 0 mins, 1.23 secs. (= 1.23 secs)
System time = 0 hrs., 0 mins, 0.05 secs. (= 0.05 secs)
Total execution time = 0 hrs., 0 mins, 1.28 secs. (= 1.28 secs)

However, I don't know where do tosolve.alpha and tosolve.pg locate!!!

from task-models.

To answer your last question, they are generated in a temporary directory and deleted after being loaded by the code. You can check that in the source code of the solve method and eventually modify it if you need the files to be kept elsewhere.

from task-models.