GPU Support

[ubeydtalha]

Just out of curiosity , I wonder if pymoo gpu supported calculation feature can be added ?
Especially using numba library.
Perhaps parallelization or computations that require a long time can be handled faster in this way.

[blankjul]

Yes, it can! A Problem can be defined to evaluate a whole population. This can be rapidly sped up by using parallelization, for instance using a GPU. (Usually, the problem itself is where most computationally resources need to be located, except for test problems of course).

[onurd86]

Hi,

I tried to put annotation @jit(nopython=True) to _evaluate function of the problem. However, I got the following error:

Traceback (most recent call last):
File "/Users/od/eclipse-workspace_Sept2022/HardeningSystem/simulations/generate_attack_graphs_and_record_hardening_time.py", line 457, in
results_m, results_f = hardening_optimization.find_optimal_hardenings_using_optimization(ag1, default_hardening_dict, hardening_costs = costs, operational_constraints=hardening_constraints, weights_of_h = [0.4,0.6,0.0], gen_out=False,
File "/Users/od/git/AG_Generation/AttackGraphGeneration/hardening_module/hardening_optimization.py", line 811, in find_optimal_hardenings_using_optimization
res = minimize(ag_opt,
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/optimize.py", line 67, in minimize
res = algorithm.run()
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/algorithm.py", line 141, in run
self.next()
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/algorithm.py", line 161, in next
self.evaluator.eval(self.problem, infills, algorithm=self)
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/evaluator.py", line 69, in eval
self._eval(problem, pop[I], evaluate_values_of, **kwargs)
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/evaluator.py", line 90, in _eval
out = problem.evaluate(X, return_values_of=evaluate_values_of, return_as_dictionary=True, **kwargs)
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/problem.py", line 187, in evaluate
_out = self.do(X, return_values_of, *args, **kwargs)
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/problem.py", line 227, in do
self._evaluate_elementwise(X, out, *args, **kwargs)
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/problem.py", line 245, in _evaluate_elementwise
elems = self.elementwise_runner(f, X)
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/problem.py", line 37, in call
return list(self.starmap(f, [[x] for x in X]))
File "/opt/homebrew/Cellar/python@3.10/3.10.10_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/multiprocessing/pool.py", line 375, in starmap
return self._map_async(func, iterable, starmapstar, chunksize).get()
File "/opt/homebrew/Cellar/python@3.10/3.10.10_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/multiprocessing/pool.py", line 774, in get
raise self._value
File "/opt/homebrew/Cellar/python@3.10/3.10.10_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/multiprocessing/pool.py", line 125, in worker
result = (True, func(*args, **kwds))
File "/opt/homebrew/Cellar/python@3.10/3.10.10_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/multiprocessing/pool.py", line 51, in starmapstar
return list(itertools.starmap(args[0], args[1]))
File "/opt/homebrew/lib/python3.10/site-packages/pymoo/core/problem.py", line 20, in call
self.problem._evaluate(x, out, *self.args, **self.kwargs)
TypeError: not enough arguments: expected 5, got 4

[blankjul]

this might be helpful: https://stackoverflow.com/questions/41769100/how-do-i-use-numba-on-a-member-function-of-a-class