Is there a way to obtain data from .get_log(), while at the same time suppressing stdout from .evolve()?

[caranha]

Hello!

I am trying to run some of the algorithms from pygmo a few times, and then produce runtime analysis graphs from those runs. For that, I use the information from .get_log().

However, I do not need the stdout output from .evolve() for all those runs, and I'd like to know if it is possible to suppress that while still getting access to the data from .get_log(). If I set_verbosity() to zero, I lose both std_out and get_log. I couldn't find anything that solved my problem in the documentation...

(I'd prefer not to suppress stdout at the terminal level, because I'm planning to use tqdm to show a progress bar as I go through the algorithm/problem pairs.)

Any ideas?

Thank you!

# Setup the run
algo = pg.algorithm(uda)
algo.set_verbosity(10)
pop = pg.population(prob, size = popsize)

pop = algo.evolve(pop)

# Extract the data
log = algo.extract(type(uda)).get_log()