pymoo

pymoo copied to clipboard

SBX crossover

2

Hello, if I compare your CBX algorithm with the one given in [1] (equation (5), (6) and (7) and the following paragraph), and if I didn't make some mistakes in...

KCrux

Hello, I wrote some pymoo `Callback`s and store the data into `algorithm.data`, and I found that finally I could not retrieve it from the `Result` object. So I think in...

jacktang

RVEA not reaching specified number of evaluations

6

**RVEA does not complete the specified number of evaluations set by the termination condition.** I've run RVEA on several problems, both those implemented in pymoo and those implemented elsewhere, and...

jordan-cork

AssertionError: This implementation of MOEAD does not support any constraints.

1

AssertionError: This implementation of MOEAD does not support any constraints. hello，How to deal with this？ In pymoo, what algorithm can be selected to solve multi-objective optimization (with constraints), other than...

niurencheng

Encountered problems while using pymoo parallelization, hoping to seek help

2

``` class MyProblem(Problem): def __init__(self): super().__init__(n_var=13, # x n_obj=2, # n_ieq_constr=4, # n_eq_constr=1, # xl=np.array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]), # xu=np.array([6.63, 3.18,...

877040623

Is fast_best_order_sort intended to be usable by NonDominatedSorting?

1

#495 has some benchmarks that seem to show good speed for `fast_best_order_sort` from [the cython code](https://github.com/anyoptimization/pymoo/blob/main/pymoo/cython/non_dominated_sorting.pyx#L32), but it doesn't seem to be importable [here](https://github.com/anyoptimization/pymoo/blob/main/pymoo/util/function_loader.py#L22).

JonathanRaynerMonumo

resolve bug that prevents two equally seeded reference direction calculations (energy reduction method) to produce the same output

1

Test with: __________________________________________ import numpy as np from pymoo.util.ref_dirs import get_reference_directions a = get_reference_directions("energy", n_dim=3, n_points=100, seed=2) b = get_reference_directions("energy", n_dim=3, n_points=100, seed=2) print(np.array_equal(a, b)) ___________________________________________ Behavior before change: Expected...

NikHoh

As the extreme point calculation is happening as a weighted and aggregated optimization (max(__F * weights)), I think it would be better to normalize __F beforehand. Otherwise, the resulting extreme...

NikHoh

Based on the discussion [here](https://github.com/anyoptimization/pymoo/issues/537) this PR rewrites `JoblibParallelization` to improve reliability and consistency of using `joblib`.

cheginit

Issue with `JoblibParallelization`

2

The current implementation of `JoblibParallelization` that uses an already instantiated `Parallel` can lead to some issues such as [this](https://github.com/joblib/joblib/issues/1243) one. A better approach would be to take as input all...

cheginit