Decc

- Add new parameter `ensure_evaluations` to DECC;
- This parameter allows to stop the optimization process mid-cycle (previously, the processes only checked its condition after a cycle had been finished);
- Add new return information: number of FEs used to obtain the best solution;

- Add clipping to DECC-G;
- Ensure that all values are within the bounds provided for the function;
- Apply clip after multiplication by weights;

Minor refactor: use `np.squeeze(...)` instead of accessing index 0 of best_fitness in DECC and DECC-G optimizers;

Add the package to PyPI with correct classifiers and URLs: https://pypi.org/project/decc/

First release, contains two optimizer:

- DECC, with DECC-O and DECC-H variants;
- DECC-G;