Implementation of pennylane optimizers

[raulconchello]

Implementation of the classical optimizers of the PennyLane library

The following optimizers have been added:

• AdagradOptimizer
• AdamOptimizer
• GradientDescentOptimizer
• MomentumOptimizer
• NesterovMomentumOptimizer
• RMSPropOptimizer
• RotosolveOptimizer
• SPSAOptimizer.

These have been implemented without changing the existing structure.

In the setup.py, the dependency with PennyLane has been introduced.

Also, a test for these methods is added.

[codecov]

Codecov Report

Merging #101 (9b76633) into dev (03681c2) will decrease coverage by 6.94%.
The diff coverage is 48.29%.

@@            Coverage Diff             @@
##              dev     #101      +/-   ##
==========================================
- Coverage   87.01%   80.07%   -6.95%     
==========================================
  Files          67       93      +26     
  Lines        9006    10969    +1963     
==========================================
+ Hits         7837     8783     +946     
- Misses       1169     2186    +1017     

Impacted Files	Coverage Δ
openqaoa/optimizers/__init__.py	100.00% <ø> (ø)
setup.py	100.00% <ø> (ø)
...enqaoa/optimizers/pennylane/fourier/reconstruct.py	9.33% <9.33%> (ø)
openqaoa/optimizers/pennylane/math/quantum.py	11.15% <11.15%> (ø)
...enqaoa/optimizers/pennylane/math/is_independent.py	13.79% <13.79%> (ø)
...a/optimizers/pennylane/math/matrix_manipulation.py	15.66% <15.66%> (ø)
openqaoa/optimizers/pennylane/math/utils.py	23.40% <23.40%> (ø)
...enqaoa/optimizers/pennylane/math/multi_dispatch.py	36.78% <36.78%> (ø)
...nqaoa/optimizers/pennylane/math/single_dispatch.py	43.06% <43.06%> (ø)
openqaoa/optimizers/pennylane/numpy/random.py	56.52% <56.52%> (ø)
... and 22 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[Q-lds]

@raulconchello you need to import pytest in the notebook_test file, see

openqaoa/tests/test_notebooks.py

Line 7 in 1887f41

import pytest

[Q-lds]

Damn! I almost forgot! We also need to update the documentation:

• Add pennylane as a requirement here https://github.com/raulconchello/openqaoa/blob/dev_PennyLane_optimization_methods/docs/requirements.txt
• Add the classes here: https://github.com/raulconchello/openqaoa/blob/dev_PennyLane_optimization_methods/docs/source/optimizers.rst

[Q-lds]

Another thing that came to mind: we need to make sure that the optimisers work well also from the point of view of the workflows.

The first step is to add them here: https://github.com/raulconchello/openqaoa/blob/115b5766c38767cbe5d8fdf18f5f6f0225bf8012/openqaoa/workflows/parameters/qaoa_parameters.py#L29

(Again, the fact that we are not doing this but that the tests is using the workflows suggests that the tests have not been running correctly)

[shaohenc]

Multiple comments pending response