Added SOP Strategy

[drkupi]

Added the SOP Strategy to the list of strategies in pySOT. Key updates include a new class SOPStrategy, and some supporting functions in the utils.py file.

[codecov[bot]]

Codecov Report

Merging #32 into master will increase coverage by 0.22%. The diff coverage is 93.72%.

@@            Coverage Diff             @@
##           master      #32      +/-   ##
==========================================
+ Coverage    92.3%   92.53%   +0.22%     
==========================================
  Files           8        8              
  Lines        1365     1634     +269     
==========================================
+ Hits         1260     1512     +252     
- Misses        105      122      +17

Impacted Files	Coverage Δ
pySOT/utils.py	92.89% <100%> (+3.9%)	:arrow_up:
pySOT/auxiliary_problems.py	91.78% <100%> (+0.05%)	:arrow_up:
pySOT/strategy.py	89.73% <91.86%> (+1.32%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update aab0f98...c75f842. Read the comment docs.

[drkupi]

Hi David,

I have attempted to resolve all of your specific and general comments. Here is the update on the general comments:

• Code styling has been improved and checked using flake8

• I have tried to remove redundant code snippets.

• The overall code coverage has been improved to 93%.

[dme65]

Hi Taimoor,

The code mostly looks good! I've a few minor comments that shouldn't take long to fix.

@dbindel: Do you think you'll have a chance to review the PR in the next few days so we can merge it?

[drkupi]

I made changes as per your feedback, and updated the pull request.

[drkupi]

All changes are done.

On Thu, Apr 25, 2019 at 5:17 AM David Eriksson notifications@github.com wrote:

@dme65 requested changes on this pull request.

Comments

In pySOT/utils.py https://github.com/dme65/pySOT/pull/32#discussion_r278321260:

• """
• flag = 1
• for i in range(nc):
• if X_c[i, dim+4] == 0:
• sigma_new = sigma
• else:
• sigma_new = np.power(1/2, X_c[i, dim+4])*sigma
• d = scp.distance.euclidean(X, X_c[i, 0:dim])
• if d < sigma_new*d_thresh/np.sqrt(len(X)):
• flag = 0
• break
• return flag

• +class SopRecord():

Can you move this back to strategy.py and rename it _SopRecord? Same for SopCenter

In pySOT/utils.py https://github.com/dme65/pySOT/pull/32#discussion_r278321297:

• while j < ndf_count:
• if domination(F[:, N], F[:, ndf_index[j-1]], M):
• df_index.append(ndf_index[j-1])
• ndf_index.remove(ndf_index[j-1])
• ndf_count -= 1
• elif domination(F[:, ndf_index[j-1]], F[:, N], M):
• df_index.append(N)
• ndf_index.remove(N)
• ndf_count -= 1
• break
• else:
• j += 1
• return (ndf_index, df_index)

• +def ND_Front(F):

Rename to nd_front

In pySOT/utils.py https://github.com/dme65/pySOT/pull/32#discussion_r278321370:

• :type M: int
• :return: Boolean check for domination
• :rtype: Bool
• """
• d = False
• for i in range(0, M):
• if fA[i] > fB[i]:
• d = False
• break
• elif fA[i] < fB[i]:
• d = True
• return d

• +def ND_Add(F, df_index, ndf_index):

Rename to nd_add

In pySOT/utils.py https://github.com/dme65/pySOT/pull/32#discussion_r278321618:

@@ -390,3 +390,182 @@ def optimize(self): population = new_population

     return best_individual, best_value

+ +POSITIVE_INFINITY = float("inf") + +def domination(fA, fB, M):

• """Returns True if a fA dominates fB
• :param fA: Vector A of objective functions
• :param fB: Vector B of objective functions
• :param M: Number of objective functions in A and B
• :return: Boolean
• """

I prefer the more compact and Pythonic version, which will likely be faster as well due to interpreter overhead.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/dme65/pySOT/pull/32#pullrequestreview-230360750, or mute the thread https://github.com/notifications/unsubscribe-auth/AEB7SHFI36SMTJBEXJWAK43PSDE5ZANCNFSM4HFFC4TQ .

[dme65]

Thanks, Taimoor!

LGTM