176 descriptor set optimizations under constrained knn query is inefficient #220
Conversation
…descriptor use new ID property names
|
Target CPP Coverage: 63.9716% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 63.9716% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 63.9716% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 63.9716% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
| @@ -785,23 +784,25 @@ def test_findDescByBlobWithLink(self): | |||
| all_queries.append(query) | |||
|
|
|||
| response, blob_array = db.query(all_queries, [descriptor_blob]) | |||
| self.assertEqual(response[0]["status"], -1) | |||
There was a problem hiding this comment.
Do we still need test_findDescByBlobWithLink since we no longer have capability to link the knn search using _ref? This is basically just a AddDescriptor/AddEntity test now. Or should we change KNN search to search by constraints to leverage defined _ref?
There was a problem hiding this comment.
Good question, left it in for the moment for us to discuss. It does technically exercise a slightly different path above when its doing its setup, but the search itself is mostly a repeat of the early KNN test.
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
|
Target CPP Coverage: 64.0287% Target Python Coverage: 98.02% |
No description provided.