[Draft] Expansion of Observable_config #664
Conversation
|
Please rebase the branch on main. A previous PR of mine modified the |
smitchaudhary
changed the title
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #664 +/- ##
==========================================
- Coverage 85.79% 85.78% -0.02%
==========================================
Files 197 197
Lines 16971 16984 +13
Branches 1452 1453 +1
==========================================
+ Hits 14561 14570 +9
- Misses 1994 1998 +4
Partials 416 416 ☔ View full report in Codecov by Sentry. |
|
Thanks. Not very comprehensive, but some of the things here are being discussed in the PR here where the configs were introduced. Might give some insight into why certain things were done. (Having said that, feel free to just ask instead of going on a wild chase). |
qadence/constructors/hamiltonians.py
Outdated
| def ising_hamiltonian_config( | ||
| x_scale: TParameter = 1.0, | ||
| x_shift: TParameter = 0.0, | ||
| x_transformation_type: ObservableTransform = ObservableTransform.NONE, # type: ignore[assignment] | ||
| zz_scale: TParameter = 1.0, | ||
| zz_shift: TParameter = 0.0, | ||
| zz_transformation_type: ObservableTransform = ObservableTransform.NONE, # type: ignore[assignment] | ||
| trainable_transform: bool | None = None, | ||
| ) -> List[ObservableConfig]: | ||
| zz_ham = zz_hamiltonian_config( | ||
| scale=zz_scale, | ||
| shift=zz_shift, | ||
| transformation_type=zz_transformation_type, | ||
| trainable_transform=trainable_transform, | ||
| ) | ||
| x_ham = ObservableConfig( | ||
| detuning=X, | ||
| scale=x_scale, | ||
| shift=x_shift, | ||
| transformation_type=x_transformation_type, | ||
| trainable_transform=trainable_transform, | ||
| ) | ||
| return [zz_ham, x_ham] |
There was a problem hiding this comment.
I think this can be changed to return a single config that when the user passes to QNN.from_configs gives you the Ising Hamiltonian as the observable instead of giving us a list of 2 ObservableConfigs, which would result in the QNN giving 2 outputs, one for the ZZ as observable, and the other as X as observable. See ising_hamiltonian.
No description provided.