QiskitDevice.probabilities() changed to QiskitDevice.probability() to…

[rafaelha]

… restore compatibility with pennylane python package

[antalszava]

Hi @rafaelha, thank you so much for your pull request! 🎊

Adjusting the name of the QiskitDevice.probabilities method is a change that is upcoming in this plugin as we are porting to the use of QubitDevice from PennyLane.

I'd be curious, was there perhaps a bug or any problems that you have faced while using QiskitDevice.probabilities?

[rafaelha]

Hi @rafaelha, thank you so much for your pull request! 🎊

Adjusting the name of the QiskitDevice.probabilities method is a change that is upcoming in this plugin as we are porting to the use of QubitDevice from PennyLane.

I'd be curious, was there perhaps a bug or any problems that you have faced while using QiskitDevice.probabilities?

Hi, yes, running a qnode on a qiskit device that returns qml.probs raises NotImplementedError: Returning probability not currently supported by qiskit.ibmq. This seems to be because pennylane calls the function Device.probability (https://github.com/XanaduAI/pennylane/blob/a776dbecfb0c1d32bb04012a69ee0e559d58c3b5/pennylane/_device.py) instead of QiskitDevice.probabilities. Renaming QiskitDevice.probabilities to QiskitDevice.probability fixed that error for me. Cheers!

[antalszava]

I see that makes sense! Thank you so much once again for your addition! 😊

The tests are failing due to the test suite pulling the newest version (0.18) of Qiskit. We are investigating this right now and creating a fitting solution.

[antalszava]

Thanks so much again for this change @rafaelha! 💯 😊

Opened two related pull requests:

• Qiskit v0.18 updates #81: includes changes due to the new version of Qiskit and allows passing the test suite on Travis
• IBMQ probs test #82: includes a test for the failing case that you have identified

Checked locally, the test from #82 picks up the case that was mentioned and the changes from this PR made it pass for me locally!

[josh146]

Thanks @rafaelha! @antalszava, it looks like we are ready to merge in #80, #81, and #82. Do we want to update the changelog separately, since these three PRs are all somewhat related?

[codecov]

Codecov Report

Merging #80 into master will decrease coverage by 0.58%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master      #80      +/-   ##
==========================================
- Coverage   99.12%   98.53%   -0.59%     
==========================================
  Files           7        7              
  Lines         341      341              
==========================================
- Hits          338      336       -2     
- Misses          3        5       +2     

Impacted Files	Coverage Δ
pennylane_qiskit/qiskit_device.py	98.63% <100.00%> (ø)
pennylane_qiskit/ibmq.py	91.30% <0.00%> (-8.70%)	⬇️

---

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 dfba326...2887b4c. Read the comment docs.

[antalszava]

Thanks for catching that @josh146!

@rafaelha, could you please modify the CHANGELOG.md file such that it includes:

• a one-sentence description of this PR (for example in the Bugfix section);
• yourself as a contributor in the list of contributors for the current release (# Release 0.9.0-dev).

Thanks! 😊

[josh146]

Thanks @rafaelha, updated changelog looks perfect!