qiskit-ignis
qiskit-ignis copied to clipboard
Missing tutorials and documentation for ignis
It seems that some of ignis tutorials were removed from: https://github.com/Qiskit/qiskit-tutorials/tree/master/tutorials/noise and should be replaced by new tutorials or enhanced API documentation. Some of them still appear here: https://github.com/Qiskit/qiskit-tutorials/tree/master/legacy_tutorials/ignis
This includes:
- [ ] Advanced RB methods (interleaved, purity, cnot-dihedral) - https://github.com/Qiskit/qiskit-tutorials/pull/862, https://github.com/Qiskit/qiskit/pull/866, https://github.com/Qiskit/qiskit-tutorials/pull/885 - should be API docstring (not tutorials)
- [x] Tomography (QST, QPT, GST)
- [ ] Discriminator - https://github.com/Qiskit/qiskit-tutorials/pull/861
- [x] Entanglement - https://github.com/Qiskit/qiskit/pull/878, https://github.com/Qiskit/qiskit-ignis/pull/328/commits/601c1a42b3f96b954954b9e968587eaec55646a1
- [x] Logging - should be API docstring (not tutorial)
The tomography tutorial was removed due to non-essential bugs (e.g. not adding validate=False
to state_fidelity
calls). I opened a PR with an updated version:
https://github.com/Qiskit/qiskit-tutorials/pull/973
However, recently I had a lot of problems with cvxpy
(the optimization library that tomography uses) - it complains about missing linkage with two linear algebra libraries, blas
and lapack
. While I managed to install them correctly on Linux, everything I tried on Windows failed. We should take this into account.
Ok, now I see that indeed, building the tutorial failed because of this problem. This is a bigger issue than simply a "tutorial not working" problem.
We now have the tutorials for tomography and entanglement: https://qiskit.org/documentation/tutorials/noise/index.html
As for the discriminator - see this issue: https://github.com/Qiskit/qiskit-tutorials/issues/952 It seems that the discrimination tutorial needs to call the IQX in order to execute.
For logging, the tutorial is not needed, since there are API docs. For advanced RB methods, the various tutorials will be combined as part of the API docs.