Julien Gacon

If I use just a UCCSD ansatz and plug that into the VQE the HF initial state is not automatically added, is it? In that scenario it seems to me...

If we would remove `preferred_init_points` then the expectation would never be that VQE/UCC sets the initial points for you, so no matter whether you transpile or not -- you have...

Could we put a that information in the deprecation warning in the circuits and say that if used for a `GroundStateSolver` or so one should use `InitialPoint`?

This problem should be resolved, there was an issue with the runtime programs in general. 🙂 @lasys could you check if the code runs for you again?

There's the option to reconnect the callback after the stream broke with `job.stream_results(callback)` though that's something the `VQEClient` would need to take care of since we don't expose the `job`....

Hi @alonso-pena, the "exact value" in the tutorial references to the exact expectation value that QAE approximates. It does not correspond to the true/analytic value of the european call option....

That would be another option, but detecting that the stream broke and automatically reconnecting is probably more difficult to implement. Also, we could think of just allowing a flag to...

Is this still active? It would be a much appreciated feature! 🚀

That's probably because `"{0, 100, 0, 40}"` is passed as string to matplotlib but it expects an array or list. That's another case of #184: currently it is only possible...

You need to uncomment a few lines, as noted in the code here: https://github.com/lava/matplotlib-cpp/blob/60dcd64c8fd4766e5426f57decfb765422a1d3fe/matplotlibcpp.h#L303 (I'm running on Catalina, too)