shall we include quantum Fourier Trafo in package?

Question

shall we include quantum Fourier Trafo in package?

urbach opened this issue 4 years ago · comments

see vignettes/qft.Rmd

Johann Ostmeyer · Answer 1 · Fri Sep 18 2020 17:58:34 GMT+0800 (China Standard Time)

Very nice that it works for the general case now!
Yes, I would say that we should include it. It makes the package much more powerful for external users should we publish it. And I don't think it harms much that the students can see an R-implementation and not only pseudo-code. The interesting part will be to implement something like breaking RSA that requires the qft but where the qft itself is not the challenging part.

Carsten Urbach · Answer 2 · Fri Sep 18 2020 19:00:50 GMT+0800 (China Standard Time)

Very nice that it works for the general case now! Yes, I would say that we should include it. It makes the package much more powerful for external users should we publish it. And I don't think it harms much that the students can see an R-implementation and not only pseudo-code. The interesting part will be to implement something like breaking RSA that requires the qft but where the qft itself is not the challenging part.

okay, agreed. I have an even more general version, see phase_estimation.Rmd, which works on a sub-set of the qubits. The phase estimation doesn't work yet, though...

Johann Ostmeyer · Answer 3 · Sat Sep 19 2020 01:02:51 GMT+0800 (China Standard Time)

It does now.

Carsten Urbach · Answer 4 · Sat Sep 19 2020 03:25:03 GMT+0800 (China Standard Time)

It does now.

Of course, the 2 pi... The order I had also realised already. thanks. it does not work for general alpha, though, because then we don't get a product state, still a bit confused.

Johann Ostmeyer · Answer 5 · Sat Sep 19 2020 19:42:09 GMT+0800 (China Standard Time)

It did. 4 digits just don't allow precise results because only two of them are really reliable. I added a few comments.

Carsten Urbach · Answer 6 · Sat Sep 19 2020 21:45:53 GMT+0800 (China Standard Time)

Let's discuss that in person... -- CU Mobil, www.carsten-urbach.eu

Carsten Urbach · Answer 7 · Thu Sep 24 2020 03:48:38 GMT+0800 (China Standard Time)

I think the measurement was still wrong. Now it should be the correct projective measurement. Maybe we also need a version of `measure` which measures a range of qubits. Currently its only either a single one or the whole state.