I know that $QFT_n|0\rangle$ is equivalent to $H_n|0\rangle$ (mathematical proof).

And it is also easy to prove that $QFT_1$ is equivalent to $H_1$ (applied to one QuBit).

From looking at the circuit below it seems clear to me that the gates should also be equivalent if $|x_1\rangle$ is in any state and all other QuBits are $|0\rangle$. This should be true because none of the controlled $R$ gates are applied to $|x_1\rangle$.

I do not know how to prove this mathematically. Can anyone provide an elegant proof?

from wikipedia 3

  • $\begingroup$ No my question is slightly different. I ask for a prove that x_1 qubit can be in any state, and qft is still equivalent to walsh. $\endgroup$
    – birneee
    Aug 1 '20 at 11:39
  • $\begingroup$ "This should be true because none of the controlled R gates are applied" is a mathematical proof (if you replace "should be" by "is"). $\endgroup$ Aug 2 '20 at 12:58

I think your explanation based on the circuit is perfectly adequate.

For a more rigorous "proof", why not simply take the output of the circuit? Substitute in $x_i=0$ for all $i$ and see that all the outputs are $(|0\rangle+|1\rangle)/\sqrt{2}$ for $i\neq 1$ and $(|0\rangle+(-1)^{x_1}|1\rangle)/\sqrt{2}$ for $i=1$, exactly as it would be for the Hadamard transform.

Alternatively, simply repeat the proof in the answer you cite for two inputs: $|0\rangle^{\otimes n}$ and $|1\rangle|0\rangle^{\otimes(n-1)}$. If it works for those two, by linearity it must work for any input on the first qubit.


Not the answer you're looking for? Browse other questions tagged or ask your own question.