Finding the state of the system after Oracle. about grove HOT 6 CLOSED

@rasa97 Thanks for the issue report! We just refactored the Deutsch-Josza module significantly. Would you mind looking at it again and see if this is still an issue?

from grove.

The refactored one works for single qubit. However, when I run the code for 2-qubits, there are errors. For balanced, I get:

The code is print dj.is_constant(qvm, {'00':'0', '01':'1', '10':'0', '11':'1'})

line 147, in unitary_function
value = int(mappings[bitstring])
KeyError: '0'

For constant, I get the error:

line 59, in is_constant
     constant = all([bit == 0 for bit in bitstring])
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

The code is print dj.is_constant(qvm, {'00':'0', '01':'0', '10':'0', '11':'0'})

from grove.

@rasa97 Thank you for the followup! I'm opening up a PR to fix this!

from grove.

I'm curious to know why the inputs should be represented in a dict. Isn't list enough? Its easier that way, right?

from grove.

@rasa97 I agree that a list has less overhead, but it also makes implicit assumptions about ordering. Moreover, the structure of a dict highlights the structure of the problem and makes it more apparent. In the end you will have to provide 2**n inputs, so an additional factor of 2 will not make too much of a difference.

What's your thought on this?

from grove.

For the sake of convenience, list is better I guess. If the user has an idea about the ordering, it would be easier to just type in the bits. The pattern in which it should be typed can be highlighted in docs.

from grove.