Implement Negate (two's complement)
#1144
Conversation
| # TODO use QInt once classical sim is fixed. | ||
| # classical sim currently only supports unsigned. |
There was a problem hiding this comment.
Is this true? I thought classical simulation supports both signed and unsigned integers. Is there an open issue you are referring to?
There was a problem hiding this comment.
Thanks, looks like it's bottlenecked by speed and #1142 should help fix it?
| x: Any signed value stored in two's complement form. | ||
| """ | ||
|
|
||
| dtype: QDType |
There was a problem hiding this comment.
Should we assert that dtype is a signed type ? So QInt or QFxp(signed=True) ?
There was a problem hiding this comment.
It actually works for unsigned types as well, it's the unary minus operator in c++. Perhaps we can use the c++ naming convention?
There was a problem hiding this comment.
Thanks, we should at least have a link explaining the behavior for unsigned types in the docstring. I don't have strong opinions on naming convention.
There was a problem hiding this comment.
updated docstring to explain this
There was a problem hiding this comment.
will leave the Bloq as Negate for now
|
|
||
|
|
||
| @frozen | ||
| class Negate(Bloq): |
There was a problem hiding this comment.
I think it's also worth adding a Not(Bloq) in arithmetic/bitwise.py which simply computes the bitwise NOT via OnEach(self.dtype.num_qubits, XGate()) and is a natural choice for unsigned types
There was a problem hiding this comment.
yep this is useful! will add in a follow-up PR and cleanup this decomp then.
No description provided.