Closed hagenw closed 7 years ago
I would suggest to remove this flag in the future. As you said, this is more a workaround than a solution. In general we 'only' have to ensure that the input spectrum really is symmetric.
Not a big fan of an extra function though. This suggests it's a different kind of ifft, which in fact it's not at all. As a quick workaround we could name it 'force' symmetric, what it actually is in my opinion.
This is obsolete and was solved by #155
This is a proposal to solve the Octave bug #133.
This is not really a proper solution, but a workaround as it allows Octave for running through without an error by replacing
ifft(...,'symmetric')
byreal(ifft(...))
. I don't know yet if it would be desirable to implement the exact same behavior of the Matlab version asreal()
simple disregards the output and is not checking for conjugate symmetric input.