That is not a bug, since v is an eigenvector iff -v is, even if you
want it normalised. If you mind whether some entries are positive,
just test and negate if necessary.
John
On 12 November 2013 09:41, Evans Doe Ocansey wrote:
> Hi
>
> We're getting different results for an eigenvector in Sage
Hi
We're getting different results for an eigenvector in Sage 5.11 and Sage
5.12. We think 5.11 is the correct answer and 5.12 has flipped the values
to negative.
sage@hummingbird-lan:~$ cat eigenbug.sage
A=matrix(RDF,[[0, 1/2, 1/3, 0, 0], \
[1/2, 0, 0, 1/3, 0], \
[1/2
