Open magv opened 4 years ago
This is not completely trivial to fix, because d in a rather special object. It would need quite special code. The easiest work around is to use (d(idx1,idx2))^2.
On 9 Jan 2020, at 14:22, Vitaly Magerya notifications@github.com wrote:
I idx1, idx2; L EXPR = d_(idx1, idx2)^2; .sort .end
On 09/01/2020 15:01, Jos Vermaseren wrote:
The easiest work around is to use (d_(idx1,idx2))^2.
That's a much better workaround. Thanks.
Hi, folks. I get an "Internal error in code generator" when d_() squared appears; here's a short snippet:
Here's the result of running it:
Changing
d_(idx1, idx2)^2
intod_(idx1, idx2)*d_(idx1, idx2)
produces the correct4
as the result.For the moment one can work around this by using e.g.
delta
in stead ofd_
in the input and then running an additionalid
, but would it be possible to fix the internal error instead?