Open saraedum opened 11 years ago
Attachment: trac_13651.patch.gz
There was a problem in _internal_lshift()
. I rewrote it to use multiplications of the form x^e/p
which can be done without a precision loss.
It initially took me a while to understand what _ntl_rep_abs()
does. I reformatted it heavily in this patch but I (hopefully) did not change what it does at all. I hope it's easier to understand now.
The doctests in the padics directory pass. Let's see what the patchbot thinks about everything else.
This also fixes a todo
in the code:
sage: ZZ(W(5)) # todo: this should be different...
381469726562505
It seems that roed wrote that todo a few years ago. (added him as CC)
Branch: u/saraedum/ticket/13651
Changed keywords from none to sd59
looks good but branch doesn't apply currently
New commits:
b87ff51 | Trac #13651: fixed a precision problem in _internal_lshift(), _ntl_rep_abs() of padics |
(changed the status so it shows up in my list of tickets needing work)
Branch pushed to git repo; I updated commit sha1. New commits:
c3875b9 | Merge branch 'develop' into t/13651/ticket/13651 |
Changed keywords from sd59 to sd59, days71
Currently,
_ntl_rep_abs()
is incorrect for some extension fields::The correct result here would be
([3],0)
.The attached patch fixes this.
CC: @roed314
Component: padics
Keywords: sd59, days71
Author: Julian Rueth
Branch/Commit: u/saraedum/ticket/13651 @
c3875b9
Issue created by migration from https://trac.sagemath.org/ticket/13651