Project

General

Profile

John Abbott's activity

From 03 Jun 2018 to 02 Jul 2018

02 Jul 2018

18:42 CoCoALib 0.50 hour (Feature #1198 (New): Non-standard DegRevLex (NonStdDegRevLex, WDegRevLex))
Some thoughts John Abbott
18:42 CoCoALib Feature #1198: Non-standard DegRevLex (NonStdDegRevLex, WDegRevLex)
Here are some thoughts about the design/impl.
If the GradingDim is 1 then we have a vector of positive "weights": ...
John Abbott

28 Jun 2018

11:00 CoCoALib 1.50 hour (Bug #1199 (Closed): GCD bug with high degree arg)
Looking for simpler example John Abbott
11:00 CoCoALib Bug #1199 (In Progress): GCD bug with high degree arg
I wrote a program to try all primes in increasing order. None up to about 12000. Then restarted from 30000, and fou... John Abbott
10:54 CoCoALib Bug #1199 (Closed): GCD bug with high degree arg
... John Abbott

27 Jun 2018

15:08 CoCoALib Feature #1198 (New): Non-standard DegRevLex (NonStdDegRevLex, WDegRevLex)
There is a family of cases where a "DegRevLex" ordering with "non-standard" weights is useful.
This can be implement...
John Abbott

26 Jun 2018

15:25 CoCoALib Feature #82: C++11 compatibility questions
I read somewhere on the internet that it is probably better to jump straight to C++14 rather than to C++11; the argum... John Abbott
15:25 CoCoALib 0.25 hour (Feature #82 (Closed): C++11 compatibility questions)
Better C++14 John Abbott
14:40 CoCoALib Feature #1197 (In Progress): IsZeroDet: new fn
Implement @IsZeroDet@ which should allow a fast modular impl. John Abbott
14:38 CoCoALib 0.25 hour (Slug #691 (Closed): Matrix determinant over ZZ)
Must port the old code John Abbott
14:38 CoCoALib Slug #691: Matrix determinant over ZZ
CoCoALib currently uses DetByCRT rather than the "clever" algorithms used in CoCoA-4.7. This issue should be about p... John Abbott
14:32 CoCoALib Slug #1110 (Feedback): Determinant of matrix over QQ (whose entries are actually integers)
Done as part of work on Habilitationsschrift.
Now both dets take about 1.8s
John Abbott
14:32 CoCoALib 20.00 hours (Slug #1110 (Closed): Determinant of matrix over QQ (whose entries are actually integers))
Already done John Abbott
11:51 CoCoALib Design #1124: Move examples directory into doc -- or just change makefile?
One problem is that the @Makefile@ in the @examples/@ directory states explicitly that all executables depend on @CoC... John Abbott
11:51 CoCoALib 0.25 hour (Design #1124 (Closed): Move examples directory into doc -- or just change makefile?)
Another idea John Abbott
11:31 CoCoALib Support #1196 (Closed): Split MatrixOps
Consider splitting the file @MatrixOps.C@ into subfiles with the same prefix. Currently it is over 2000 lines long.
John Abbott
11:29 CoCoALib 0.20 hour (Slug #1187 (New): Matrix rank is slow (over QQ))
Slow mod p too John Abbott
11:29 CoCoALib Slug #1187: Matrix rank is slow (over QQ)
Rank is slower than det even for matrices over a small finite field.
For example a random 500x500 matrix: det takes ...
John Abbott

25 Jun 2018

15:28 CoCoALib Feature #4 (In Progress): Squarefree GCD-free basis
Is this not subsumed by issue #259?
The only remaining question is the "squarefree" part. We could just make the ...
John Abbott
15:28 CoCoALib 0.20 hour (Feature #4 (Rejected): Squarefree GCD-free basis)
Subsumed? John Abbott
15:25 CoCoALib Feature #259: Squarefree(?) GCD-free basis
I have checked in the code. There is doc, but no tests.
I am not happy with the class names: *@GCDFreeBasis_BigIn...
John Abbott
15:25 CoCoALib 0.25 hour (Feature #259 (Closed): Squarefree(?) GCD-free basis)
Bad names John Abbott
15:06 CoCoALib 0.15 hour (Design #1159 (Closed): Add global enum "verify/DontVerify")
Subsumed John Abbott
15:06 CoCoALib Design #1159 (Closed): Add global enum "verify/DontVerify"
This issue has been subsumed by #1167. Closing! John Abbott
14:37 CoCoALib Design #1182 (Feedback): "mod" for BigInt
Anna has not complained, so moving to *@Feedback@*.
John Abbott
14:37 CoCoALib 0.10 hour (Design #1182 (Closed): "mod" for BigInt)
Apparently almost done John Abbott
14:10 CoCoALib Feature #1154 (Feedback): SmallFpImpl: new ctor arg to say do-not-check-that-arg-is-prime
I think that @SmallPrime@ solves this matter reasonably well. It does require making 2 ctors (copy-and-paste), but t... John Abbott
14:10 CoCoALib 1.00 hour (Feature #1154 (Closed): SmallFpImpl: new ctor arg to say do-not-check-that-arg-is-prime)
Resolved John Abbott

21 Jun 2018

22:39 CoCoALib 8.00 hours (Feature #259 (Closed): Squarefree(?) GCD-free basis)
Mostly done? John Abbott
22:39 CoCoALib Feature #259 (Resolved): Squarefree(?) GCD-free basis
I have impls of GCDFreeBasis for RingElem and for BigInt.
Not yet checked in/ No tests; one simple example.
John Abbott
15:39 CoCoALib Support #1193: Naming convention mistakes
In *@BigIntOps@* there are @quorem@ and @divexact@ (names "inherited" from GMP); but there is also @QuoRem@ in anothe... John Abbott
15:39 CoCoALib 0.25 hour (Support #1193 (New): Naming convention mistakes)
Some notes John Abbott
15:38 CoCoALib Support #1193 (New): Naming convention mistakes
I have noticed some functions whose names do not follow the naming conventions: _e.g._ *@quorem@* and *@divexact@*
...
John Abbott

19 Jun 2018

17:29 CoCoA-5 Design #1192 (New): RealRoots: rational roots (but not binary rational roots)
What should @RealRoots@ do with an input like @3*x-1@?
Currently, if the root is integer or a binary rational, @Re...
John Abbott
16:46 CoCoA-5 Bug #1191 (Feedback): ApproxSolve: log(0) on some examples
The problem was caused by a new better impl of @RootBound@ which can return zero (if arg is just @x@); the old @RootB... John Abbott
16:46 CoCoA-5 0.25 hour (Bug #1191 (Closed): ApproxSolve: log(0) on some examples)
Resolved John Abbott
16:32 CoCoA-5 0.10 hour (Bug #1191 (Closed): ApproxSolve: log(0) on some examples)
Oh goodness! John Abbott
16:32 CoCoA-5 Bug #1191: ApproxSolve: log(0) on some examples
Here is a surprise:... John Abbott
16:25 CoCoA-5 0.10 hour (Bug #1191 (Closed): ApproxSolve: log(0) on some examples)
Univariate failing case John Abbott
16:25 CoCoA-5 Bug #1191: ApproxSolve: log(0) on some examples
It even fails in a univariate case... @-3*x^2+2*x@.
Investigating...
John Abbott
16:17 CoCoA-5 0.25 hour (Bug #1191 (Closed): ApproxSolve: log(0) on some examples)
More failing examples John Abbott
16:17 CoCoA-5 Bug #1191 (In Progress): ApproxSolve: log(0) on some examples
Something odd must have happened: @ApproxSolve@ is failing on almost trivial examples.
Here are some simpler failing...
John Abbott
16:15 CoCoA-5 Bug #1191 (Closed): ApproxSolve: log(0) on some examples
The following input attempts to compute log(0).... John Abbott
11:44 CoCoA-5 0.30 hour (Bug #1190 (Closed): HilbertBasisKer: SEGV (again))
Nasty John Abbott
11:44 CoCoA-5 Bug #1190: HilbertBasisKer: SEGV (again)
CoCoA should never SEGV --> high priority.
Also it would help to make @HilbertBasisKer@ interruptible.
John Abbott
11:43 CoCoA-5 Bug #1190 (Closed): HilbertBasisKer: SEGV (again)
I get SEGV with the following input:... John Abbott
11:27 CoCoA-5 Bug #1189: HilbertBasisKer: unhelpful error with some inputs
High priority because it's embarassing, but hopefully easy to fix.
John Abbott
11:27 CoCoA-5 0.30 hour (Bug #1189 (Closed): HilbertBasisKer: unhelpful error with some inputs)
Check inputs? John Abbott
11:25 CoCoA-5 Bug #1189 (Closed): HilbertBasisKer: unhelpful error with some inputs
@HilbertBasisKer@ gives an unhelpful error sometimes:... John Abbott
11:08 CoCoA-5 Support #1188 (Closed): Product of empty list: manual and code disagree
What should CoCoA-5 do when asked to compute the product of the empty list?... John Abbott
10:56 CoCoALib 0.25 hour (Slug #1187 (New): Matrix rank is slow (over QQ))
Modular heuristic approach John Abbott
10:56 CoCoALib Slug #1187: Matrix rank is slow (over QQ)
A potentially fast heuristic is to reduce mod p and compute the rank of the reduced matrix. This should be quick; if... John Abbott
10:50 CoCoALib Slug #1187 (New): Matrix rank is slow (over QQ)
The rank function is too slow:... John Abbott

15 Jun 2018

23:56 CoCoA 0.20 hour (Bug #1186 (New): Web site: mailto link for bugs, suggestions etc broken)
Half right John Abbott
23:56 CoCoA Bug #1186: Web site: mailto link for bugs, suggestions etc broken
The "description" of the email address is correct, but the address inside the "mailto" link is incomplete.
John Abbott
22:07 CoCoA Bug #1186 (New): Web site: mailto link for bugs, suggestions etc broken
On the main web page for CoCoA, in the bottom left corner, there is a @mailto@ link for sending email to the cocoa te... John Abbott

14 Jun 2018

21:04 CoCoALib Design #832: Generalize grading matrix
I have just started working on an implementation of isomorphic-to-subring. For example, given @QQ[x,y,z]@ it should ... John Abbott
21:04 CoCoALib 0.25 hour (Design #832 (New): Generalize grading matrix)
Perhaps not full rank? John Abbott
17:28 CoCoALib Bug #935 (Feedback): ElimMat, ElimHomogMat: complain about zero or negative weights
Anna says that @ElimMat@ should allow a grading mat with 0 rows, but @ElimHomogMat@ should not.
I have adjusted the ...
John Abbott
17:28 CoCoALib 0.20 hour (Bug #935 (Closed): ElimMat, ElimHomogMat: complain about zero or negative weights)
Done John Abbott
16:16 CoCoALib 0.25 hour (Bug #935 (Closed): ElimMat, ElimHomogMat: complain about zero or negative weights)
Status? John Abbott
16:16 CoCoALib Bug #935 (In Progress): ElimMat, ElimHomogMat: complain about zero or negative weights
What is the status of this issue?
What do we allow as "grading matrices"?
Current code requires entries to be int...
John Abbott
17:11 CoCoA-5 0.20 hour (Feature #978 (Closed): CommonDenom: for polys and lists?)
Mostly done? John Abbott
17:11 CoCoA-5 Feature #978 (Resolved): CommonDenom: for polys and lists?
This appears already to have been done (and documented).
Feedback? Close?
John Abbott
16:48 CoCoALib Design #1159: Add global enum "verify/DontVerify"
Close this issue and refer to #1167?
John Abbott
16:48 CoCoALib 0.10 hour (Design #1159 (Closed): Add global enum "verify/DontVerify")
Close? John Abbott
16:29 CoCoALib Slug #1181 (Feedback): CpuTime is costly!
Changing to *feedback* even though there is scope for improvement.
It seems to be difficult to do this properly in...
John Abbott
16:29 CoCoALib 0.25 hour (Slug #1181 (Closed): CpuTime is costly!)
Feedback John Abbott
15:53 CoCoALib Feature #202: MatrixView/function for viewing a single row or column (RowMat, ColMat)
I have just checked the code: it is in @MatrixView@ (as one would expect).
I am not so happy about the names *@Row...
John Abbott
15:53 CoCoALib 0.25 hour (Feature #202 (Closed): MatrixView/function for viewing a single row or column (RowMat, ColMat))
Fn names? John Abbott

13 Jun 2018

14:51 CoCoA-5 Feature #346: Describe should indicate package name
The CoCoA fn *@Starting@* says when a symbol comes from a package, and which package it comes from, so @describe@ sho... John Abbott
14:51 CoCoA-5 0.25 hour (Feature #346 (Closed): Describe should indicate package name)
Starting can do it! John Abbott

12 Jun 2018

18:35 CoCoALib Feature #1155 (Feedback): Create a new "prime source" iterator
Time to move to feedback.
Only remaining doubt: the function @NextPrime@ needs to be reconsidered (what happens wh...
John Abbott
18:35 CoCoALib 1.00 hour (Feature #1155 (Closed): Create a new "prime source" iterator)
Seems OK John Abbott
18:32 CoCoALib Feature #1131 (Feedback): Sturm sequence
Works OK in ad hoc testing -- fixed a couple of bugs.
John Abbott
18:32 CoCoALib 1.00 hour (Feature #1131 (Closed): Sturm sequence)
Seems OK now John Abbott
18:25 CoCoALib 2.50 hours (Feature #1176 (In Progress): interval arithmetic)
Partly done John Abbott
18:25 CoCoALib Feature #1176 (In Progress): interval arithmetic
A first prototype is in *@BigRatInterval@*
I have even written some doc!
John Abbott
18:22 CoCoALib Design #1145: File names, coding conventions: Fns, Ops, Operations? part 2
I have renamed the doc files too. John Abbott
18:22 CoCoALib 0.10 hour (Design #1145 (Closed): File names, coding conventions: Fns, Ops, Operations? part 2)
Doc done John Abbott
18:20 CoCoALib Design #410: New file: BigRatOps
I have split the sources (or did Anna do it?). Updated the doc, and all other files.
John Abbott
18:20 CoCoALib 0.25 hour (Design #410 (Closed): New file: BigRatOps)
Done John Abbott
18:18 CoCoALib 1.50 hour (Design #1180 (Closed): BigRat(0) unexpectedly compiles! (calls ctor with mpq_t arg))
Finished John Abbott
18:18 CoCoALib Design #1180 (Closed): BigRat(0) unexpectedly compiles! (calls ctor with mpq_t arg)
I have sorted out the doc.
John Abbott
16:50 CoCoALib Bug #1130 (In Progress): SqFreeFactor: should it work over ZZ?
One potential problem is that strictly speaking the content must be factorized too, and it is costly to tell whether ... John Abbott
16:50 CoCoALib 0.20 hour (Bug #1130 (In Progress): SqFreeFactor: should it work over ZZ?)
Some notes John Abbott
16:30 CoCoALib Design #764 (Resolved): ExternalLib: simplify configuration
Can we close this?
John Abbott
16:30 CoCoALib 0.10 hour (Design #764 (Closed): ExternalLib: simplify configuration)
Mostly done? John Abbott
16:22 CoCoALib Design #1162 (Feedback): Check semantics of NoSmallFactorSeq
I have no ideas for improving this currently. So have moved to "feedback".
John Abbott
16:22 CoCoALib 0.25 hour (Design #1162 (Closed): Check semantics of NoSmallFactorSeq)
No ideas John Abbott
15:58 CoCoALib Bug #1185 (Feedback): factor: multivariate, wrong RemainingFactor (sign wrong)
I have tested my change -- seems fine.
Checked in.
John Abbott
15:58 CoCoALib 0.50 hour (Bug #1185 (Closed): factor: multivariate, wrong RemainingFactor (sign wrong))
Checked in John Abbott

10 Jun 2018

22:28 CoCoALib Bug #1185 (Resolved): factor: multivariate, wrong RemainingFactor (sign wrong)
I believe I have fixed the problem (without really finding what was the original bug).
I have changed the (ancient) ...
John Abbott
22:28 CoCoALib 2.00 hours (Bug #1185 (Closed): factor: multivariate, wrong RemainingFactor (sign wrong))
Fixed (I think) John Abbott

07 Jun 2018

10:57 CoCoALib Bug #1185: factor: multivariate, wrong RemainingFactor (sign wrong)
I have a workaround for the problem: in the file @factor.C@ I ensure that all factors have positive leading coeff, an... John Abbott
10:57 CoCoALib 1.50 hour (Bug #1185 (Closed): factor: multivariate, wrong RemainingFactor (sign wrong))
Workaround John Abbott

05 Jun 2018

23:49 CoCoALib 0.50 hour (Bug #1185 (Closed): factor: multivariate, wrong RemainingFactor (sign wrong))
More failing examples John Abbott
23:49 CoCoALib Bug #1185 (In Progress): factor: multivariate, wrong RemainingFactor (sign wrong)
More failing examples:... John Abbott

04 Jun 2018

18:31 CoCoALib Bug #1185 (Closed): factor: multivariate, wrong RemainingFactor (sign wrong)
Gereon Kreme reports the following bug (JAA has confirmed)... John Abbott
 

Also available in: Atom