Re: [sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Harald Schilly
On Fri, Jul 12, 2019 at 3:38 PM Simon King wrote: > > sage: DOT_SAGE > > '/home/user/.sage/' > > Do you literally mean "user", not the specific name of a single user? yes, all cocalc projects run under the same user "user" in their own container. $ echo $USER user $ id uid=2001(user) gid=2001(us

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Simon King
Hi Harald, On 2019-07-12, Harald Schilly wrote: > Interestingly, this works: > > matrix(GF(16), [[1,0], [0, 1]]) while matrix(GF(25), [[1,0], [0, 1]]) fails > with the above error. MeatAxe is only used for finite non-prime fields of order <255, and I forgot to add: Only in odd characteristic (

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Simon King
Hi Harald, On 2019-07-12, Harald Schilly wrote: > Ok... To me, this sounds like we have to uninstall MeatAxe on CoCalc. > > The CoCalc setup is quite easy to explain, and so far I wasn't aware of any > issues. Sage is in a globally shared read-only directory > /ext/sage/sage- and permissions ar

Re: [sage-support] problem with matrix over finite field

2019-07-12 Thread Vincent Delecroix
Dear Hal, For the time being, you can use the following workaround matrix(GF(25, a), [[1,0], [0, 1]], implementation='generic') Le 11/07/2019 à 23:20, Hal Snyder a écrit : This works on sage-8.5: sage: a = var('a') : matrix(GF(25, a), [[1,0], [0, 1]]) : [1 0] [0 1] but not on sage-8.

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Simon King
Hi Dima, On 2019-07-12, Dima Pasechnik wrote: > Hi Simon, > isn't MeatAxe interfaced via a library, rather than via files?! > Doing arithmetics on small matrices storing them on a disk is insanely > inefficient... Maybe. Do you mean one should instead compute a new multiplication table in memory

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Simon King
Hi Harald, On 2019-07-12, Harald Schilly wrote: > On Friday, July 12, 2019 at 11:48:40 AM UTC+2, Simon King wrote: >> >> sage: DOT_SAGE >> '/home/king/.sage/' > > It's > > sage: DOT_SAGE > '/home/user/.sage/' Do you literally mean "user", not the specific name of a single user? Does that per

Re: [sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Dima Pasechnik
by the way, libgap will get a direct interface to GAP matrices, which might be a useful matrix backend for various purposes. See https://github.com/gap-system/gap/pull/3554 On Fri, Jul 12, 2019 at 1:29 PM Harald Schilly wrote: > > Ok... To me, this sounds like we have to uninstall MeatAxe on CoC

Re: [sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Harald Schilly
Ok... To me, this sounds like we have to uninstall MeatAxe on CoCalc. The CoCalc setup is quite easy to explain, and so far I wasn't aware of any issues. Sage is in a globally shared read-only directory /ext/sage/sage- and permissions are set such that a user "user" can read/exec it. Hence, onl

Re: [sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Vincent Delecroix
Le 12/07/2019 à 11:56, Harald Schilly a écrit : I've no idea why this is happening on CoCalc. My first guess is there is an optional package installed, which is causing this. As Simon already said, there is meataxe installed. While Interestingly, this works: matrix(GF(16), [[1,0], [0, 1]]) w

Re: [sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Dima Pasechnik
Hi Simon, isn't MeatAxe interfaced via a library, rather than via files?! Doing arithmetics on small matrices storing them on a disk is insanely inefficient... Dima On Fri, Jul 12, 2019 at 11:48 AM Simon King wrote: > > Hi Hal, > > On 2019-07-11, Hal Snyder wrote: > > sage: a = var('a') > > ...

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Harald Schilly
On Friday, July 12, 2019 at 11:48:40 AM UTC+2, Simon King wrote: > > > sage: DOT_SAGE > '/home/king/.sage/' It's sage: DOT_SAGE '/home/user/.sage/' Which I bet is the wrong place to look. (all cocalc projects are essentially linux docker containers, where /home/user is $HOME, though)

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Harald Schilly
On Friday, July 12, 2019 at 11:48:40 AM UTC+2, Simon King wrote: > > > sage: DOT_SAGE > '/home/king/.sage/' It's sage: DOT_SAGE '/home/user/.sage/' Which I bet is the wrong place to look. (all cocalc projects are essentially linux docker containers, where /home/user is $HOME, though)

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Harald Schilly
I've no idea why this is happening on CoCalc. My first guess is there is an optional package installed, which is causing this. Interestingly, this works: matrix(GF(16), [[1,0], [0, 1]]) while matrix(GF(25), [[1,0], [0, 1]]) fails with the above error. -- You received this message because you

[sage-support] Re: problem with matrix over finite field

2019-07-12 Thread Simon King
Hi Hal, On 2019-07-11, Hal Snyder wrote: > sage: a = var('a') > : matrix(GF(25, a), [[1,0], [0, 1]]) > : > p025.zzz: No such file or directory > --- > RuntimeError Traceback (most recent