[sage-devel] Re: Sage class patches
On Wed, Jul 2, 2008 at 3:23 AM, William Stein [EMAIL PROTECTED] wrote: On Tue, Jul 1, 2008 at 8:03 AM, Robert Miller [EMAIL PROTECTED] wrote: William, There were several awesome projects in your Sage class last quarter that I haven't seen any trac tickets about. For example, did Ryan Dingman ever send you a patch for his tree generation stuff? What about that Hamiltonian cycle code? Sage chat? Multi-edges? I tried dingman.sws, that worked fine, but then I tried rutherford.sws and the Sage notebook said: Error uploading worksheet 'Error decompressing saved worksheet.'. terminal: 2008/07/02 09:27 +0200 [-] cd /home/ondra/.sage/temp/fuji/4492/dir_0; tar -jxf /home/ondra/.sage/sage_notebook/tmp.sws 2008/07/02 09:27 +0200 [-] 2008/07/02 09:31 +0200 [-] cd /home/ondra/.sage/temp/fuji/4492/dir_0; tar -jxf /home/ondra/.sage/sage_notebook/tmp.sws bzip2: (stdin) is not a bzip2 file. tar: Child returned status 2 tar: Error exit delayed from previous errors I also tried to decompress by hand and indeed, it doesn't work. So I looked what's wrong and the following makes it clear what's wrong: [EMAIL PROTECTED]:~/Desktop/Downloads$ file dingman.sws dingman.sws: bzip2 compressed data, block size = 900k [EMAIL PROTECTED]:~/Desktop/Downloads$ file rutherford.sws rutherford.sws: RAR archive data, v1d, os: Win32 Ondrej --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~--~~~~--~~--~--~---
[sage-devel] And the winner in the X-monthly let's use up space on sage.math competition is ...
William Stein as usual with a little over 400 GB. In total we are using 711G. If you are on the list below you are using at least 1GB space. You should check if any of those data can be delete and then get rid of them. That does not mean that you must delete data, just a friendly reminder that we are running low on space and in case there are old Sage trees you no longer need please get rid of them. Cheers, Michael 1.9Gabhinav 4.1Gacgetchell 4.3Gadam 1.4Gaklemm 2.0Gallan 4.3Gbgranger 15G binegar 1.5Gbober 4.7Gboothby 5.0Gbrettnak 1.6Gbroune 7.1Gburcin 24G burhanud 4.9Gclarita 7.3Gcraigcitro 2.3Gcswiercz 9.4Gcwitty 1.9Gdav 11G dbm25 1.8Gdeldotdr 6.6Gdfdeshom 1.1Gdisabled_accounts 4.7Gdmharvey 1.5Gekirkman 2.9Gelliottd 3.9Ggeorgesk 13G gfurnish 1.8Gghtdak 1.2Ggregorybard 1.3Gjared 5.0Gjason 1.7Gjec 2.7Gjen 1.1Gjetchev 4.5Gjkantor 1.1Gjsp 4.7Gjvoight 1.1Gkathy 1.3Gkedlaya 2.4Gmabshoff 2.5Gmacaulay2 1.9Gmalb 1.7Gmanny 5.4Gmhansen 3.9Gmoretti 1.1Gnathan 1.1Gncalexan 1.2Gnoam 7.0Gnovoselt 3.8Gpadicgroup 4.4Gpage 4.3Gpernet 3.0Gpurbon 1.4Grishikesh 3.4Grlmill 5.7Grobertwb 5.0Groed 28G savitt 5.3Gsimuw 3.1Gsuperstein 3.7Gtclemans 1.2Gtkosan 2.6Gtornaria 407Gwas 4.4Gwatkins 2.7Gwbhart 3.1Gwdj 6.6Gyqiang --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~--~~~~--~~--~--~---
[sage-devel] Re: Computing large Bernoulli numbers
Returning to a slightly old thread. I have implemented a new algorithm for computing large Bernoulli numbers. Running on 10 cores for 5.5 days, I computed B_k for k = 10^8, which I believe is a new record. (Recall the Mathematica blog post from April was for k = 10^7.) Essentially it's the multimodular algorithm I suggested earlier on this thread, but I figured out some tricks to optimise the crap out of the computation of B_k mod p. Patch is up for review here: http://sagetrac.org/sage_trac/ticket/3542 Preprint is here (comments welcome): http://math.harvard.edu/~dmharvey/bernmm/bernmm.pdf One data point, on a 2.6GHz opteron: sage: time x = bernoulli(10^5, algorithm=pari) CPU times: user 0.16 s, sys: 0.00 s, total: 0.16 s Wall time: 20.57 s sage: time y = bernoulli(10^5, algorithm=bernmm) CPU times: user 6.54 s, sys: 0.00 s, total: 6.54 s Wall time: 6.54 s sage: time z = bernoulli(10^5, algorithm=bernmm, num_threads=3) CPU times: user 6.54 s, sys: 0.03 s, total: 6.57 s Wall time: 2.71 s sage: x == y True sage: x == z True Timings for some bigger k: k = 10^7: PARI/GP = 75 h Mathematica = 142 h bernmm (1 core) = 11.1 h bernmm (10 cores) = 1.3 h k = 10^8: bernmm (10 cores) = 131h = 5.5 days david --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~--~~~~--~~--~--~---
