Greetings! I've been working on familiarising myself with with finite field linear algebra implementations in Sage and fflas-ffpack and would like to reach out to the prospect mentors (Clément Pernet and Vincent Neiger) in order to understand the scope of the project. I've explored the routines fflas-ffpack offers through these papers -
https://dl.acm.org/doi/pdf/10.1145/1391989.1391992 https://dl.acm.org/doi/10.1145/1005285.1005304 these course notes on dense and sparse linear algebra - https://courses.csail.mit.edu/18.337/2006/book/ as well as the documentation given in the fflas-ffpack repository. I'm confused about which particular fflas-ffpack routines should be used in Sage as the GSoC project. Currently, only the vector-matrix multiplication, matrix-matrix multiplication and echelonize functions in finite field matrices use ffpack, even though ffpack offers other methods like rank and determinant. Linbox is being called for the other methods in Sage (which I would assume is slower than directly calling ffpack, as is noted in the documentation for echelonize). I have some questions - - Would it be worthwhile to switch the defaults from linbox to ffpack for some of these methods as is done in echelonize? - Which pivoting strategies are being talked about in the GSoC project description? - Would a project that just involves making all useful fflas-ffpack routines available through Sage be appropriate for a 300 hour project, or would FLINT routines necessarily need to be included as well? As I work on this, discussing these issues with the mentors would be amazing. Is there an IRC / Slack / Discord channel where I could be in direct contact with the mentors? Thanks! Regards, Karan -- You received this message because you are subscribed to the Google Groups "sage-gsoc" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-gsoc+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-gsoc/7e58c0cf-d4f0-4078-b15a-cf1f564b302fn%40googlegroups.com.
