This is what happened at Sage Days 46: Hawaii:
* Vivek Venkatachalam <[email protected]>: I explored the possibility of adding vector-based (SVG) visualizations to Sage using the D3 framework. This allows fluid, responsive interactive components to be generated by caching data on the client. The output of this project involved preparing a demonstration of SVG line plots with a one-dimensional UI control and document-cached data. * Xander Faber: familiarized himself with Benjamin Hutz's arithmetic dynamics code for computing rational preperiodic points for rational functions defined over Q. He then wrote new code for extending the local parts of the algorithm to the case of arbitrary number fields. What remains is to implement the lifting part of the algorithm via the Chinese remainder theorem. * William Stein: I discussed the design of both interacts and the Sage Cell server with Jason Grout, and helped make it more scalable and highly available by deploying on multiple physical and virtual machines. We discussed the design of SageMathCloud with participants, greatly improved the xterm functionality (adding resizing, font sizing, and color schemes), and I worked with Vivek on his D3-based visualization code. I also wrote an %fork decorator, to enable easy async execution in a worksheet, and wrote a %exercise decorator on the flight home. * Simon Spicer: Authored Patch 14189: Extend modular degree and congruence modulus of elliptic curves over QQ to arbitrary level. Used code from the above patch to generate data to support formulation for conjectures relating to modular degrees and congruence numbers at higher levels. Reviewed the following number-theory tickets written by David Roe: Ticket 6567 - Function to test whether or not some integer is a primitive root modulo n Ticket 12109 - Function for faster evaluation of cyclotomic polynomials Ticket 12116 - perfect_power for integers Ticket 14155 - gcd for IntegerMod Ticket 14193 - Converting polynomial quotient ring elements to PARI yields plain polynomials Started work on the following tickets: Ticket 12117 - Bugfixes and improvements to Aurifeuillian factorization Ticket 7240 - Factorization of Cunningham numbers - applications Ticket 12561 - Factoring p-adic polynomials * Mirela Ciperiani: I worked on computing anticyclotomic p-adic heights and removing the uncertainty factor in the computation of cyclotomic p-adic heights of Heegner points. * Michelle Mannes: (*) With Bianca Thompson: editing paper that grew out of Sage Days for Women 3 project on dynamics over finite fields. (*) With David Lukas (with lots of support from Aly): Learned how to create & upload a patch to Sage. Made a small patch to how absolute_ideal works & uploaded it for review (#14212). (*) Reviewed patch #13130, sent comments to the author, then reviewed again (positive review, set to go in Sage 5.8beta3). (*) Reviewed patch #2217 (needs work). * Alyson Deines and Ling Long: Ling and I wrote a function which takes square roots of q-expansions and used this compute coefficients of a weight one non-congruence modular form. Based on work of Atkin and O'Brien, "Some properties of p(n) and c(n) modulo powers of 13" and the data, Ling was able to formulate two conjectures about the relationship of the q-expansion coefficients, which should be related to congruences between modular forms or p-adic modular forms. * Jennifer Balakrishnan: At Sage Days 46, I studied how to characterize the first level of the anticyclotomic tower to remove the "highly likely" condition on our construction of Heegner points and their field of definition. I also worked on computing a global anticyclotomic character from local data, as the first step to computing anticyclotomic p-adic heights on higher rank elliptic curves. * Keith Clawson: I worked on setting up a high availability cluster to run the Sage cell server. I also worked on some systems administration tasks for the servers at UW. * Kartik Venkatram: I was (and am still) working on putting NFS style factoring in sage. Several of the relevant libraries have been made into sage packages, and there are rudimentary functions to use them, but the main project is a class based approach which separates the polynomial searching, sieving, and linear algebra capabilities for separate use and optimization. Currently, the code is designed to use the binaries, but future work should wrap the c functions individually for greater versatility. The classes are inherently asynchronous, and use the underlying mpi of the nfs binaries for further parallelization. * Volker Braun: I implemented and deployed a kvm-based script to keep long-running computations alive beyond OS restarts (https://github.com/vbraun/ImmortalSage). I also finished my rewrite of matrix groups, which now makes use of LibGAP. Finally, I added options to control the graphical display of polygons and wrote an enhancement for the resolution of lazy (delayed) imports. * David Roe: I spent some time working on getting #12415 ready (including reviewing some dependencies: #14079, #14150, #14158). I also worked on cleaning up a bunch of my old tickets (#6567, #14155, #12116, #12109, #12117, #12125, #7240) and getting them reviewed. Finally, I worked a bit on templating p-adics (#12555) and found some bugs in Sage's interface with PARI (#14082, #14193, #14194). * Victoria DeQuene: I worked on getting my Jacobi theta series code ready for inclusion in Sage. Code that previously used doubles now uses ints, and the code now works with Sage 5.7 (as opposed to 4.0.2). I also documented the main functions. There is code posted at #14244 which should be ready for review soon. * Jason Grout: I worked quite a bit with William and Keith on making the Sage Cell Server scalable and fault-tolerant, including writing a new database adapter and web server and setting up virtual machines with Keith and William to deploy load-balanced virtual Sage Cell Servers. I also had long discussions with William about SageMathCloud and spent time getting more familiar with the current state of SageMathCloud. I also worked a lot on understanding the similarities and differences between William's and my new interact implementations, thinking about the proper design of interacts, and fixed a major design problem with my implementation. I also helped or had discussions in several other efforts, like Kartik's packaging of libraries into spkgs, and Vivek's work in D3 graphics.