Skip to content

Latest commit

 

History

History
36 lines (24 loc) · 1.31 KB

README.md

File metadata and controls

36 lines (24 loc) · 1.31 KB

Code for generating the data for the mostly512.net website (or whatever we're calling it!)

Called as follows (assuming you're in this directory, with GAP installed in a directory next to it):

$ ../gap/bin/gap.sh smallgrp_parallel.g
gap> write_smallgrp_csv("out.csv", [1 .. 1023], 60);

to run it on all groups of order up to 1023, using 60 parallel threads, and output the results in a file called out.csv.

On the monster server, I estimate this would take about 4 days, so this is probably feasible.

We probably want to fix the output by running

$ cat *.csv | sed ':a;N;$!ba;s/\\\n//g' | sort -Vu | sed '1h;1d;$!H;$!d;G' > all.csv

a pipeline which does the following:

* concatenates all csv files in the directory
* cleans up line breaks
* sorts the results
* moves the header line back to the start

This has now been run to completion for [1 .. 1023], resulting in a file of size 915 MB (or 49MB after gz compression). That's probably too big to put in this repository.

To get a json file out of this (in the format KB suggested) we can run

$ sed '1s/^.*$/\[/;s/^\(\([0-9]*,\)\{2\}\(true,\|false,\)\{4\}\([0-9]*,\)\)\([0-9]*,[0-9]*\),\(.*\)$/  [\1[\5],"\6"],/g;$a\]' all.csv > all.json

and use all.json to create a server (I actually compressed this and committed it to this repo anyway!)