Skip to main content

Timeline for Compute the first N digits of e

Current License: CC BY-SA 2.5

11 events
when toggle format what by license comment
Mar 13, 2011 at 20:33 comment added Eelvex @Jesse: Use 450x then. It's enough for every n<1000 and takes just a few seconds (~9s here).
Mar 13, 2011 at 19:46 comment added Jesse Millikan @Eelvex: This brings up a point... This performs horribly for larger values of n because of the resulting denominator, which is (for example) */!i.999x if it's not reduced. I wasn't even patient enough to let +/%!i.999x finish. Hints here, possibly...
Mar 13, 2011 at 18:20 comment added Eelvex Ah, you're right, that's unfortunate. (That "mess" is shorter than anything else I could come up with - unless, off course, you use a fixed number for i. like: (0 j.<:n)":+/%!i.999x; then tacitly:(+/%!i.999x)":~0 j.<: )
Mar 13, 2011 at 18:04 comment added Jesse Millikan @Eelvex: But I need extra terms up until N=30 or so. (Any hints on reducing that mess of composes, or is that about right?)
Mar 13, 2011 at 17:32 history edited Jesse Millikan CC BY-SA 2.5
added 13 characters in body
Mar 13, 2011 at 14:49 comment added aaaaaaaaaaaa Would you care to explain your code?
Mar 13, 2011 at 10:09 comment added Eelvex For 1000 digits this: (0 j.<:1000)":+/%!i.x:450 takes just a few seconds.
Mar 13, 2011 at 10:05 comment added Eelvex This converges very fast (for 100 digits you need "just" 70 terms) so you can drop that +9.
Mar 13, 2011 at 6:45 history edited Jesse Millikan CC BY-SA 2.5
deleted 4 characters in body
Mar 13, 2011 at 5:52 history edited Jesse Millikan CC BY-SA 2.5
added 103 characters in body
Mar 13, 2011 at 4:18 history answered Jesse Millikan CC BY-SA 2.5