Tripartite Palindromic Partition of Integer (HP 50g) Challenge

[2old2randr]

(02-21-2023 05:38 PM)3298 Wrote:  Anyway, I'm in the process of translating the proof's algorithm into SysRPL code. Between SysRPL speed in general and the lack of backtracking in that algorithm (the "correction step" isn't really backtracking, right? 'cause it's not repeated, it merely looks like another set of if-then-else at the end), I'm expecting it to be quicker to execute ... once I am done implementing all those pesky branches of that algorithm at least. Although that'll probably take another few days.

Yes, my user RPL version (not quite complete) does 21-22 digit numbers in less than 3 seconds on a physical 50g so your sysRPL version should be practically instantaneous.

One thing that tripped me up several times is that the authors do not have consistent indexing subscripts between the 4 arrays that they use. There is 0-based indexing, 1-based indexing and 2-based indexing at different places ...

Sudhir