Robin Lee Powell wrote:
On Tue, Oct 26, 2010 at 01:56:00PM -0400, Nora LeChevalier wrote:At 02:39 AM 10/26/2010, Robin wrote:Oh, also, it means the actual formally written out immediately before is wrong, and possibly some of the surrounding verbiage. It also makes negative lujvo possible, which seems icky. -Robin[snip]On Tue, Oct 26, 2010 at 02:33:56AM -0400, John Cowan wrote:At the last minute we (which I think means Nora) decided to subtract everything from 32500 so that higher values = better lujvo. 12.1 obviously got missed, so your fix is just the wrong way round.In the program, a max of 6 pieces was handled.Ah. That changes things.Looking at calculations in my lujvo-making program ( I presume this is what you're referring to?), higher is better score: Luj.LujScore := (32 - length(Luj.Lujvo)) * 1000 + ApostTot * 500 + (5 - HyphenCount) * 100 + TypeTot * 10 + VowelTot;Thaaaaaaat's not *remotely* like what's in *any* version of the book. Every book version has both positive and negative segments. It also doesn't have the 6 piece limit.
Actually it is almost identical to the book examples (with the 32500 subtraction). You just have to expand the formula algebraically. The above becomes
32000 - (1000 * L) + (500 * A) + 500 - (100 * H) + (10 * R) + Vwhich is the same as subtracting the book formula from 32500 (flipping all the signs):
(1000 * L) - (500 * A) + (100 * H) - (10 * R) - V becomes 32500 - (1000 * L) + (500 * A) - (100 * H) + (10 * R) + V which is Nora's equation.
I'd really prefer to change the book as little as possible. That means lower score is better. Is it *really* important that we flip it around, especially at the expense of limiting the alg's applicability?
It isn't important at all. Indeed, the parenthetical in the first paragraph of section 12.1 shows that flipping it was the explicit intent. John (apparently) changed it from highest scoring to lowest scoring (probably to eliminate the 6 rafsi limit) but didn't correctly change the example calculations.
So make them match the formula in the book (which requires removing the "32500 -" and adjusting other signs to exactly match the formula, and then using a calculator to get the correct total for the low-scoring algorithm; I think only 12.3 has the correct total for the stated formula, if my head-calculator is working and I flipped the right signs.)
lojbab -- You received this message because you are subscribed to the Google Groups "BPFK" group. To post to this group, send email to bpfk-list@googlegroups.com. To unsubscribe from this group, send email to bpfk-list+unsubscribe@googlegroups.com. For more options, visit this group at http://groups.google.com/group/bpfk-list?hl=en.