View Single Post
Old 05-13-2022, 11:46 PM   #474
Deobulakenyo
Guru
Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.Deobulakenyo ought to be getting tired of karma fortunes by now.
 
Posts: 706
Karma: 2153490
Join Date: Aug 2021
Location: Stupid Philippines
Device: Kobo Libra 2, Boyue Likebook P78
Quote:
Originally Posted by CyberPaul View Post
I improved the formulas. With below I got pretty much same estimate shown on kobo store.
Code:
program:floor($$#words / 12000)
program:ceiling($$#words / 11150)
Simpler and more elegant
With this approach lower and higher estimates start to diverge when the books are big, and this is ok. Same is happening on kobo store. Having always higher = lower + 1 does not make sense when number of words is significant and uncertainty on estimate is increased.
I also materialized these two attributes as custom columns so I don't need to transfer them onto the kobo device to see the estimates produced
Thank you for this. This is a lot better than the minus one formula.
Deobulakenyo is offline   Reply With Quote