All Kinds of Fun!
Some Notes for a Later Post
How good are the engines that are created with Yould? Here a good way to find out:
First, create a complete word list used in an engine
yould -t cicero.yould -s > wordlist.txt
Second, generate 100 words with the engine using –real-words flag
yould -t cicero.yould -r -n100 > co100.txt
Third, sort the 100 words in alphabetical order
sort co100.txt > sco100.txt
Fourth, run a comparison that prints out only the words in sco100.txt that are not in wordlist.txt
comm -23 sco100.txt wordlist.txt > compare.txt
Fifth, count the number of lines in compare.txt
wc -l compare.txt
You can script this as well, just needed to make some notes before doing some explanation later. This works on Windows if you use unixutils from sourceforge. Very handy toolset to have.
| Print article | This entry was posted by John Payne on January 7, 2008 at 4:25 pm, and is filed under yould. Follow any responses to this post through RSS 2.0. Both comments and pings are currently closed. |
Comments are closed.
Recent Comments