Difference between revisions of "Western Abenaki/Transducer"

Revision as of 14:07, 14 March 2023

This is a WIP transducer for Western Abenaki

Evaluation

Code

Coverage Data

TOP UNKNOWN WORDS:

	14 ^ni/*ni$
 	8 ^idamow/*idamow$
 	7 ^ȣnda/*ȣnda$
 	7 ^paami/*paami$
 	7 ^oji/*oji$
 	7 ^itekw/*itekw$
 	6 ^ȣda/*ȣda$
 	5 ^tȣni/*tȣni$
 	5 ^niji/*niji$
 	5 ^kanwa/*kanwa$
 	5 ^aidid/*aidid$
 	4 ^waji/*waji$
 	4 ^nokmessek/*nokmessek$
 	4 ^alȣmiwi/*alȣmiwi$
 	4 ^ala/*ala$
 	4 ^achi/*achi$
 	3 ^wji/*wji$
 	3 ^wassakol/*wassakol$
 	3 ^pmȣwzimek/*pmȣwzimek$
 	3 ^pazgo/*pazgo$

coverage: 64 / 689 (~0.09288824383164005806)

remaining unknown forms: 625

ambiguity: 797 / 689 (~1.15674891146589259797)

Thu Mar 2 08:54:14 PM EST 2023: 64/689 ~0.09288824383164005806 \| 797/689 ~1.15674891146589259797

Reg Tests

Corpus 1 of 6: Noun Pluralization-gen

 15/15 (100.0%) tests pass (13/15 (86.67%) match gold)

Corpus 2 of 6: Pronouns and Possession-gen

 26/26 (100.0%) tests pass (0/26 (0.0%) match gold)

Corpus 3 of 6: Locative Case-gen

 8/8 (100.0%) tests pass (0/8 (0.0%) match gold)

Corpus 4 of 6: Adjective Agreement-gen

 9/9 (100.0%) tests pass (0/9 (0.0%) match gold)

Corpus 5 of 6: Verbs-gen

 43/43 (100.0%) tests pass (0/43 (0.0%) match gold)

Corpus 6 of 6: Adverbs-gen

 10/10 (100.0%) tests pass (0/10 (0.0%) match gold)