One more time, sending only the trouble-making text:
_______________________________________________________________
Trying again and adding more...
A neat measure of neatness,
neat=. #@:|:@:":@:(5!:2)@< neat e ('waveVC' ; 'waveJQ' ; 'waveR' ;
'waveR0' ;
'cheat')┌───┬───┬───┬───┬──┐│117│114│108│102│94│└───┴───┴───┴───┴──┘ but it is
not invulnerable.
_______________________________________________________________
I agree, counting the number of tokens (especially focusing on the string used
to define the verb) is not particularly productive, nt=. # @: ;: @: lr @:
< nt e ('waveVC' ; 'waveJQ' ; 'waveR')┌──┬──┬──┐│41│40│32│└──┴──┴──┘
Counting the number of structural components is, in my opinion, better as a
measure of the complexity of (honest) fixed tacit verbs,
nsc=. # @: (<S:0) @: (5!:2) @: < nsc'nsc'11 nsc0=. [: # [: (<S:0) [:
(5!:2) < nsc0 e ('nsc';'nsc0')┌──┬──┐│11│11│└──┴──┘ nsc1=. # @: ((<S:0) @:
((5!:2) @: <)) nsc e ('nsc';'nsc0';'nsc1')┌──┬──┬──┐│11│11│11│└──┴──┴──┘
nt e ('nsc';'nsc0';'nsc1')┌──┬──┬──┐│15│11│15│└──┴──┴──┘ waveR0=. [: <"_1 [:
|.&|:^:(<5) [: (|."_1~ ([: >. +/ .=&' ' % _2:)) [: > ;:
nsc e ('waveVC' ; 'waveJQ' ; 'waveR' ;
'waveR0')┌──┬──┬──┬──┐│32│31│28│28│└──┴──┴──┴──┘
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
