All,
I am sorry for not having included Don's message in my most recent
post in this thread. It must have been confusing.
To continue that discussion where I mentioned using M. ,
here are some experiments I performed. Do they suggest that M. is
actually reducing the time requirement? I am not sure.
ts =: 6!:2 , 7!:2...@]
F=: I. - ] < I. { [,_:
FM =: I. M. - ] < I. { [,_:
FMM =: I. M. - ] < I. M. { [,_:
FMMN =: I. - ] < I. M. { [,_:
test =: ?.100000#8
ts '2 3 5 F test'
0.10805 1.70515e6
ts '2 3 5 FM test'
0.066664 1.70515e6
ts '2 3 5 FMM test'
0.0505471 1.70515e6
ts '2 3 5 FMMN test'
0.0625262 1.70515e6
ts '2 3 5 F test'
0.0843358 1.70515e6
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
