Thanks! I know about dissect and do indeed use it a lot. In this case, dissect doesn’t make it really clear (at least to me) that only changes in y are fed to repeated calls with ^: as you said.
Regards, F. PS Interestingly enough, although I got the answer from Pascal Jasmin, I did not received yours email but grabbed it from the email archive. > Welcome to J, and to the Forum! > > This looks right to me. In the last case the verb is 3&(*:@[) which > always produces 9: > > 3&(*:@[) "0 i. 10 > 9 9 9 9 9 9 9 9 9 9 > 3&(*:@[) 1 2 3 4 > 9 > > You execute it 4 times, getting 4 results of 9. > > The difference with the first case is that there you are executing > > x *:@] y > > which squares y, not x; and y changes with each iteration. > > If you haven't seen Dissect, give it a look. It can help with things > like this (I didn't try it on this example). > > You say you see a change between 8.07 and 9.01: please report that. I > don't think there should be a difference except for rare cases. > > Henry Rich > > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
