hh wrote > [@Bernd: here a PONG to your PING.] > > Obviously each list member has to be inspected once for the comparison. > And each "smaller than"-member has to be inspected once again to find > the maximum. There is no other way, this is a math fact. > > Here is my solution, that seems to be pretty fast. > > -- v is a real num, L is a list of real comma-delimites nums > -- if L has NO member < v then return MINVAL (=your setting) > -- Empty members are treated as 0 (doesn't harm, if all x>0) > -- Pre-sorting of L gives no advantage. > > function greatestMemberSmallerThan v,L > put -10^15 into z[true] --> because max(empty)=0 > repeat for each item x in L > -- if x is empty then next repeat --> here not necessary > put comma & x after z[x<v] > end repeat > return max( z[true] ) > end greatestMemberSmallerThan
Hermann, you know that math is my weak spot (among others) but whatever math says in my code I don't touch every member of the list, in my case the sorting has "touched" ever member and I can safely do my "binary search" type of thing. Missing values are no problem but empty items would. did you benchmark my snippet and yours say on a list of 100000 items? Kind regards Bernd -- View this message in context: http://runtime-revolution.278305.n4.nabble.com/function-for-greatest-object-in-list-less-than-or-equal-to-a-value-tp4697221p4697311.html Sent from the Revolution - User mailing list archive at Nabble.com. _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
