23 Jun
2012
23 Jun
'12
10:55 a.m.
Yes, that was my reason for using an "e.g.". --Dan << On 2012-06-23, at 9:21 AM, Tom Karzes wrote:
It's even worse than that, since both the altered vector and its paired vector are removed. So as few as half of the vectors being altered by the child could result in the empty subset (i.e., if only one vector in each pair is altered). For example, suppose all vectors whose initial element was originally +1 are altered to contain a zero element. Then all vectors would be removed and you'd end up with the disallowed empty subset.
Tom
Dan Asimov writes:
If, e.g., all vectors have been altered then this suggestion would lead to all vectors being removed, but the puzzle calls for a nonempty subset summing to 0.