Using 6 numbers from (0) to (2) blocks in combinations block can be used to make ternary numbers, but I somehow come across this:
Notice that Column F starts with [1, 2, 0, 1, 2, 0, ...] rather than [0, 1, 2, 0, 1, 2, ...], despite using [0, 1, 2] in order in the inputs.
Can this be fixed?
It's probably a case of the order isn't guaranteed and that you'll need to sort the data if sorted order is required
6 or 7...
strange (ok at 8 again,)
Okay, that's weird. Technically @cymplecy is right, we don't guarantee the order, so this isn't technically a bug; on the other hand, the code is trying to produce the order you expect. It's failing, apparently, because a computation that should produce 1 is instead producing a floating point value slightly more than 1, and the ceil() function is turning that into 2. It's probably fixable by making the code less clever, but Jens might not want to do that, to avoid making it slower. Anyway, don't hold your breath.
