I noticed something strange, but I’m not sure if it’s a bug or intended behavior.
When every field in a schema has the same “blank” percentage (say, 50%), like this:
The resulting data contains rows that are either entirely blank or entirely non-blank, like this:
This isn’t what I would have expected. I would have expected that, for a given row, each value would have had a 50% change of being blank, regardless of whether other values in the same row were blank. Put another way, I would have expected it to look something like this:
Is the current behavior, where every row is entirely blank or entirely non-blank, a bug, or am I misunderstanding something?