Talking to a user recently, they were confused about how to pass divisions= in set_index. Specifically, they missed that the list had to have N+1 elements in it, and the N+1th element was the upper bound for the last partition.

The API docs currently link to https://docs.dask.org/en/latest/dataframe-design.html#partitions, which does describe this in words, but takes some close reading to get.

A simple example of setting divisions using a list of values, both in the set_index docs and in the dataframe design section, might be helpful.