Hi team, IIRC, in the current implementation, for all the provided dataset, we load all its columns by default.
For a wide table, say, with over 1K columns, it leads to high latency and unfriendly UI (I have to scroll to the very right side to find the columns I care about).
So I'm wondering if it's possible to get users select / exclude the columns they're (un)interested in, so we don't need to read and load every single column?
Hi team, IIRC, in the current implementation, for all the provided dataset, we load all its columns by default.
For a wide table, say, with over 1K columns, it leads to high latency and unfriendly UI (I have to scroll to the very right side to find the columns I care about).
So I'm wondering if it's possible to get users select / exclude the columns they're (un)interested in, so we don't need to read and load every single column?