Even before JMP17 I was looking for faster ways to open some specific rows from large JMP tables to improve the user-experience of tools I write. Finally JMP17 provided us with options to open x first, x last or x random rows. I would suggest that JMP should also let user open data table using specific rows (not sure if this can be technically done). Syntax for these is First(x), Last(x), Random(x) and if x is between 0-1 fraction of the total number of columns is used. And the new syntax could be same as Subset has for example with Rows([1,2,3,4]) dt = Open("$SAMPLE_DATA/CrimeData.jmp", Rows([2,4,6,8,30])); .
One of my use cases would be following:
I have a data table with 10million rows and 100 columns. First I open the table with only one (or multiple) column and ALL rows. Then I calculate which rows I am interested in, close the one column table and then re-open with just the interesting rows. This should make opening large tables much faster if I don't need all the data. Currently I can do this with if the table is sorted and I need sorted data.
