nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Select 10 highest values from a variable split by date

Valerie_Nedbal

Hi,

Is there a script available how to select 10 highest values from a variable grouped by a date variable. for example from the big class data set, how to find the 3 highest values of weight or height, grouped by age. Hope it is clear.

Cheers,

Valerie

Accepted answers

Valerie_Nedbal

Thanks this works!

Thanks in advance,

Best regards,

Valerie

image001.jpg

All comments

brady_brady

Hi Valerie, this should work, where num is the numeric column and group is the group column:

dt << select where(colrank(:num, :group, <<tie("average"))>(colnumber(:num, :group)-10));

Cheers,

Brady

brady_brady

Valerie didn't get this to work--I realized I had used JMP 13, which adds the BY variable support for ColRank().

Without that, you've got to sort the table, and use the Row() function, which has BY support in JMP12.

This code selects the top 3 weights, for each age (note you'd need more logic in the event that there are ties.)

//selects the top 3:

dt = open("$Sample_Data/Big Class.jmp");

dt << sort(by(:age, :weight), order(ascending, descending), replacetable);

dt << select where (row() <= colminimum(row(), :age) + 2);

Valerie_Nedbal

Thanks this works!

Thanks in advance,

Best regards,

Valerie

image001.jpg