How to summarize by variable column names

MuttonChops — Wed, 10 May 2017 23:22:12 GMT

I have a table with variable numeric, continuous column names. I need to get the mean of each column, grouped by a different column. The script below shows what I'm trying to do and it makes a generic table that works for this purpose. I'm getting hung up on the "Mean(ColNames)" line because it can't seem to process that.

If I replace "Mean(ColNames)" with "Mean(:Col2, :Col3)" or "Mean(:Name("Col2"), :Name("Col3"))" then it works but obviously doesn't meet my need of being able to use variable columns. Am I close or way off?


// Make a table to work with
nr = 20;
nc = 3;
dt = NewTable("Random_Table");
For(c=1, c<=nc, c++,
  dt << NewColumn("Col"||Char(c), Numeric, Continuous, Formula(RandomInteger(1, 10)));
  );
dt << AddRows(nr);
Column(1) << Data Type(Character);

//Get column names for the numeric continuous columns
ColNames = dt << Get Column Names(Numeric, "Continuous");

dt << Summary(
		Group(:Col1),
		Mean(ColNames),
		statistics column name format( "column" )
);

Re: How to summarize by variable column names

txnelson — Thu, 11 May 2017 01:01:18 GMT

Put and Eval() around the columns list and it will work

names default to here(1);
dt=open("$SAMPLE_DATA\big class.jmp");

ColNames = dt << Get Column Names(Numeric, "Continuous");

dt << Summary(
		Group(:sex),
		Mean(eval(ColNames)),
		statistics column name format( "column" )
);

topic Re: How to summarize by variable column names in Discussions

How to summarize by variable column names

Re: How to summarize by variable column names