I would like to script a running count or cumulative sum of repeated values within a column.
For example, first appearance should equal 1, max value of count by value specified should equal the cumulative sum of the value within the column.
Ideally, the output for :count_color would not need the data table to be sorted by :color
Using the Eval command under formula generates an output, but all values = 0
Perhaps I have misinterpreted. My column formula is below.
// Running Count
Eval(Eval Expr(Col Sum(Row() <= Expr(Row()), :color)))
New Column( "count",
Format( "Best", 12 ),
Formula( Name Expr( Eval( Eval Expr( Col Sum( Row() <= Expr( Row() ), :color ) ) ) ) )
There is a bug in JMP that you've helped uncover; the resulting table does not include the Name Expr( ) wrapper around the formula when the table's <<getscript command is used. Thanks!
EDIT: May want to ignore this. This gives a total count of each item that can be used in a script later on, but doesn't actually address the OP's real question.
A somewhat more roundabout way to do this is to use the Summary platform and then read the resulting table into whatever script object you would like. I've used an associative array here.
// create temporary summary table new = Data Table( "Untitled 8" ) << Summary( Group( :color ) ); // make sure "Untitled 8" is the correct title of the original data, containing a column named "color"
// transfer result to associative array result = Associative Array(); // empty A.A. For Each Row(new, result << Insert Item(:color, :N Rows)); // populate A.A. Show(result); // display A.A. to log Close(new, No Save); // close temporary summary table