How to create a data table of 3 columns per row from numeric string

Question

I'm new to JSL. I have a text string that contains values separated by a "#". The numbers are to be grouped in 3 columns

I know I can write a code to put the values in columns being separated by the "#", but how do I then create a table of 3 values in each row, until the table is complete. I know I'll also probably need to remove the first "#" to prevent issues as first 3 numbers would be in 2nd to 4th column.

For example:

#1#0.93#0.99#2#1.92#1.99#3#2.9#2.99#4#3.86#3.98#5#4.85#4.98

can go to this:

1

0.930.9921.921.9932.92.9943.863.9854.854.98

But how do I write the JSL code to get it to this:

Value1Value2Value310.930.9921.921.9932.92.9943.863.9854.854.98

In this example there are 5 rows in table but the number of rows could vary.

txnelson · Answer

Here is one approach

Names Default To Here( 1 );
theString = "#1#0.93#0.99#2#1.92#1.99#3#2.9#2.99#4#3.86#3.98#5#4.85#4.98";

dt = New Table( "Start",
	add rows( 1 ),
	New Column( "string", character, set each value( theString ) )
);
dt << Text To Columns( delimiter( "#" ), columns( :string ) );
dt << delete columns( :string );
colNames = dt << get column names( string );
For Each( {col}, colNames,
	Column( dt, col ) << Data Type( Numeric, Format( "Best", 12 ) ) <<
	Set Modeling Type( "Continuous" )
);

dt << Stack(
	columns( colNames ),
	Source Label Column( "Label" ),
	Stacked Data Column( "Value" ),
	Number of Series( 3 ),
	Output Table( "Final" )
);

dtFinal = Current Data Table();
dtFinal << delete columns( {:Label, :Label2, :Label3} );
dtFinal:value << set name( "Value 1" );

close( dt, nosave);

jthi · Answer

Use Text To Columns to get this

Then stack your data with multiple series stack

and finally remove extra columns, rename column headers and fix column data/modelling types

JMP created script is a good starting point for the final JSL

// Text to columns
Data Table("Untitled 6") << Text to Columns(columns(:Column 1), Delimiters("#"));

// Stack data table
// → Data Table("Untitled 11")
Data Table("Untitled 6") << Stack(
	columns(
		:Column 1 1, :Column 1 2, :Column 1 3, :Column 1 4, :Column 1 5, :Column 1 6,
		:Column 1 7, :Column 1 8, :Column 1 9, :Column 1 10, :Column 1 11,
		:Column 1 12, :Column 1 13, :Column 1 14, :Column 1 15
	),
	Source Label Column("Label"),
	Stacked Data Column("Data"),
	Drop All Other Columns(1),
	Number of Series(3),
	Output Table("Untitled 11.jmp")
);

// Delete columns
Data Table("Untitled 11") << Delete Columns(:Label, :Label 2, :Label 3);

// Recode column names
Local({dt = Data Table("Untitled 11"), names},
	names = Recode(
		dt << Get Column Names(String),
		{Map Value(
			_rcOrig,
			{"Data", "Value1", "Data 2", "Value2", "Data 3", "Value3"},
			Unmatched(_rcNow)
		)}
	);
	For Each({name, i}, names, Column(dt, i) << Set Name(name));
);

// Standardize column attributes
Local({old dt = Current Data Table()},
	Current Data Table(Data Table("Untitled 11"));
	For Each({col, index}, {:Value1, :Value2, :Value3},
		col << Data Type(Numeric, Format("Best", 12)) <<
		Set Modeling Type("Continuous")
	);
	Current Data Table(old dt);
);