cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Check out the JMP® Marketplace featured Capability Explorer add-in
Choose Language Hide Translation Bar
noviJSL
Level II

Parse a concatenated string which has comma

Hi all,

I have a datatable with multiple columns, some of which I want to add together. As a part of my script, I am selecting some of the columns, storing the column names in a list and then hoping to add the selected columns using the "Sum" feature.

 

In order to do that I need to build a string ":DATA1, :DATA2, :DATA3" and Parse it in a column. 

New Column( "Addition", Numeric, Continuous, Formula( Parse( Sum( :DATA1, :DATA2, :DATA3 ) ) ) );

 

A snippet of the code I am using is below:

New Column( "Addition", Numeric, Continuous, Formula( Parse( Sum( :DATA1, :DATA2, :DATA3 ) ) ) );
list = {"DATA1", "DATA2", "DATA3"};
For( i = 1, i <= N Items( list ), i++,
    If(
        i == 1, string = ":" || Char( Eval( list[Eval( i )] ) ),
        i > 1, string = string || "," || ":" || Char( Eval( list[Eval( i )] ) )
    )
);
Parse( string );

 

On execution, I get the following error:

Unexpected ",". Perhaps there is a missing ";" or ",".
Line 1 Column 7: :DATA1►,:DATA2,:DATA3

The remaining text that was ignored was
,:DATA2,:DATA3
:DATA1

 

Is there anything I am missing?

Thanks in advance!

 

 

1 ACCEPTED SOLUTION

Accepted Solutions
vince_faller
Super User (Alumni)

Re: Parse a concatenated string which has comma

Look at the string in which it returns. 

":DATA1,:DATA2,:DATA3"

 

Try running just that without the quotes, you'll get the same thing.  

If you have column names as strings, you can do something like this.  

 

Names default to here(1);
dt = New Table( "Example",
	Add Rows( 1 ),
	New Column( "Data1", Set Values( [1] )),
	New Column( "Data2", Set Values( [2] )),
	New Column( "Data3", Set Values( [3] )),
);
// list of column names as strings
list = dt << Get Column Names("string"); // {"Data1", "Data2", "Data3"}
SumExpr = Expr(Sum());
for(i=1, i<=nitems(list), i++, 
	insert into(SumExpr, Column(dt, list[i]));
);
// see what the sumExpr looks like
show(nameexpr(sumExpr));
// then you have to do some substitute in order to actually get it to run
just_for_show = EvalExpr(
	dt << New Column("Another Data", Formula(Expr(nameexpr(SumExpr))));
);
show(nameexpr(just_for_show));
just_for_show();// this will actually run the expression you just created. 

// or you could use the list in Sum Directly. 
// list of column names as references
list = dt << Get Column Names();
list=list[1::3];
// can do a list directly in Sum()
// should still substitute it in so it's no dependant on then list variable
just_for_show =Substitute(
	Expr(
		dt << new Column("Yet Another Data", Formula(Sum(DV_LIST)))
	), 
	Expr(DV_LIST), list
);

show(nameexpr(just_for_show));
just_for_show(); // actually evaluates again.  
Vince Faller - Predictum

View solution in original post

2 REPLIES 2
vince_faller
Super User (Alumni)

Re: Parse a concatenated string which has comma

Look at the string in which it returns. 

":DATA1,:DATA2,:DATA3"

 

Try running just that without the quotes, you'll get the same thing.  

If you have column names as strings, you can do something like this.  

 

Names default to here(1);
dt = New Table( "Example",
	Add Rows( 1 ),
	New Column( "Data1", Set Values( [1] )),
	New Column( "Data2", Set Values( [2] )),
	New Column( "Data3", Set Values( [3] )),
);
// list of column names as strings
list = dt << Get Column Names("string"); // {"Data1", "Data2", "Data3"}
SumExpr = Expr(Sum());
for(i=1, i<=nitems(list), i++, 
	insert into(SumExpr, Column(dt, list[i]));
);
// see what the sumExpr looks like
show(nameexpr(sumExpr));
// then you have to do some substitute in order to actually get it to run
just_for_show = EvalExpr(
	dt << New Column("Another Data", Formula(Expr(nameexpr(SumExpr))));
);
show(nameexpr(just_for_show));
just_for_show();// this will actually run the expression you just created. 

// or you could use the list in Sum Directly. 
// list of column names as references
list = dt << Get Column Names();
list=list[1::3];
// can do a list directly in Sum()
// should still substitute it in so it's no dependant on then list variable
just_for_show =Substitute(
	Expr(
		dt << new Column("Yet Another Data", Formula(Sum(DV_LIST)))
	), 
	Expr(DV_LIST), list
);

show(nameexpr(just_for_show));
just_for_show(); // actually evaluates again.  
Vince Faller - Predictum
noviJSL
Level II

Re: Parse a concatenated string which has comma

Thank you for helping and the explanation! The "Yet Another Data" column in your script does exactly what i want. The "Another Data" column seems to be returning an empty cell with an error "need arguments to be scalars, matrices, or lists..". I will try to figure out why this is happening.