topic Re: Calculation Result Discrepancy, Col Mean, etc. in Discussions

Calculation Result Discrepancy, Col Mean, etc.

TamedZebra — Fri, 30 Jan 2026 01:17:54 GMT

Why do calculation results differ when using variables compared to using direct strings? I suspect that when variables are used, the grouping function for the variable part might not be working. How can this be resolved?

with the following formula

Col Mean(
	:data,
	:A,
	:B,
	:C_D == "ccccccc"
)

This is the formula.

Col Mean(
	:data,
	:A,
	:B,
	Column(Eval Insert("C_^get_input_1^" == get_input_2))
)

Re: Calculation Result Discrepancy, Col Mean, etc.

TamedZebra — Fri, 30 Jan 2026 01:22:55 GMT

Sorry, everyone. I forgot to mention that get_input_1 and get_input_2 can be assumed to be already obtained as strings, respectively.

Re: Calculation Result Discrepancy, Col Mean, etc.

jthi — Fri, 30 Jan 2026 05:29:49 GMT

First one is comparing column vs string and second one we cannot be sure what is being compared (string to string which would always be 1 or 0?). At least it isn't evaluated and you will need to evaluate the values into the formula

Re: Calculation Result Discrepancy, Col Mean, etc.

TamedZebra — Fri, 30 Jan 2026 08:31:41 GMT

I want to achieve the same result as the expression in the first line. When the variables were given as get_input_1="D" and get_input_2="cccccc" respectively, I naturally expected the same result, but the reality was different...

Re: Calculation Result Discrepancy, Col Mean, etc.

jthi — Fri, 30 Jan 2026 09:36:19 GMT

Eval(Substitute(
	Expr(Col Mean(
		:data,
		:A,
		:B,
		_input1_ == _input2_
	),
	Expr(_input1_) = NameExpr(AsColumn(dt, get_input_1)),
	Expr(_input2_) = get_input_2
	)
));

but you need to add new column and formula to that example

Re: Calculation Result Discrepancy, Col Mean, etc.

mmarchandFSLR — Fri, 30 Jan 2026 12:56:15 GMT

~~Column(Eval Insert("C_^get_input_1^" == get_input_2)) evaluates to Column( "C_D" == "ccccccc" ), so your formula would be~~

edit

Including non-string inside Eval Insert() actually causes it to evaluate to the empty string.

Col Mean(
	:data,
	:A,
	:B,
	Column( "" )
);

That may be why it isn't working as expected. Jarmo's solution should be what you want.

Re: Calculation Result Discrepancy, Col Mean, etc.

TamedZebra — Mon, 02 Feb 2026 00:29:31 GMT

Thank you for your replies. Also, I apologize for my late responses.

I wrote the JSL as shown below, but an error is being output. What could be the cause?

The error message says "insufficient arguments" at the location "Substitute/###/(Expr(".

	Data Table( "data_table" ) << New Column( Eval Insert("Xbar_^get_input_process^_^get_input_line^"),
		Numeric,
		"Continuous",
		Format( "Best", 12 ),
		Formula(
			Eval(Substitute(
				Expr(Col Mean(
					:data,
					:A,
					:B,
					_input1_ == _input2_
					),
					Expr(_input1_) = NameExpr(AsColumn(dt, Eval Insert("new_^get_input_1^"))),
					Expr(_input2_) = get_input_2
					)
					));
	)) << Run Formulas;

Re: Calculation Result Discrepancy, Col Mean, etc.

jthi — Mon, 02 Feb 2026 05:50:37 GMT

Most likely something like this, but I cannot test it as I don't have data or the earlier part of the script

Names Default To Here(1);


dt = Data Table("data_table");

Eval(Substitute(
	Expr(Data Table("data_table") << New Column(colname,
		Numeric,
		"Continuous",
		Formula(Col Mean(:data, :A, :B, _input1_ == _input2_)),
	)),
	Expr(colname), Substitute(Xbar_^get_input_process^_^get_input_line^),
	Expr(_input1_) = NameExpr(AsColumn(dt, Eval Insert("new_^get_input_1^"))),
	Expr(_input2_) = get_input_2
));

Re: Calculation Result Discrepancy, Col Mean, etc.

TamedZebra — Mon, 02 Feb 2026 08:15:34 GMT

Thank you for your consideration, even though it was difficult to experiment without being able to provide actual data.

I modified it as follows to fit my data table.

Names Default To Here(1);


dt = Data Table("data_table");

Eval(Substitute(
	Expr(Data Table("data_table") << New Column(colname,
		Numeric,
		"Continuous",
		Formula(Col Mean(:data, :A, :B, _input1_ == _input2_)),
	)),
	Expr(colname), Eval Insert(Xbar_^get_input_process^_^get_input_line^),
	Expr(_input1_) , NameExpr(AsColumn(dt, Eval Insert("new_^get_input_1^"))),
	Expr(_input2_) , get_input_2
));