nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

scripting categorical analysis of multiple columns

MFVIT

Hello,

I have a data table (in attachment) with the first column (COD CAMP) that is always the same but the other columns can be different in different tables (different number and different names). I want to script the categorical analysis of the other columns vs. COD CAMP. In this way may analysis will work whatever the data table is (of course it must contain COD CAMP). I tried the following jsl code with several variants but without success. Can anyone help? Many thanks.

dt3 = Current Data Table();
n3 = dt3 << Get Column Names();
Show( n3 );

lstColNames_n3 = dt3 << Get Column Names();
Show( lstColNames_n3 );

Show( lstCols_n3 );
lstCols_n3 = {};
For( i = 1, i <= N Items( lstColNames_n3 ), i++,
    colName = lstColNames_n3[i];
    col = Column( dt3, i );
    Insert Into( lstCols_n3, col );
);
Show( lstCols_n3 );

For( i = 1, i <= N Items( lstColNames_n3 ), i++,
    colName = lstColNames_n3[i];
    col = Column( dt3, colName );
    Insert Into( n2, col );
);

dt3 << New Script(
    "Categorical by COD CAMP",
    Categorical( X( :COD CAMP ), Responses( Eval( n2 ) ), Legend( 0 ) )
);

JAR.jmp

scripting categorical.jsl

Accepted answers

txnelson

Here is a script that produces your original request, using 1 execution of the Categorical platform.

Names Default To Here( 1 );
dt3 = Current Data Table();
n3 = dt3 << Get Column Names();

Remove From( n3, 1, 1 );

theExpr =
			"\[
dt3<<New Script(
		"Categorical by COD CAMP",
		Categorical(
			X( :COD CAMP ),]\";
For( i=1,i<=N Items( n3 ), i++,
	theExpr = theExpr || "\[Name( "Responses,!Responses" )(:]\" || n3[i] || "),";
);

theExpr = theExpr || "Legend( 0 ) ); );";

eval(Parse(theExpr));

All comments

txnelson

Here is a solution to your question.

Most of the code that you supplied repeated the same retrieval of column names multiple times. I suspect you do not understand what the code is doing. The script below simplifies your script.

Names Default To Here( 1 );
dt3 = Current Data Table();
n3 = dt3 << Get Column Names();

Remove From( n3, 1, 1 );

Eval(
	Parse(
		Eval Insert(
			"\[
dt3<<New Script(
		"Categorical by COD CAMP",
		Categorical(
			X( :COD CAMP ),
			Responses(	Eval(^n3^)
			),
			Legend( 0 )
		)
	);
]\"
		)
	)
);

I strongly suggest that you take the time to read the Scripting Guide.

MFVIT

Hello Jim,

I appreciate your help. I am not expert in JSL, I try to extract the code from JMP and from suggestions posted on the JMP community. I am aware that there are some repetitions in the code, sorry for not having cleaned it. This type of coding (probably with some differences because at some point I tried to change different things with a trial and error procedure) worked for some platforms when I have a list of columns (list of references or list of names.. depending on the platform if I remember well) but not for others. For sure I have to go deeper in JSL coding... But so far it has given me a lot of satisfaction!!

I tried the code you wrote but the analysis is performed only for the fist attribute (INTENSITA' PROFUMO IN APPLICAZIONE), not for all of them. I do not receive any error message.

Many thanks for your time,

txnelson

I think the best thing to do, is to start over.

Using your sample data table, can you interactively run the Categorical platform and produce the results you want?

If you can do that, then click on the red triangle and save the script to a script window. That will provide the starting point for what the JSL needs to be to replicate the results for whatever data table you want to run the analysis on.

If you can provide that JSL, I can help you.

I ran the platform interactively and produced the results below. I am not sure this is what you want. Below the output is the JSL required to produce the output.

Categorical(
	X( :COD CAMP ),
	Name( "Responses,!Responses" )(:INTENSITA' PROFUMO IN APPLICAZIONE),
	Name( "Responses,!Responses" )(:QUANTITA' EROGATA),
	Name( "Responses,!Responses" )(:PERSISTENZA PROFUMO DURANTE LA GIORNATA),
	Name( "Responses,!Responses" )(:INTENSITA' PROFUMO A FINE GIORNATA),
	Legend( 0 )
);

MFVIT

I follow exactly what you suggest to capture the script and so far this has
worked fine. I started from this also in this particular project. Then I
started to modify the part where the response names are. In fact, I want to
make a code that works whatever the name and the number of responses are.
So I started to use the concept of list of names and references to
variables. I did this with success with other platforms. In this project I
load the variables from an excel file (it can be one variable or 10 and the
names can change). And I want to perform the categorical analysis on the
variables that are on that excel file. For fit model, text explorer, fit y
by x, I was able to reach my goal. With the categorical I was not.
Thank you very much.