How to use N missing values in a distribution frequencies table

[#1]

Hello All,

I have a large table where I have empty cells in some places. I am doing a count using the distribution -> frequencies table. However, the probability is based only on the cells with a value inside. The frequencies table still prints out how many values are missing ("N missing"). I was hoping to be able to put that value in the combined table I make from the frequencies table.

Any help is appreciated.

Here is an example of what I am talking about

Ekarshalev_0-1680202078519.png

[#2]

Why not just recode that column so that missing becomes "missing" and then that will appear in the graph as a category?

[#3]

Thx dale,

I have not played with recoding much but it seems like a pretty fast solution.

[#4]

You can easily do this using Transformation variables.

In the dialog box for the Distribution, simply right click and create a virtual transformational variable;  If you specify the formula as in the example, you can use that new virtual variable for the distribution.

txnelson_0-1680204012611.png

To do this is JSL, here is an example

Distribution(
	Transform Column(
		"sex expanded",
		Nominal,
		Set Property( "Value Order", {Numerical Order( 0 )} ),
		Formula( If( :sex == "", "Missing", :sex ) )
	),
	Nominal Distribution( Column( :sex expanded ) )
);

[#5]

Thanks txnelson,

This seems like a pretty easy way to do it.

[#6]

Aside from recoding as mentioned by others, it might be worth while to add to the JMP Wish List a request for an option to display missing values as a category/level in distributions and other places where categorical data is used.

[#7]

Thanks Chris,

I will be sure to do that