topic Re: help! in SAS Programming

help!

ayeshasiddique — Mon, 24 Aug 2015 15:38:30 GMT

I have a dataset that was made on sas and now its handed to me in the sas coded form on pdf with the column names as follows:

I need help in interpreting this data. What do the labels mean and how do i interpret it?

Num	Variable	Type	Len	Format	Informat	Label

1	NEWID	Num	8	Z4.		Patient ID (Pseudonym)
2	VISIT	Char	2	$2	$2	VISIT
3	FORM13ND	Num	8	NOTDONE	2	FORM 13 NOT DONE
4	CHOLOWO	Num	8	YN	2	fm13q01a: Cholesterol lowering (e.g. statin) on-going
5	CHOLOWN	Num	8	YN	2	fm13q01b: Cholesterol lowering (e.g. statin) newly prescribed

Re: help!

Reeza — Mon, 24 Aug 2015 16:02:41 GMT

Please review the guidelines on how to frame a question:

When you say labels, what are you referring to, the column label or the column names?

If the label column it is the Variable Label that is viewed in the SAS data set.

A SAS variable has a variable name NEWID, and that variable has a descriptive label Patient ID (Pseudonym). The variable label is useful for presentation and describing the variable contents.

Re: help!

ayeshasiddique — Mon, 24 Aug 2015 16:08:35 GMT

so sorry for not making this clear. i mean to ask what do each column headings mean. and what does the data in each column mean?

I know the meaning of

num

variable

type

Label

But what do the following mean?

Len

Format

Informat

Also the data in these columns seems coded...i need help in interpreting that?

Re: help!

ballardw — Mon, 24 Aug 2015 17:07:12 GMT

The information you post looks to be a description of the variables in a dataset. Label in this case is the text that describes (hopefully) the contents of the associated variable. This label text will appear in many procedures when the variable is used.

From what I see you have as example: your data set has a variable named CHOLOWO, which is how to refer to it in code, that is numeric, has an associated format named YN, which I would guess is two valued likely 1 and 0, and is a flag that indicates whether the patient identified with NEWID had been prescribed a statin prior to this visit and likely the information is collected on a Form 13 as question 1, field A (complete guess: first check box for question 1 was checked/unchecked).

Re: help!

Tom — Mon, 24 Aug 2015 17:11:18 GMT

NUM is most likely the order of the variable in the table.

NAME is the variable name. This is the value you use in your programs to reference variable.

TYPE tells if the variable is a fixed length character string or floating point number.

LENGTH is really only useful for character variables as it sets the maximum length.

FORMAT is how the data is displayed. The first variable is using the Z4. format so values will be displayed with 4 digits with leading zeros. The $2. format attached to the character variable is not really needed as it does not change the default behavior. The other variables have custom or user formats attached. You will need to find separate documentation for how those are defined.

INFORMAT is how input is converted, does not normally matter for existing data since you will not be reading in new data from a raw text file.

LABEL is just with is says. It is a label that is applied to the variable. It is normally displayed in reports and to help the user understand what the variable means. Your guess is probably better than ours what the labels mean in the context of the data you are being provided.