Serial Attached SCSI | Technology Tales

Dealing with variable length warnings in SAS 9.2

11^th January 2012

A habit of mine is to put a LENGTH or ATTRIB statement between DATA and SET statements in a SAS data step to reset variable lengths. By default, it appears that this triggers truncation warnings in SAS 9.2 or SAS 9.3 when it didn't in previous versions. SAS 9.1.3, for instance, allowed you to have something like the following for shortening a variable length without issuing any messages at all:

data b;
    length x $100;
    set a;
run;

In this case, x could have a length of 200 previously and SAS 9.1.3 wouldn't have complained. Now, SAS 9.2 and 9.3 will issue a warning if the new length is less than the old length. This can be useful to know, but it can be changed using the VARLENCHK system option. Though the default value is WARN, it can be set to ERROR if you really want to ensure that there is no chance of truncation. Then, you get error messages and the program fails where it normally would run with warnings. Setting the value of the option to NOWARN restores the type of behaviour seen in SAS 9.1.3 and versions before that.

The SAS documentation says that the ability to change VARLENCHK can be restricted by an administrator, so you might need to deal with this situation in a more locked down environment. Then, one option would be to do something like the following:

data b;
    drop x;
    rename _x=x;
    set a;
    length _x $100;
    _x=strip(x);
run;

While It's a bit more laborious than setting the VARLENCHK option to NOWARN, the idea is that you create a new variable of the right length and replace the old one with it. That gets rid of warnings or errors in the log and resets the variable length as needed. Of course, you have to ensure that there is no value truncation with either remedy. If any is found, then the dataset specification probably needs updating to accommodate the length of the values in the data. After all, there is no substitute for getting to know your data and doing your own checking should you decide to take matters into your hands.

There is a use for the default behaviour, though. If you use a specification to specify a shell for a dataset, then you will be warned when the shell shortens variable lengths. That allows you to either adjust the dataset or your program. Also, it provides additional information when you get variable length mismatch warnings when concatenating or merging datasets. There was a time when SAS wasn't so communicative in these situations and some investigation was needed to establish which variable was affected. Now, that has changed without leaving the option to work differently if you so do desire. Sometimes, what can seem like an added restriction can have its uses.

Transferring data between SAS and R

5^th June 2008

A question regarding the ability to transfer of data between SAS and R set me off on a spot of investigation a while back, and I have always planned to share the results of my labours. Once I managed to locate the required documentation, things became clearer with further inspection. Functions from the foreign package seem to offer the most from the data import and export point of view, so they're what I'll be featuring in this posting.

Here, I am starting with importing, and using the read.ssd function makes life so much easier for getting SAS data into R. When I discovered that the foreign package may not be loaded by default, that could be determined easily using the following command:

search()

If package:foreign isn't in the list, then you need to issue the following function call:

library(foreign)

Of course, if the foreign package isn't installed, none of this will work. It should live in the library sub-folder of the main R installation directory, but if it isn't there, then downloading the relevant binary package from CRAN is in order. Assuming that all is installed, then a command like the following will perform the needful:

read.ssd("c:/data","data1",sascmd="C:/Program Files/SAS Institute/SAS/V8/sas.exe")

This creates a temporary SAS program that converts the SAS data set into a transport file for reading by another R function that is called in the background, read.xport. From my experience, it all seems to work fairly seamlessly.

To get data out of R and into SAS is a multi-stage process, even with the foreign package. While there are other ways, using the write.foreign seems more useful than most. Here is an example function call:

write.foreign(data1,"C:/test.txt","C:/test.sas",package="SAS",dataname="data1",validvarname="V7")

While no SAS data sets are created at this stage, a text file is generated along with a SAS program for converting it into a data set. Running the SAS program is a separate step that follows the creation of the two files. Even if it is less streamlined than read.ssd, write.foreign does make it easier to transfer data into SAS than having to write a program from scratch to read in write.table output.

In summary, R can neither read nor write SAS data sets by itself, so you need SAS installed to really make things happen. SAS gets called by read.ssd and I feel that it would be better if was called by write.foreign also rather than a SAS program generated for execution later on. Even so, it is good to see some custom functionality being provided that makes life easier. There's also the hmisc package, but my experiences while working with that on S-Plus have been such that it compares less favourably with foreign on the reliability front. Saying that, things may have changed since I last tried it.

Porting SAS files to other platforms and versions

1^st October 2007

SAS uses its transport file format to port files between operating and, where the need arises, different software versions. As with many things, there is more than one method to create these transport files: PROC CPORT/CIMPORT and PROC COPY with the XPORT engine. The former method is for within version transfer of SAS files between different operating systems (UNIX to Windows, for instance) and the latter is for cross-version transfer (SAS9 to SAS 8, for example). SAS Institute has a page devoted to this subject which may share more details.