turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-10-2011 06:39 PM

I have been looking for a command in SAS/IML which finds the location of the matched values in two different vectors.

I came across this website: http://stackoverflow.com/questions/4952548/value-matching-in-sas-iml where the same question was asked and the suggested answer was to use XSECT function to find the intersection of the two sets. However, XSECT function gives the common elements between two sets rather that giving the location of those elements.

For example, in the following code:

**proc iml**;

x={**10 11 ****12 13 ****12 13 ****14 15**};

v={**12 13**};

z=xsect(x,v);

print z;

**quit**;

the SAS result is: Z= {12 13}, but I’mlooking for a vector like {3 4 5 6} which is the location of common elements of vector V in vector X.

Would anybody help me with the appropriate function inSAS/IML?

Thank you very much.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Reza

11-10-2011 08:39 PM

In SAS 9.3, the function you want is the ELEMENT function teamed up with the LOC function:

idx=loc(element(x,v));

If you haven't upgraded to SAS 9.3, you can use the UNIQUE-LOC technique, which I recently blogged about.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Reza

11-11-2011 06:51 AM

For example, here's one one approach:

u = unique(v);

free idx;

do i = 1 to ncol(u);

k = loc(x=u*);*

idx = idx || k;

end;

print idx;

This code isn't maximally efficient because it doesn't preallocate the IDX array (see http://blogs.sas.com/content/iml/2011/06/20/pre-allocate-arrays-to-improve-efficiency/).

Still it's not too bad. Use this if your vectors aren't too long. If this step is being used in the innermost loop of a long-running computation, write back. The code can be improved, but at the cost of additional complexitites.