[S] Large (sparse) matrix is killing me

Arman Maghbouleh (arman@csli.stanford.edu)
Thu, 22 Oct 1998 19:13:33 -0700 (PDT)

Are there any existing functions for working with large sparse

One matrix I am dealing with is 18698x614 and will contain floating
point numbers. At 12 bytes per float this translates into 137,766,864
bytes which brings my Sun Ultra with 120 megs RAM, and 400 megs virtual
memory (S-PLUS 3.4) to a halt. I expect only 150 non-zero values per row.

It would be great if there were some object oriented code that
allowed me to use the usual matrix indexing mechanisms such as []. If
there is nothing available, I'll just store values in each row as a
list and write limited access functions.

As usual, I will post a summary of any replies sent directly to me.

Thanks, Arman.

Arman Maghbouleh Linguistics Dept. Stanford University

This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news