[kepler-dev] Token.toString method(s)

Christopher Brooks cxh at eecs.berkeley.edu
Mon Feb 27 07:11:35 PST 2006

I just checked in changes to DoubleMatrixToken and MatrixToken that
should do the trick.  I ended up using a Set, and I don't allocate it
unless there is a nil token.  We are definitely introducing overhead
in a few places, but I tried to be smart about it by only allocating
the set when we have a nil token.

We should probably review this code asap, perhaps Tuesday.
Maybe some of the Keplerites would like to call in? We would
review 1-2pm on Tuesday.  I'd probably review
Token, DoubleToken, ArrayToken, MatrixToken and DoubleMatrixToken
We'd review just the methods that involve nil.
If there is sufficient interest, I could set up a conference call.


    At 01:07 AM 2/26/2006, Christopher Brooks wrote:
    >One issue here is that it would be nice if when we
    >have a DoubleMatrix that has a nil value, then if
    >we print it out and then read it back in with an
    >ExpressionReader, then we get the same matrix with
    >a nil value.  This points to printing nil as nil
    >not NaN.  So, we will need to keep a list of which
    >elements are nil. We probably want something fast,
    >like an ArrayList here, which as roughly O(n) time for
    >adds,  The ArrayList docs say:
    I hope this can be done in a way that doesn't
    introduce overhead when nil is not in use (the
    common case).  The intent of matrix tokens (vs.
    array tokens) is to maximize efficiency of numerical
    Edward A. Lee
    Professor, Chair of the EE Division, Associate Chair of EECS
    231 Cory Hall, UC Berkeley, Berkeley, CA 94720
    phone: 510-642-0253 or 510-642-0455, fax: 510-642-2845
    eal at eecs.Berkeley.EDU, http://ptolemy.eecs.berkeley.edu/~eal  

More information about the Kepler-dev mailing list