[Purdue-pm] PM meeting this Tuesday, June 15th
Mark Senn
mark at ecn.purdue.edu
Thu Jun 10 09:47:53 PDT 2010
Rick> We do have several challenge problems out on the table. The
Rick> most recent is from Joe and is summarized below. I had
Rick> suggested that the fastest way to access the data was via an
Rick> array. And indeed Mark sent out an elegant [Perl 5 -mark]
Rick> solution that used an array.
I'll try to send a Perl 6 solution to the list before the next meeting.
Rick> I am not sure if we can do
Rick> better than that. That is the solution is O(1) ignoring, as is
Rick> typical, setup time. And space is O(x) where 'x' is the size of
Rick> the array. However what if we have a very large and sparse
Rick> array? And if we do wish to consider setup time plus minimizing
Rick> memory space? Then what type of solution can be found?
Doing a Google search for
perl sparse arrays
gave around 20,900 results. I thought (next line over 80 chararacters)
http://etutorials.org/Programming/Perl+tutorial/Chapter+4.+Arrays/Recipe+4.4+Implementing+a+Sparse+Array/
and
http://stackoverflow.com/questions/863426/counting-array-elements-in-perl
were helpful. (Mathematica---my second favorite programming language,
has sparse arrays built-in.)
Rick> And if we do wish to consider setup time plus minimizing
Rick> memory space? Then what type of solution can be found?
I don't want to consider it---too hypothetical in the context of the
simple challenge problem. Use hashes for the 1D case. Maybe spin this
off into a 2D or more challenge problem to better investigate setup time and
minimizing memory space? -mark
More information about the Purdue-pm
mailing list