Re: Hashing for big list of last names

*** top-posting fixed ***
sprash wrote:
Ben Pfaff wrote:
sprash25@xxxxxxxxx writes:

I would like to design a hash table for a big list of last names
of patients.

Any recommendations on the hashing type (chaining, open
addressing, double hashing) and the function I should use?

It entirely depends on your constraints. If you describe your
application in some more detail, perhaps we can give
recommendations.

To put it simply, the module I am writing essentially reads the
list of last names of Patients from a CSV file which is a very
large list. Next, based on the inputs I get from another
application, I need to manipulate this list by searching and
inserting new last names and also delete certain last names from
my list.

There is existing legacy code that is implementing this using
chaining ( function is based on the first character of the last
name). This causes a lot of collision for common characters such
as A, D etc and very few for X, Q etc. Basically the distribution
is understandably uneven.

I am wondering if there are any better ideas out there.

To all practical purposes it is all done for you, and available
under GPL. See:

<http://cbfalconer.home.att.net/download/>

and look for the hashlib and id2id-20 packages.

--
"I have a creative mind. You (singular) are eccentric.
He is insane. We are losing sight of reality.
You (plural) are smoking crack. They are certifiable."
Declension of verbs, per Lewin Edwards



--
Posted via a free Usenet account from http://www.teranews.com

.

Relevant Pages