[Biopython] Bio.trie

Ruchira Datta ruchira.datta at gmail.com
Wed Jan 5 03:10:53 UTC 2011


I had also seen that.  Sorry for my delay in replying.  A trie is an
important data structure with many uses.  The use I had in mind is: tries
are the lowest-latency way of implementing an autosuggest/autocomplete
(e.g., if you want to allow users to pick a member of the NCBI taxonomy by
scientific name).

--Ruchira

On Tue, Jan 4, 2011 at 6:59 PM, Andrew Dalke <dalke at dalkescientific.com>wrote:

> On Dec 29, 2010, at 4:24 AM, Michiel de Hoon wrote:
> > We would like to know though how many users Bio.trie has, so we can
> decide whether it is worthwhile to update this module. If you are using
> Bio.trie, please let us know (preferably via the mailing list). If there are
> no current users, I suggest that we deprecate and later remove this module
> from Biopython.
>
> I am not a user but the other day I was looking through the Python bug list
> and came across:
>
>   http://bugs.python.org/issue9520
>
>   The best existing implementation I've been able to find so far
>   is one in the BioPython. Compared to defaultdict(int) on the
>   task of counting words. Dataset 123,981,712 words (6,504,484
>   unique), 1..21 characters long:
>     * bio.tree - 459 Mb/0.13 Hours, good O(1) behavior
>     * defaultdict(int) - 693 Mb/0.32 Hours, poor, almost O(N) behavior
>
>
>
>                                Andrew
>                                dalke at dalkescientific.com
>
>
>
> _______________________________________________
> Biopython mailing list  -  Biopython at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biopython
>



More information about the Biopython mailing list