[Cdk-user] Re: RTool Algo-Explanation?

[Irilenia N. <iri...@kc...>]

Dear Sneha,

The reference listed in the Java code is not correct. In fact Thierry  
who is a co-author of the code has told me that he was going to  
publish it, but I think he may be too busy in his job to do this any  
time soon. To my knowledge there is no publication of the algorithm  
(and hence, it may be time to update the java doc).
When I used the Java classes referring to this algorithm, I came  
across a few bugs - we discussed them in the list last year and I  
believe they are probably taken care of (although I had used a  
modified version of the classes because I was in a hurry at the  
time). I think however, that RTools may not be updated. If I remember  
well, its presence was redundant as individual classes had all the  
methods needed, so there was some duplication of code there. It may  
well be that RTools is not updated and hence the bugs are still there  
- I am not sure, and at the moment I do not have the time to check,  
but it may be worth looking at the code of the  
UniversalIsomorphismTester class and see whether this is more up to  
date.

I may be a bit out of date myself with the advice I'm giving (since I  
used this last time about six months ago, and certainly before the  
last release), so please take this advice carefully.

Good luck with it,

Irilenia

p.s. The fact that the code is not published is not quite correct - I  
think it exists inside Thierry's PhD thesis, but this is not that  
easy to get hold of, and it is in french. When I wanted to understood  
what it was doing, I had no option but to go through the code and  
understand the main steps of it (which is not too difficult to do),  
but I would have preferred myself an explanation on paper.

-----------------------------
Irilenia (Irene) Nobeli
Randall Division of Cell and Molecular Biophysics
New Hunt's House (room 3.14)
King's College London, Guy's Campus
London, SE1 1UL
U.K.
iri...@kc...
+44(0)207-8486329