Ok, I did some more testing.
First of all, now I've run all tests with the jmol app, not the applet.
Then I tested the following:
-Paint 500 bps with a script like the one I pasted on my last email
-Loading the full script at once: 17 seconds
-Splitting the script each 100 lines, then loading all the files: 12
sec
So the gain isn't very notable.
More testing:
Instead of loading some residues at once and then painting all of
them, I changed the script to select each residue individually and
then change its colour;
select C1; color [0,0,0];
select N2; color [0,0,0];
select C3; color [0,0,0];
...
select N100; color [0,10,0];
select C101; color [0,10,0];
...
Then, the running time of the full script is 42 seconds (2.5 times
slower). As the color commands are the same, I suppose that the issue
is related do the selects.
Now for a bigger DNA (2000 bps)
-with the script split each 100 lines: 2min 26sec
-the full script (3000+ lines): i cancelled at 8 minutes...
So, clearly, there is an issue with the selects. I think that the cost
of the function isn't lineal but more like exponential, as multiplying
the residues number by four doesn't multiply the script running time
by four but by 12!
How does jmol paint the molecule on load? Clearly it goes a different
way, because the initial load+paint is just 5-6 seconds
Thanks for the help,
Carles
El 10/01/2008, a las 12:46, Angel Herraez escribió:
> Hi Carles
>
> This sound strange indeed. I agree that it shouldn't take long.
>
> I think that maybe the 100 kB is too long for some unknown memory,
> stack or cache (I'm just guessing). Try cutting the script file in
> several pieces and see what happens when you call them one after
> another (either manually or maybe with a master script)
>
> Another idea: test this in the Jmol application.
>
>
> -------------------------------------------------------------------------
> Check out the new SourceForge.net Marketplace.
> It's the best place to buy or sell services for
> just about anything Open Source.
> http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
> _______________________________________________
> Jmol-users mailing list
> Jmol-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/jmol-users
--
Carles Fenollosa Bielsa
Barcelona Supercomputing Center / Life Sciences
Nexus II / Jordi Girona, 29 / 08034 Barcelona
+34 93 413 76 02
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
