Viewing huge trees

26 views
Skip to first unread message

Connor Skennerton

unread,
Mar 15, 2015, 4:21:10 PM3/15/15
to archae...@googlegroups.com
I tried to view the SILVA 119 16S rRNA tree, which contains approximately 900,000 tips and the rendering was incredibly slow for scolling or zooming. Does anyone have a sense of what the maximum size trees is before the rendering engine starts to really slow down? Are there any configuration options that might help speed things up? I already tried increasing the memory. 

Christian

unread,
Mar 15, 2015, 4:27:34 PM3/15/15
to archae...@googlegroups.com
I would guess that 200,000 external nodes is about the max Archaeopteryx can reasonable handle, before performance comes to a crawl. I will test this soon...

Connor Skennerton

unread,
Mar 23, 2015, 6:57:18 PM3/23/15
to archae...@googlegroups.com
Hi Christian,

I did some testing of my own with randomly generated trees with 50000, 100000, 200000 and 500000 leaves. I would say that with 100000 leaves there was a bit of a slow down and then at 200000 (as you predicted) there was a big slow down in rendering. 

My interest is that I'm trying to find a viable replacement to ARB, which is a huge pain to install and use, yet is still the most capable open source tree viewer for huge trees. You know the codebase of Archaeopteryx - what do you think it would take to improve performance to the point where such large trees could run smoothly? I'm happy to try things out with the code myself but I'd love your opinion on what parts of the codebase could increase the performance; or whether you think such an endeavor was foolhardy. 

Christian Zmasek

unread,
Mar 31, 2015, 3:31:32 PM3/31/15
to archae...@googlegroups.com
Hi, Connor:

I think I did most I could to improve Archaeopteryx' performance.

To improve it further in a significant manner, I am afraid a move from Java Swing to OpenGL to utilize modern GPUs would be necessary. Sadly, I lack both expertise and time to do so.

Christian





--
--
You received this message because you are subscribed to the Google
Groups "Archaeopteryx" group.
To unsubscribe from this group, send email to
archaeoptery...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/archaeopteryx?hl=en
---
You received this message because you are subscribed to the Google Groups "Archaeopteryx" group.
To unsubscribe from this group and stop receiving emails from it, send an email to archaeoptery...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply all
Reply to author
Forward
0 new messages