Based on the size of your grid (2x10) it looks like it
is the cell renderer code that is causing problems for you. My
comments about poor performance of treeviews don't apply in your case,
because your treeview is quite small compared with the ones I'm running
into problems with.
You probably need to knock up a version of your code in C so that you
can profile the gtkcellrenderer code and find out where the bottleneck
is.
A couple of other things you could try.
- Use a dictionary instead of that long if/elif
clause to convert letters to the colour strings (I suspect this won't
help much, because I doubt if it is spending much time there).
- Have one column per character in your strings and
use the trick explained in this PyGTK FAQ entry to set the background
colours of each cell:
http://www.async.com.br/faq/pygtk/index.py?req=show&file=faq13.031.htp
-- I'm not sure if this will help or not, but it is worth a try. It
will mean that the renderers are no longer dealing with the markup
stuff, rather they will be getting a single colour from another column
of your model.
- Find some other way of displaying your data.
One option is the solution to all problems, matplotlib:
http://matplotlib.sourceforge.net
I hope one of these works for you.
John
Frederico Arnoldi wrote:
Hi John,
Thanks for reply my question.
Some things you can do to improve
performance:
* set_fixed_height_mode
* use fixed with columns in the view
* use your own custom tree_model.
I have already tried the first two methods, but didn't work. The
third, I have been reading about how to create my own model, but I
didn't try it yet.
Finally, I'm not entirely sure whether your
performance issues are a variation on what I am running into. Can you
post a simple code snippet?
Here is a bit of the code:
# Creating the Treeview #
self.liststore = gtk.ListStore(str,str)
#sequence treeview => _This is the slow Treeview_
self._seqs_treeview.set_model(self.liststore)
seqsselection = self._seqs_treeview.get_selection()
seqsselection.set_mode(gtk.SELECTION_MULTIPLE)
#spp treeview
self._spp_treeview.set_model(self.liststore)
#After opened the desired file and selected the desired rows to
color, the color data is loaded into the Treeview#
(model, paths) = sppselection.get_selected_rows()
tmpliststore = {}
names = []
uncolored = {}
for path in paths:
kiter = model.get_iter(path)
name = model.get_value(kiter,0)
sequncolor = model.get_value(kiter,1)
seq2 =[]
names.append(name)
for letter in sequncolor:#THIS STEP turn my treeview too slow
if letter == "A" :
seq2.append('<span
background="">A</span>')
elif letter == "T":
seq2.append('<span
background="">T</span>')
elif letter == "C":
seq2.append('<span
background="">C</span>')
elif letter == "G":
seq2.append('<span
background="">G</span>')
elif letter == "R":
seq2.append('<span
background="">R</span>')
elif letter == "M":
seq2.append('<span
background="">M</span>')
elif letter == "S":
seq2.append('<span
background="">S</span>')
elif letter == "N":
seq2.append('<span
background="">N</span>')
elif letter == "D":
seq2.append('<span
background="">D</span>')
elif letter == "E":
seq2.append('<span
background="">E</span>')
elif letter == "K":
seq2.append('<span
background="">K</span>')
elif letter == "V":
seq2.append('<span
background="">V</span>')
elif letter == "I":
seq2.append('<span
background="">I</span>')
elif letter == "P":
seq2.append('<span
background="">P</span>')
elif letter == "L":
seq2.append('<span
background="">L</span>')
elif letter == "H":
seq2.append('<span
background="">H</span>')
elif letter == "Y":
seq2.append('<span
background="">Y</span>')
elif letter == "Q":
seq2.append('<span
background="">Q</span>')
elif letter == "F":
seq2.append('<span
background="">F</span>')
elif letter == "W":
seq2.append('<span
background="">W</span>')
else:
seq2.append('<span
background="">%s</span>'%letter)
joinseq=join(seq2)
strseq=str(joinseq)
colorseq = strseq.replace("> <","><")
tmpliststore[name] = colorseq
....
....
#appending the liststore
for seqname in names:
self.liststore.append([seqname,tmpliststore[seqname]])
...
...
self.cellspp = gtk.CellRendererText()
self.cellseqs = gtk.CellRendererText()
....
....
#setting the cell attributes to the appropriate liststore column
self.spp.set_attributes(self.cellspp, markup=0)
self.seqs.set_attributes(self.cellseqs,markup=1)
# end #
How many rows can you put in your tables
before the performance starts to fall off?
Each row contain at least 500 letters, and about seven rows turn
the Treeview slow, and more than 15 turns the Treeview too slow.
I believe that the root of the problem is the great number of
markups, because when I just open the file with only normal text, even
20 rows are not enough to turn the treeview slow. However, when I color
the letters to turn the data analysis easier, the performance problem
apear.
Thanks again,
Fred
Frederico Arnoldi wrote:
I have been working in a tool that display
large sequences of data. Using
normal text it work's fine, but when I substitute the normal text by
markup
text it get to slow. I have already tried set_fixed_height_mode, but
didn't
work. Is it an algorithm or a processor/RAM memory problem?
Is there another way to improve the treeview performance?
_________________________________________________________________
MSN Messenger: converse online com seus amigos .
http://messenger.msn.com.br
|
_______________________________________________
pygtk mailing list [email protected]
http://www.daa.com.au/mailman/listinfo/pygtk
Read the PyGTK FAQ: http://www.async.com.br/faq/pygtk/
