gato-user Mailing List for Gato (Graph Animation Toolbox)

Hello,

I have a question with reference to Gato and it's possibilities to display graphs. I would like to show the user an overview what each colour in the running algorithm demonstrates. Is there an uncomplicated way to put a frame with description or to open a second window with coloured text
(edge or vertex colour)? 
I would be glad if you could tell me if this is possible oder even if you still found a way to do this.


Best regards,
Bernd


-- 
NEU: FreePhone - kostenlos mobil telefonieren!			
Jetzt informieren: http://www.gmx.net/de/go/freephone

On Mar 31, 2011, at 8:35 AM, Bernd Perscheid wrote:
> I've got a problem with gato when I want to start an algorithm.
> I open the algorithm and a graph, and when I press the Start-Button I get
> the message "No handlers could be found for logger "Gato"".
> The system I use is Windows 7.

Which version of Gato (SVN, binaries from http://schliep.org) and which 
Python-Version is that?

If you start Gato.py from the command line, you can use the additional flags
-d -v which make Gato more talkative, e.g. in the directory containing Gato.py

   C:\path-to-python\python Gato.py -v -d 


Best,
Alexander

Hello,

I've got a problem with gato when I want to start an algorithm.
I open the algorithm and a graph, and when I press the Start-Button I get
the message "No handlers could be found for logger "Gato"".
The system I use is Windows 7.
Could you please tell me what to do to make the program work?

Best regards,
Bernd
-- 
NEU: FreePhone - kostenlos mobil telefonieren und surfen!			
Jetzt informieren: http://www.gmx.net/de/go/freephone

Alexander Schliep wrote:

>
> On Tue, 22 Feb 2005, David Grant wrote:
>
>> I tried using an .alg and .pro files to colour the nodes and edges 
>> but I couldn't get it to work properly.
>
>
> An empty prolog file (I will make that optional at some point) and the 
> following algorithm work:
>
> # Use SetAllVerticesColor
> w = G.edgeWeights[0] # The first set of weigths
>
> for v in G.Vertices():
>     A.SetVertexColor(v,'red')
>
> for u,v in G.Edges():
>     print "w[(%d,%d)] = %f" % (u,v,w[(u,v)])
>     A.SetEdgeColor(u,v,'red')
>
>
Hey it worked great and the debugging mode was helpful. I have just 
implemented the following in the last half an hour:

w = G.edgeWeights[0] # The first set of edge weights
n = G.vertexWeights[0] # The first set of node weights

defaultNodeColour = 'black'
nodeColours = {1: 'blue', 0: 'red'}
defaultEdgeColour = 'black'
edgeColours = {1: 'blue', -1: 'red'}

for v in G.Vertices():
    try:
        A.SetVertexColor(v,nodeColours[int(n.label[v])])
    except:
        A.SetVertexColor(v,defaultNodeColour)

for u,v in G.Edges():
    try:
        A.SetEdgeColor(u,v,edgeColours[int(w[(u,v)])])
    except:
        A.SetEdgeColor(u,v,defaultEdgeColour)

It colour my nodes/edges according to a dictionary and it has a default 
for nodes/edges which don't have acceptable weights.

Thanks again,

David

Alexander Schliep wrote:

>
> On Tue, 22 Feb 2005, David Grant wrote:
>
>> I tried using an .alg and .pro files to colour the nodes and edges 
>> but I couldn't get it to work properly.
>
>
> An empty prolog file (I will make that optional at some point) 

Good idea to use an empty prolog file, I was thinking that, but didn't 
know if it would work.

> and the following algorithm work:
>
> # Use SetAllVerticesColor
> w = G.edgeWeights[0] # The first set of weigths
>
> for v in G.Vertices():
>     A.SetVertexColor(v,'red')
>
> for u,v in G.Edges():
>     print "w[(%d,%d)] = %f" % (u,v,w[(u,v)])
>     A.SetEdgeColor(u,v,'red')
>
Thanks, I will try it out in a bit.

> If you start ./Gato.py -v you will actually see error messages in the 
> terminal.
>
Thanks. Not being able to debug my program was one of the difficulties I 
had.


David

On Tue, 22 Feb 2005, David Grant wrote:
> I tried using an .alg and .pro files to colour the nodes and edges but I 
> couldn't get it to work properly.

An empty prolog file (I will make that optional at some point) and the 
following algorithm work:

# Use SetAllVerticesColor
w = G.edgeWeights[0] # The first set of weigths

for v in G.Vertices():
     A.SetVertexColor(v,'red')

for u,v in G.Edges():
     print "w[(%d,%d)] = %f" % (u,v,w[(u,v)])
     A.SetEdgeColor(u,v,'red')


If you start ./Gato.py -v you will actually see error messages in the 
terminal.

best,
A:




--
Dr. Alexander Schliep                                    sc...@mo...
Group Leader Algorithmics, Dept. Comp. Mol. Bio        phone: +49-30-8413-1166
Max Planck Inst. for Molecular Genetics                 fax: + 49-30-8413-1152
Ihnestrasse 73, 14195 Berlin, Germany        http://algorithmics.molgen.mpg.de

Alexander Schliep wrote:

>
> On Mon, 21 Feb 2005, David Grant wrote:
>
>>> Sorry about that. I found a script at 
>>> http://mail.python.org/pipermail/pythonmac-sig/2004-December/012239.html 
>>>
>>>
>>> which works. I will convert to all spaces.
>>
>
> It is checked in now.

Thanks.

>
>> I just did the same using boa-constructor. In its file menuu there is 
>> something called "re-indent entire file." It worked like a charm. I'm 
>> following the instructions on cvshome.org's cvs manual on maintaining 
>> 3rd party sources and a modified branch of those sources. Hopefully 
>> my changes to add some customized node colours won't be too 
>> monumental and drastic, and I can incorporate any new Gato 
>> releases/bug fixes into my tree.
>
>
> Are you actually talking about extending Gred, or do you actually
> want to change Gato?
>
It's probably more the former not the latter. It's mostly cosmetic, I 
want all edges with -1 weights to be coloured red and all edges with -1 
weight to be coloured red. I would like it during editing, and during 
export, and I wan't the capability to disable it, and return Gato to its 
normal functionality. To me it is somewhat similar to the "directed 
graph" and "euclidean" options. Except unlike Euclidean, my option 
doesn't require making any changes to the Graph object itself (unless I 
want an "Ising" boolean option stored in the Graph object). Turning on 
"directed graph" alters the drawing of the edges, so I should be able to 
copy that to change colours of the edges. Changing the colours of the 
nodes should be easy enough. I should be able to add a new parameter 
'colour' with a default argument to the node creating function. And if 
"Ising" is enabled I can colour nodes appropriately.

I tried using an .alg and .pro files to colour the nodes and edges but I 
couldn't get it to work properly.

David

On Mon, 21 Feb 2005, David Grant wrote:
>> Sorry about that. I found a script at 
>> http://mail.python.org/pipermail/pythonmac-sig/2004-December/012239.html
>> 
>> which works. I will convert to all spaces.

It is checked in now.

> I just did the same using boa-constructor. In its file menuu there is 
> something called "re-indent entire file." It worked like a charm. I'm 
> following the instructions on cvshome.org's cvs manual on maintaining 3rd 
> party sources and a modified branch of those sources. Hopefully my changes to 
> add some customized node colours won't be too monumental and drastic, and I 
> can incorporate any new Gato releases/bug fixes into my tree.

Are you actually talking about extending Gred, or do you actually
want to change Gato?

Best,
A.


--
Dr. Alexander Schliep                                    sc...@mo...
Group Leader Algorithmics, Dept. Comp. Mol. Bio        phone: +49-30-8413-1166
Max Planck Inst. for Molecular Genetics                 fax: + 49-30-8413-1152
Ihnestrasse 73, 14195 Berlin, Germany        http://algorithmics.molgen.mpg.de

Alexander Schliep wrote:

>
> On Sun, 20 Feb 2005, David Grant wrote:
>
>> Alexander Schliep wrote:
>>
>>>
>>> - mixed tabs and spaces
>>>
>>> I never noticed, Emacs does just fine. It is a trivial fix with 
>>> tabnanny
>>
>
> Sorry about that. I found a script at 
> http://mail.python.org/pipermail/pythonmac-sig/2004-December/012239.html
>
> which works. I will convert to all spaces.

I just did the same using boa-constructor. In its file menuu there is 
something called "re-indent entire file." It worked like a charm. I'm 
following the instructions on cvshome.org's cvs manual on maintaining 
3rd party sources and a modified branch of those sources. Hopefully my 
changes to add some customized node colours won't be too monumental and 
drastic, and I can incorporate any new Gato releases/bug fixes into my tree.

David

Alexander Schliep wrote:

>
> On Sun, 20 Feb 2005, David Grant wrote:
>
>> Alexander Schliep wrote:
>>
>>>
>>> - mixed tabs and spaces
>>>
>>> I never noticed, Emacs does just fine. It is a trivial fix with 
>>> tabnanny
>>
>
> Sorry about that. I found a script at 
> http://mail.python.org/pipermail/pythonmac-sig/2004-December/012239.html
>
> which works. I will convert to all spaces.

Ah, that script looks interesting. So it looks at what Python thinks is 
an indent. That would handle the lines where there are 2 spaces, and a 
tab, to make up 4 spaces total (the tab equaling 2 spaces). Those were 
lines my sed script couldn't handle and after it converted the tab to 8 
spaces, I had 10 spaces of indentation which isn't even a multiple of 4.

Let me know if it is in CVS. I'll check later today though anyways. I'd 
like to use the all-spaces code as my main trunk, and then I'll make my 
own branch, then hopefully I can merge in any new bug fixes to Gato into 
my local tree later.

>
> Actually Python can parse the mixed tab and space just fine. There are 
> no ambiguities, and the script seems to work fine on it.

Yeah it is interesting that Python works. It sees the script as Emacs 
and IDLE do, not as vi does. And that script you found seems to take 
that same approach.


-- 
David J. Grant
http://www.davidgrant.ca:81

On Sun, 20 Feb 2005, David Grant wrote:
> Alexander Schliep wrote:
>
>> 
>> - mixed tabs and spaces
>> 
>> I never noticed, Emacs does just fine. It is a trivial fix with tabnanny

Sorry about that. I found a script at 
http://mail.python.org/pipermail/pythonmac-sig/2004-December/012239.html

which works. I will convert to all spaces.

> Emacs and Idle seem to work fine, but vi/vim and boa both have problems with

What is vi? ;)

> it. tabnanny doesn't seem to work at all. Not sure why, it doesn't give any 
> useful output when run on the Gato.py file besides this:
>
> Gato.py 123 '\tFrame.__init__(self,parent)\n'
>
> Doing this:
>
> #sed -e "N;s/\t/        /g" Gato.py > Gato-new.py
>
> seems to work partly. But there are some lines which have mixed tabs and 
> spaces in the same line, and those have to be fixed manually.

Actually Python can parse the mixed tab and space just fine. There 
are no ambiguities, and the script seems to work fine on it.

Best,
Alexander


--
Dr. Alexander Schliep                                    sc...@mo...
Group Leader Algorithmics, Dept. Comp. Mol. Bio        phone: +49-30-8413-1166
Max Planck Inst. for Molecular Genetics                 fax: + 49-30-8413-1152
Ihnestrasse 73, 14195 Berlin, Germany        http://algorithmics.molgen.mpg.de

Alexander Schliep wrote:

>
> - mixed tabs and spaces
>
> I never noticed, Emacs does just fine. It is a trivial fix with tabnanny
>
>
Emacs and Idle seem to work fine, but vi/vim and boa both have problems 
with it. tabnanny doesn't seem to work at all. Not sure why, it doesn't 
give any useful output when run on the Gato.py file besides this:

Gato.py 123 '\tFrame.__init__(self,parent)\n'

Doing this:

#sed -e "N;s/\t/        /g" Gato.py > Gato-new.py

seems to work partly. But there are some lines which have mixed tabs and 
spaces in the same line, and those have to be fixed manually.


-- 
David J. Grant
http://www.davidgrant.ca:81

Dear Mr. Grant,

your first email was sent on 2/15. After two business days (due to the 
time difference) you complain rather impolitely about not getting an 
answer promptly. You demand developer access without, say, providing a 
patch or two to demonstrate your familiarity with Gato.

I am afraid, I do not think we would get along too well.

To address some of your issues:

- Gato list@sourceforge

This is brand new

- mixed tabs and spaces

I never noticed, Emacs does just fine. It is a trivial fix with tabnanny

- documentation

I guess that there could be more. We are finishing up the book,
code cleanup has to wait.

- File format

The old *.cat format was a hack anyways. We will change that to a
an XML-based format within the next three months.

- Developer access

Frankly, we want to keep the editor as simple as possible.
Some of the things you suggested sounded agreeable, some
dont.

Also, we want complete control over licensing so I'd rather
have only people I hired or myself contribute code.

You can think about whether you still want to contribute.

- Global colors etc.

This is broken right now and still in transition. Again,
this will be taken care off when we have the book off our
chest.

Best,
Alexadner Schliep

--
Dr. Alexander Schliep                                    sc...@mo...
Group Leader Algorithmics, Dept. Comp. Mol. Bio        phone: +49-30-8413-1166
Max Planck Inst. for Molecular Genetics                 fax: + 49-30-8413-1152
Ihnestrasse 73, 14195 Berlin, Germany        http://algorithmics.molgen.mpg.de

Hi,

I've been looking at GraphDisplay.py and the following function:

    def ReadConfiguration(self):
        self.gVertexRadius = 14 # XXX

        self.gFontFamily = "Helvetica"
        self.gFontSize = 10
        self.gFontStyle = tkFont.BOLD

        self.gEdgeWidth = 4

        self.gVertexFrameWidth = 0
        self.cVertexDefault = "#FF0000" # 'red'
        self.cVertexBlink = "black"
        self.cEdgeDefault = "#EEEEEE"
        self.cLabelDefault = "white"
        self.cLabelDefaultInverted = "black"
        self.cLabelBlink = "green"
       
        # Used by ramazan's scaling code
        self.zVertexRadius = self.gVertexRadius
        self.zArrowShape = (16, 20, 6)
        self.zFontSize = 10

The self.cVertexDefault doesn't work for some reason. The vertices 
always show up in blue no matter what I do. I want to figure out how the 
colouring works exactly, because I want to make the vertex colour 
proportional to its weight, so that every time I open a graph, the 
weight determines the colour in which it is displayed.

Thanks,
David Grant

Can anyone tell me if anyone has any example algorithms for use with 
Gato? I'd like to see some more examples to help me with coding my own 
algorithms.

Thanks,
David