Desired features and requests for Data Laboratory

[eduramiba]

Hello everyone, I am Eduardo Ramos.

I open this topic because I am working this summer on improving Data Laboratory features and I think it would be a good idea to listen to people requests for this section of the program. Doesn't matter if you are developer or a user of Gephi, your opinion is important

So this topic could be used for posting new ideas and discussing necessary features.

You can check the already done work in my branch: https://code.launchpad.net/~eduramiba/g ... laboratory

Also, feel free to edit the wiki page of the Data Laboratory project: http://wiki.gephi.org/index.php/Specifi ... Laboratory

Thank you,
Eduardo.

[admin]

Hi,

A node manipulation feature called "Add to group" (or a group of nodes) which would add existing nodes to an existing group would be useful to cancel the "Remove from group".

[eduramiba]

That's a good idea, thanks

I think it will show a UI to select the group to add the selected node(s).
Conditions to be enabled would be:

• The selected nodes are all in the same level
  There is at least 1 group in that level

[eduramiba]

I have done this and tested. You can try it and tell me

[Snurb]

Hi, @Gephi on Twitter suggested that I post this suggestion here:

Currently Gephi seems to treat node names as case-sensitive - it would be great to have the ability to switch that off.

(The background to this: I'm currently visualising networks of @-replies on Twitter, and people can't be relied on to be consistent in their capitalisation of Twitter user names. I can run a script over my data to convert all names to lowercase before importing to Gephi, but it would be a lot easier if Gephi could simply be set to disregard any differences between cases...)

Thanks,

Axel Bruns

[eduramiba]

Hi, @Gephi on Twitter suggested that I post this suggestion here:

Currently Gephi seems to treat node names as case-sensitive - it would be great to have the ability to switch that off.

(The background to this: I'm currently visualising networks of @-replies on Twitter, and people can't be relied on to be consistent in their capitalisation of Twitter user names. I can run a script over my data to convert all names to lowercase before importing to Gephi, but it would be a lot easier if Gephi could simply be set to disregard any differences between cases...)

Thanks,

Axel Bruns

Hi Axel.

We are studying the best way to solve this problem and it could be achieved with creating a merge nodes feature with different strategies.

I will think about this and see what can be done.

Eduardo.

[Snurb]

Hi Eduardo,

excellent, thanks. FWIW: the preferred solution for me would be to choose the most common style of capitalisation after the merge - e.g., if 'abc' occurs 10 times, 'ABC' 20 times, and 'AbC' 30 times, after merging those variants 'AbC should be chosen to represent all of them.

But then, even just an option to convert and merge all to uppercase or all to lowercase would be great.

Axel

[miturian]

hello

I am completely, utterly new to Gephi, but it seems to me that I have to create my database outside of the program? why not have a feature to add nodes / edges manually, one at a time, on an equal footing with data-import?

What I would like to use it for is to organise my articles, both the ones I've read and the ones I am going to read, by adding them one at a time with their respective buzzwords. For this use, it would make much more sense to just add them as I read them. I don't have any experience with mySQL or similar, and if I have to learn both that and Gephi, there's a good chance I won't do either...

I realise that other people must have requested this before, but then, by that logic it would already be included...

thanks =)

[eduramiba]

hello

I am completely, utterly new to Gephi, but it seems to me that I have to create my database outside of the program? why not have a feature to add nodes / edges manually, one at a time, on an equal footing with data-import?

What I would like to use it for is to organise my articles, both the ones I've read and the ones I am going to read, by adding them one at a time with their respective buzzwords. For this use, it would make much more sense to just add them as I read them. I don't have any experience with mySQL or similar, and if I have to learn both that and Gephi, there's a good chance I won't do either...

I realise that other people must have requested this before, but then, by that logic it would already be included...

thanks =)

Hi, you already can add nodes and edges in Overview and soon you will be able to do this and many more things in Data Laboratory view

Here you can check some of the new features: http://gephi.org/2010/introducing-data-laboratory/

[admin]

Hi,

I just encountered the need with the attached file sample.nodelist.csv.

During the "Import CSV" process, it'd be neat to set which column corresponds to the node id instead of generating one, and which column should be used to fill the Label column.


For sample.edgelist.csv, we shouldn't be forced to set a "source" and "target", because it is implicit if only two columns exist.


Cheers!