The tutorial shows old igraph/python versions

[szhorvat]

The tutorial shows Python 2.7 and igraph 0.6 in the example outputs. This should be updated to recent versions since we don't even support Python 2.7.

[iosonofabio]

Good catch, we should also take away star imports and include a plot with matplotlib I think

[ntamas]

Version numbers updated in 9bde4c2. I think the star imports are okay; personally, I don't mind star imports when someone is working interactively in a Python prompt, I'm only against star imports in Python files.

I'll push an update to igraph.org soon.

[iosonofabio]

The main issue is solved, but since we're here: many people copy-paste from the tutorial into their files and notebooks I believe, so I wondered about the star import too.

Not a big deal I think, but perhaps to discuss again as we ramp up our examples and docs over the next months

Edit: just realized I'm the only one who worries about the star import haha! Let's just close this issue as soon as the new version numbers hit the server.

[vtraag]

just realized I'm the only one who worries about the star import

No, I also don't like it, it obscures very much where something is coming from and makes it difficult to read. I do understand that it's sometimes easy to use though. I personally use the ig abbreviation for import, similar to np for numpy and pd for pandas.

[iosonofabio]

Oh the cavalry had arrived! Yes I use ig too, we could start promoting it that way in the docs?

[ntamas]

It's up to you guys; personally, I never use notebooks so I don't know what's the convention there. I tend to be overly pedantic in my own code and use explicit imports (i.e. from igraph import Graph) even in interactive prompts so I don't have a strong preference for or against using import igraph as ig. (This seems to have been invented by the community anyway as I have never advocated abbreviating igraph as ig anywhere).

[vtraag]

This seems to have been invented by the community anyway as I have never advocated abbreviating igraph as ig anywhere

I've been probably just imitating what is done for numpy and related packages. Explicit imports is also fine, but for a tutorial that brings overly many imports I think.