A general tip for graphs with small points: if you use a clipped voronoi diagram to trigger mouse interactions, then the graph becomes much easier to navigate. Basically, the graph gets split up into regions that make efficient but intuitive use of the space.

See here for an example with d3: http://bl.ocks.org/1405439

Otherwise, way cool.

Not sure how I would apply this technique in this particular case, maybe you can give me another hint.

For hovering over each language, it would make it much easier to select the smaller dots.

Very nice.

(Some prior threads on lines of influence)

http://news.ycombinator.com/item?id=4194460

http://news.ycombinator.com/item?id=3409774

http://news.ycombinator.com/item?id=3920619

I'm curious where you got your data for the graph? For example, I don't understand how R is influenced by Scheme, and S is isn't.

The data is from Freebase http://www.freebase.com/view/computer/programming_language and the code to get the data is in this GitHub repo https://github.com/exploringdata/programming-languages-influ....

From Wikipedia's R page: "R is an implementation of the S programming language combined with lexical scoping semantics inspired by Scheme."

Guess that explains it.

Very cool graph.

Small nit: "C" and "Pascal" overlap in position, it appears to be impossible (at least in Chrome) to select "Pascal".

Edit: never mind, it's randomized every time :) But maybe it's a good idea to add some collision avoidance.

Yes, I may be going to preprocess the graph using Gephi and/or see what I can do with sigma, it was my 1st project with this library.

What does the size of each vertex represent? Also, how do you lay out the graphs?

The size is based on the number of languages influenced across all paradigms, the layout is random.

Hm, I would have thought that PHP would have been larger/visible on the graph.

Same here. I was surprised to see it not even being listed despite OP's original data source having it.

http://www.freebase.com/view/computer/programming_language

PHP is a small node, according to the data source it has influenced only 2 languages.

The fact that PHP is widely used doesn't make it influential on other languages, in the case of PHP the opposite may be true.

To find PHP in this random layout restrict the paradigm to "Interpreted language" and move your mouse over Python, the PHP label will show up.

Is it just me, or is the programming language called Clean not on there?

http://en.wikipedia.org/wiki/Clean_(programming_language)

It's included, choose "Functional Programming" and move your mouse over Haskell, it is one of the influenced languages (blue edge).