I am new to using graph-tools and am currently trying to an adapt my running code in networkx to graph-tools in the hopes of improving runtime.

Currently I am stuck at implementing networkx’s center function in graph-tools.

The center would be the set of nodes with eccentricity equal to radius. Unfortunately I have not found a function to calculate either the eccentricity of all vertices nor the graph center.

Any help would be appreciated.

Cheers

Sophia

I have looked at the problem some more and have written some code which for the moment gives me the same results I would get with the networkx function.

```
def get_center_nodes(g):
nodes = g.get_vertices()
if len(nodes) > 2:
distances = topology.shortest_distance(g)
eccentricities = [x.a.max() for x in distances]
centers = nodes[eccentricities.index(min(eccentricities))]
return(centers)
else:
return(nodes[0])
```

In the long run I want to be able to run this on very large graphs so I am not sure if my current solution is the most efficient.

So a follow-up question:

- how can I improve the efficiency of my code?

For context, this is the code I am wanting to call the centers function in:

```
c = gt.topology.label_components(gtG)[0]
components = np.unique(c.a)
centers = []
for i in components:
u = gt.GraphView(gtG, vfilt=c.a == i)
#graph_draw(u, vertex_text=u.vertex_index, ink_scale = 0.5)
center = get_center_nodes(u)
centers.append(center)
```