Special characters cause bug in visualization

[user1246638]

Using some special characters results in all nodes being on top of eachother in the visualization. The characters include, but are probably not limited to:
"',[]{}!?:;.

At least warning would be useful, if this bug proves to be hard.

[IngoScholtes]

Could you provide a minimal example for the bug?

The following works for me:

import pathpy as pp

#%%
n = pp.Network()
n.add_edge("'", '[')
n.add_edge('"', ']')
n.add_edge(':', ';')
n.add_edge('!', '?')
n.add_edge('{', '}')
n

[maximelucas]

I've experienced the same problem using temporal networks. Here is an MWE of the bug, which displays smoothly if 'B/' is replaced by 'B':

import pathpy as pp

TN = pp.TemporalNetwork()

tedges = [('A', 'B/', 1),
            ('B/', 'C', 2),
            ('C', 'D', 3),
            ('D', 'A', 4),
           ]
for tedge in tedges :
    TN.add_edge(*tedge)
    
pp.visualisation.plot(TN)

Tested with the following characters, and none of them work for me:

symbols = ['`','~','!','@','#','$','%','^','&','*', '(',')',
           '+','=','{','[','}',']','|',':',';','"','<',',','>','.','?','/']

[IngoScholtes]

Sorry for the late follow-up.

Indeed, this is an issue that is due to limitations on which characters can be used as node IDs in HTML and CSS. So, while this does not affect the inner workings of pathpy, it will prevent HTML-based visualisations from being rendered propertly.

For now, the only quick workaround is to write a function that takes care of special characters, making sure that node names conform to what can be used as valid ID attribute in HTML and CSS.

We are working on a new release of pathpy, in which we will fix this issue among others. Thank you very much for the report.