char sets and lack of unique labels due to errors in id

[JonteVe]

Two related issues:

Gephi is not very good at special characters (trema, tilde, cedilha etc), so often the same node are imported with different names (exact set of chars).
What is a good charset to use to handle such characters?
UTF-8 apparently is not good, nor any of the UTFs listed in GEPHI.

You can somewhat remedy that by giving the node ID with misspellings the same label, but that does not always work.
Doing my first graphs I see that some labels are repeated, I would have hoped if the label is the same it should be treated as the same node, but does not seem to work for all.

[JonteVe]

forgot to mention that I am trying to import csv files (one for nodes and one for edges)

Some doubles might be "self-loops" since they are linked to each other.

Also, may labels in graphs lack some of its characters.

[eduramiba]

Can you show a sample of your file?

Anyway, ids of nodes must be unique and exactly the same, similar characters are not treated as the same character.
Source and target also need to match the exact node id.