Right now, errors are subsumed under the ~OTHER~ category.
However, they might be an entity the user is interested in.
[x] Turn them into their own category.
[ ] Encode their density into the paths somehow -- should it be apparent in a node how many paths it leads to that have errors? Like, a ratio: error-frequency / total-frequency?
Right now, errors are subsumed under the ~OTHER~ category.
However, they might be an entity the user is interested in.