from/to, contacting/contacted

[sbfnk]

The function arguments and documentation in several places refer to directionality of contacts, with one group contacting another. However, who initiates contact is usually not recorded in contact matrices (or at least I can't think of an instance where it is). Instead rows/contacts need to be identified according to how they're averaged (i.e. whether $C_ij is the average number of contacts that people in group $i$ have with those in group $j$, or the average number of contacts that people in group $j$ have with those in group $i$) - see also https://github.com/socialcontactdata/contactmatrix/pull/6#discussion_r1584780237

[Bisaloo]

It's not entirely clear to me what the specific suggestion is here. How do you propose we document the matrix orientation?

[sbfnk]

I'm not sure I have a good answer to this question. One suggestion that I think would be better than the current naming is to talk neutrally about two populations that come in contact with each other. We could refer to them as population (1) and population (2), for example - they could be the same, or they could be different (e.g. male/female). We then only further need to specify/decide if the total number of contacts of group i is given by the sum of row i or column i.

[chartgerink]

Hi :wave: We discussed this in one of the meetings with the @epiverse-trace group, and I wondered whether we might take some inspiration from network analysis (re: nodes/edges). @joshwlambert correctly raised that they often also use from/to to describe edges.

I saw they for example use tail/head as well, but that may be too directional still?