Understanding `get_polymod_contact_data`

[jfunction]

Hi - I saw this at IDM in November and am interested in using it for my research. Did CRAN do the CRAN thing and take it down? Doesn't seem to be there at the moment.

Anyway, this issue is about get_polymod_contact_data. I am trying to understand the ages argument. If I pass 1 I see ages 0,1,2,... in the result and if I pass seq(0,100,5) I also see 0,1,2,... in the result. The help says the function "Provides contact and participant POLYMOD data from selected countries." but its not clear if that is survey data, contact matrix data, or how it is processing the data (though imputation is discussed). From the output it appears to be a contact matrix, but then I'm not sure what's happening with the age groups here?

I appreciate the effort with this package and am curious to hear more about this.

[njtierney]

Hi @jfunction !

Thanks for posting here.

conmat has not been submitted to CRAN yet, it's just about to be submitted to JOSS, and then will be published on CRAN. Let me know if you are having any issues with installation?

That's a good question about the documentation, thanks for that. I can see how the documentation here could be clearer. I will try and clarify that a bit better in a separate pull request, but to answer your question for the moment, and also to help form part of the updated documentation, here is some information that should help answer your question.

The data is from the contact surveys from the POLYMOD study, and uses the data in socialmixr::polymod$participants and socialmixr::polymod$contacts. As defined in socialmixr), these are:

socialmixr::polymod$participants

the study participant, with age, country, year and day of the week (starting with 1 = Monday)

So each row represents on a given day, a participants key information, their age, where they are from

socialmixr::polymod$contacts

reported contacts of the study participants. The variable phys_contact has two levels (1 denotes physical contact while 2 denotes non-physical contact), duration_multi has five levels (1 is less than 5 minutes while 5 is more than 4 hours, increasing in the order found in Figure 1 in Mossong et al.), and frequency_multi has five levels (1 is daily, 2 is weekly, 3 is monthly, 4 is less often, and 5 is first time)

So each row represents the contacts made by a given study participant. So there are multiple rows for each participant to represent the details of the contact that they have made.

What we want at the end of this is data where each row represents the following:

• In a given setting (setting variable), the number of contacts (the contacts variable) between two age groups (age_from - the age of participant, and age_to - the age of the person the participant had contact with), and also the number of participants in that age group.

You can consider this a contact matrix, using the survey data from POLYMOD. Although there are some assumptions made - in that we don't consider the different levels of contact in different ways, but consider them all as "contact". E.g., there is no difference between physical contact, non physical contact, or duration of contact.

Regarding the ages variable, I need to have a bit of a think about the interpretation of this and how this impacts the data that we get out, I will get back to you.

[jfunction]

Thanks for the info and good luck with the JOSS article. I recall overhearing that socialmixr was getting a rewrite - it might be worth checking on that rumor for backwards compatibility concerns.