China contacts

china_contacts.r

library('socialmixr')
library('tidyverse')

## get China survey
s <- get_survey("https://doi.org/10.5281/zenodo.3366396")

## contact matrix
contact_matrix(s, age.limits = seq(0, 15, 5))

## $matrix
## contact.age.group
## age.group      [0,5)     [5,10)    [10,15)      15+
## [0,5)   0.27692308 0.04615385 0.06153846 4.430769
## [5,10)  0.17391304 1.32608696 0.21739130 4.043478
## [10,15) 0.00000000 0.03125000 2.62500000 3.734375
## 15+     0.08659218 0.11592179 0.11731844 6.638268

## == Let's check the counts
contact_matrix(s, age.limits = seq(0, 15, 5), counts = TRUE)
## $matrix
## contact.age.group
## age.group [0,5) [5,10) [10,15)  15+
## [0,5)      18      3       4  288
## [5,10)      8     61      10  186
## [10,15)     0      2     168  239
## 15+        62     83      84 4753
## $participants
## age.group participants proportion
## 1:     [0,5)           65 0.07295174
## 2:    [5,10)           46 0.05162738
## 3:   [10,15)           64 0.07182941
## 4:       15+          716 0.80359147

## == In the 10-14 age group: 168/64=2.625

## == Let's check in the raw data
s$participants %>%
  mutate(age_group =
           limits_to_agegroups(reduce_agegroups(part_age, seq(0, 15, 5)))) %>%
  group_by(age_group) %>%
  summarise(n())

## # A tibble: 4 x 2
## age_group `n()`
## <fct>     <int>
## 1 0-4          66
## 2 5-9          51
## 3 10-14        69
## 4 15+         779

ten_to_fourteen_id <- s$participants %>%
  filter(part_age >= 10 & part_age < 15) %>%
  .$part_id

s$contacts %>%
  filter(part_id %in% ten_to_fourteen_id) %>%
  mutate(age_group =
           limits_to_agegroups(
             reduce_agegroups(cnt_age_exact, seq(0, 15, 5)))) %>%
  group_by(age_group) %>%
  summarise(n())

## # A tibble: 4 x 2
## age_group `n()`
## <fct>     <int>
## 1 5-9           6
## 2 10-14       184
## 3 15+         261
## 4 NA           29

## == Numbers are slightly smaller, probably because participants with contacts
## with missing age are being removed

contact_matrix(s, age.limits = seq(0, 15, 5), counts = TRUE,
               missing.contact.age = "keep")

## == This matches the numbers returned by contact_matrix