Calculate multiple point distance matrices with dplyr-CodePudding

I know that there is a similar question about that here

Any ideas of how should I approach this?

CodePudding user response：

If you want to calc distance by bird_ID, then you should likely left_join the locations onto the original data before calculating the distance.

distances <- dt %>% 
  left_join(locations, by = "bird_ID") %>%   # NEW
  group_by(bird_ID) %>%
  select(bird_ID,                            # suppress an auto-add message
         longitude, latitude, date,
         longitude.mean, latitude.mean) %>%  # added so we have the data per-bird_ID
  mutate(
    Dist = geosphere::distHaversine(
      cbind(longitude.mean, latitude.mean),  # removed `location$`
      cbind(longitude, latitude))
  ) %>%
  tidyr::drop_na()
# # A tibble: 2 x 7
# # Groups:   bird_ID [1]
#   bird_ID longitude latitude date       longitude.mean latitude.mean    Dist
#   <chr>       <dbl>    <dbl> <chr>               <dbl>         <dbl>   <dbl>
# 1 350E         40.2     64.7 2022-05-02           52.1          68.5 674850.
# 2 350E         40.2     64.7 2022-05-02           52.1          68.5 674840.