detailed_itineraires for multiple origins

[franciscopasqual]

The detailed_itineraires function works only for one single origin at a time, and I need to obtain results (actually just the distances) for a large number of origins. I wasn't successful on automating the function using "for", because at every new loop the output overwrites the last one... Any suggestions on how can I change the "det" for every new loop?

for (i in 1:10) {

det <- detailed_itineraries(r5r_core = r5r_core,
origins = origins[i,],
destinations = destinations,
mode = mode,
departure_datetime = departure_datetime,
max_walk_dist = max_walk_dist,
max_trip_duration = max_trip_duration,
shortest_path = FALSE)
}

[xtimbeau]

hi you have 2 solutions at least 1. use foreach or map or apply or assign to a list to list your output 2. pass to detailled_itirenaries a cross product of origins and destinations in a long form (one pair of o and d per row) 2 may be more efficient 1 can be run using parallel stuff Le ven. 4 févr. 2022 à 03:06, franciscopasqual ***@***.***> a écrit :

…

The detailed_itineraires function works only for one single origin at a time, and I need to obtain results (actually just the distances) for a large number of origins. I wasn't successful on automating the function using "for", because at every new loop the output overwrites the last one... Any suggestions on how can I change the "det" for every new loop? for (i in 1:10) { det <- detailed_itineraries(r5r_core = r5r_core, origins = origins[i,], destinations = destinations, mode = mode, departure_datetime = departure_datetime, max_walk_dist = max_walk_dist, max_trip_duration = max_trip_duration, shortest_path = FALSE) } — Reply to this email directly, view it on GitHub <#224>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ANA2KEMBZ5LS6J2BXK4MICLUZMYD5ANCNFSM5NQSA3KA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[mvpsaraiva]

Thanks @xtimbeau for helping out. I definitely suggest using solution 2... do a cross product of origins and destinations, and let r5r run it all in parallel. You can use the function @rafapereirabr suggested in issue #61 to do that.

[franciscopasqual]

Thanks for the suggestions! I managed to create the cross product of all pairs but still don't know how to use that to run detailed_itineraires function, since it asks for different origins and destinations files.

det <- detailed_itineraries(r5r_core = r5r_core,
origins = df # not sure what to put in here
destinations = df # same
mode = mode,
departure_datetime = departure_datetime,
max_walk_dist = max_walk_dist,
max_trip_duration = max_trip_duration,
shortest_path = FALSE)

[rafapereirabr]

please follow the example below:

library(r5r)
library(data.table)
library(dplyr)

# build transport network
data_path <- system.file("extdata/poa", package = "r5r")
r5r_core <- setup_r5(data_path = data_path, temp_dir = TRUE)


# Read points
origins <- read.csv(file.path(data_path, "poa_points_of_interest.csv"))[1:4,]
destinations <- read.csv(file.path(data_path, "poa_points_of_interest.csv"))[3:10,]

# get all possible combinations between origins and destinations
# function
get_all_od_combinations <- function(origins, destinations){
  
  # all possible id combinations
  base <- expand.grid(origins$id, destinations$id)
  
  # rename df
  setDT(base)
  setnames(base, 'Var1', 'idorig')
  setnames(base, 'Var2', 'iddest')
  
  # bring spatial coordinates from origin and destination
  base[origins, on=c('idorig'='id'), c('lon_orig', 'lat_orig') := list(i.lon, i.lat)]
  base[destinations, on=c('iddest'='id'), c('lon_dest', 'lat_dest') := list(i.lon, i.lat)]
  
  return(base)
}

df <- get_all_od_combinations(origins, destinations)

# select/rename columns for r5r input
all_orig <- dplyr::select(df, c('id'=idorig, 'lon'=lon_orig,'lat'=lat_orig))
all_dest <- dplyr::select(df, c('id'=iddest, 'lon'=lon_dest,'lat'=lat_dest))


# run r5r
departure_datetime <- as.POSIXct("13-05-2019 14:00:00", format = "%d-%m-%Y %H:%M:%S")
dit <- detailed_itineraries(r5r_core,
                            origins = all_orig,
                            destinations = all_dest,
                            mode = c("WALK", "TRANSIT"),
                            departure_datetime = departure_datetime,
                            max_walk_dist = 1000,
                            max_trip_duration = 30)

[franciscopasqual]

This worked out perfectly! I didn't realize that the function works for multiple origins at a time if you organize them that way.

Thanks a lot @rafapereirabr!

[Torskelgen]

Is there any way to make your approach work with different departure times for each OD pair, @rafapereirabr, or do I have to do a loop?

[mvpsaraiva]

Is there any way to make your approach work with different departure times for each OD pair, @rafapereirabr, or do I have to do a loop?

Hi @Torskelgen. Currently, it's not possible. This is a good idea, and it's relatively simple to implement. We'll add this feature in the next release of the package.

[Torskelgen]

That would be really neat. Thank you for your really great work with this package!

[rafapereirabr]

The dev version of the detailed_itineraries() function now has a new parameter all_to_all, which allows users to set whether they want to query routes between all origins to all destinations (all_to_all = TRUE) or to query. By default (FALSE), the functions queries routes between the the 1st row of origins to the 1st row of destinations, then the 2nd row of origins to the 2nd row of destinations, and so on.