merge two GTFS files about gtfs2gps

[rafapereirabr]

I've added a new function merge_gtfs_feeds to our package (commit 55eda53 ) . This is just a starting point, I believe we can improve the function from here. As it stands, the function does not check for duplicated IDs.

[pedro-andrade-inpe]

Nice! I only don´t know whether we should consider the question of GTFS files with and without frequencies. Could GTFS files of these two types be merged?

[rafapereirabr]

Merging two GTFS feeds of different types (detailed and frequency based) is not a problem per se. However, our core function gtfs2gps currently does not work well with mixed-type GTFS feeds. As it stands, the function assumes all data in a feed are of a single type. We know from the data of Rio de Janeiro, however, that a single GTFS feed might have some trips/routes organized in one way or another. I have created a separate issue for this. This should be fun to tackle.

[rafapereirabr]

• Add ERROR if there are any route_id or stop_ids shared between different GTFS feeds
• test output 1 (merging 2 GTFS feeds) and output 2 (getting the gtfs2gps outputs of two files)

[rafapereirabr]

quick approach to check overlapping values

spo <- system.file("extdata/saopaulo.zip", package = "gtfs2gps")
poa <- system.file("extdata/poa.zip", package = "gtfs2gps")
gtfs_list <- list(spo, poa)

# read all feeds
all_feeds <- lapply(gtfs_list, read_gtfs)


##### STOPS ----------------------

# extract stops from all GTFS feeds
stops <- sapply(all_feeds, "[", 'stops')

# extract stop ids
stop_ids <-  sapply(stops, "[[", 'stop_id')

# check inersection between ids
output_intersect <- Reduce(intersect, stop_ids)

# if there is any overlap
if( length(output_intersect) = 0){ 
  stops <- rbindlist(stops, fill = T)
    }



##### AGENCY ----------------------
# extract stops from all GTFS feeds
agency <- sapply(all_feeds, "[", 'agency')

# extract stop ids
agency_ids <-  sapply(agency, "[[", 'agency_id')

[rafapereirabr]

then we need to update the trips, routes, shapes and stop ids:

for(i in 1:length(agency_ids)){
 stop_ids[i] <- sprintf("%s_%s", agency_ids[[i]], stop_ids[[i]])
 # route_ids[i] <- sprintf("%s_%s", agency_ids[[i]], route_ids[[i]])
 # trip_ids[i] <- sprintf("%s_%s", agency_ids[[i]], trip_ids[[i]])
 # ...
}