osm-tonie / gtfs Goto Github PK

See discussion starting @ https://community.openstreetmap.org/t/ptna-news-for-public-transport-network-analysis/8383/15

both: 'rides' and 'sum_rides' should then be calculated in gtfs-analyze-ptna-sqlite.pl

see osm-ToniE/ptna#139
for the gtfs aggregation and analysis part

trips are identical w.r.t. name of stops.

trips end at same stop_area but at different platforms

1. trip ends at: de:12:98:1

2. trip ends at: de:12:98:14

so 1. trip is wrongly identified as a sub-route of the sedond one.

When searching unique trip_ids, identical stop_list-strings of trips of different route_ids (of same route_short_name?) where mixed, so that it can happen that all trip_ids of a route_id are considered as similar and thus the route_id vanishes.

Roughly: number of days where service is provides multiplied by number of departures per day

ptna_trips has "list_service_ids" which includes a list of service_ids (corresponds to each departure time in list_departures)

• calendar: for each relevant service_id find the number of service days (end_date - start_date) * (week_days with '1') / 7

• calendar_dates: for each relevant service_id add the number of "exception_type == 1" and subtract the number of "exception_type == 2"

• store the number of service days for each service in an extra sql table

• ptna_trips: for this trip_id take the service_ids and their individual number out of "list_service_ids" and multiply those numbers with the individual number of service days per service_id

This is linked to issue #3 in 'gtfs-feed' repository

Create a script which uses the feed specific scripts

• get_release_date.sh

• get_release_url.sh

• cleanup.sh

to do the necessary stuff.

ptna-www repo shows the "Service Times"

• departure times sorted by time
• as is for the durations

--> departure times and durations do not fit together when departure times were unsorted and # of durations times > 1

Fix this in this repo rather than in ptna-www

ptna-www will rely on those lists being sorted and well aligned - well not sort again

PRAGMA synchronous = OFF

Using a Transaction
BEGIN TRANSACTION
INSERT/UPDATE
...
...
END TRANSACTION

Using a Prepared Statement

https://stackoverflow.com/questions/1711631/improve-insert-per-second-performance-of-sqlite