Re: [OSRM-talk] Some question/issue using the trips service
Never mind! I figured it out after reading the docs yet a few more times. For anyone else looking, the waypoints are returned in the order you send them, if you want them in route order sort the list on waypoint_index field. Sorry for the noise, -Steve On 10/11/2020 3:50 PM, Stephen Woodbridge wrote: Hi, I'm trying to use the trips service and I'm confused by the waypoint_index. If I pass 50 coordinates on input, is this index back to the original list of coordinates? Because, the lon,lat of the waypoint in the result does not appear to map back to the index of the coordinate in the request. I can see these don't match exactly, but close enough to know something is amiss, probably my understanding. For example, Here is a list of covered bridges in Vermont with an arbitrary START location in the first location: [['', 'START', '', 43.16477, -72.8958, '', ''], ['Belvidere', 'Mill', 'W off Rte. 109', '44.74373', '-72.74143', '1895', 'Null'], ['Belvidere', 'Morgan', 'W off Rte. I 09', '44.74357', '-72.72809', '1887', 'Null'], ['Cambridge', 'Grist Mill', 'E off Rte. 108', '44.63668', '-72.82537', '1872', 'Null'], ['Cambridge', 'Poland', 'N off Rte. 15', '44.65134', '-72.81465', '1887', '2001'], ['Charlotte', 'Holmes Creek', 'W off Rte. 7', '44.33312', '-73.28230', '1870', 'Null'], ['Charlotte', 'Sequin', 'E off Quinlan Bridge', '44.28906', '-73.15032', '1850', '1994'], ['Charlotte', 'Quinlan\\u0027s', 'Off Rte. 7 to N Ferrisburgh', '44.27641', '-73.18388', '1849', '1950'], ['Danville', 'Greenbanks Hollow', 'S off Rte. 2', '44.37762', '-72.12200', '1886', '2002'], ['Enosburg', 'Hopkins', 'W next to Rte. 118', '44.92062', '-72.67309', '1875', 'Null'], ['Fairfax', 'Maple Street', 'S off Rte. 104', '44.66369', '-73.01043', '1865', '2002'], ['Fairfield', 'East Fairfield', 'Adjacent to Rte. 36', '44.78616', '-72.86229', '1865', 'Null'], ['Island Pond', 'Island Pond Footbridge', 'At railroad depot off Main St.', '44.81629', '-71.88105', '2003', 'Null'], ['Johnson', 'Power House', 'W next to Rte. lOOC', '44.63611', '-72.67041', '1872', '2002'], ['Johnson', 'Scribner', 'E off Rte. lOOC', '44.63822', '-72.64855', 'Null', 'Null'], ['Lemington', 'Columbia', 'Off Rte. 102 to NH', '44.85324', '-71.55169', '1912', 'Null'], ['Lunenburg', 'Mount Orne', 'South Lancaster Rd. to NH', '44.46024', '-71.65270', '1911', '1999'], ['Lyndon', 'Schoolhouse', 'Rte. 5 E, Lyndon Corner', '44.51617', '-72.01014', '1879', 'Null'], ['Lyndon', 'Chamberlin', 'W off Rte. 5', '44.51659', '-72.01647', '1881', 'Null'], ['Lyndon', 'Millers Run', 'On Rte. 122', '44.54220', '-72.00989', '1878', '1995'], ['Montgomery', 'Comstock', 'S next to Rte. 118', '44.89963', '-72.64484', '1883', 'Null'], ['Montgomery', 'West Hill', 'S off Rte. 118', '44.86762', '-72.64831', '1883', 'Null'], ['Montgomery', 'Fuller', 'N next to Rte. 118', '44.90334', '-72.64011', '1890', '2000'], ['Montgomery', 'Hectorville', 'W off Rte. 118', '44.85378', '-72.61359', '1883', '2003'], ['Montgomery', 'Hutchins', 'W off Rte. 118', '44.85862', '-72.61255', '1883', 'Null'], ['Montgomery', 'Longley', 'W next to Rte. 118', '44.90723', '-72.65594', '1863', 'Null'], ['Morristown', 'Red', 'W off Rte. l00', '44.51869', '-72.67772', '1896', 'Null'], ['Shelburne', 'Shelburne Museum', 'Facing Rte. 7', '44.37700', '-73.22952', '1845', '1950'], ['Stowe', 'Emily\\u0027s', 'E off Rte. 100', '44.44048', '-72.67983', '1844', '1969'], ['Waterville', 'Village', 'W off Rte. 109', '44.69015', '-72.77098', '1877', 'Null'], ['Waterville', 'Montgomery', 'E next to Rte. 109', '44.70573', '-72.76023', '1877', '1998'], ['Waterville', 'Codding Hollow', 'E next to Rte. l09', '44.71216', '-72.75624', '1877', 'Null'], ['Westford', 'Browns River', 'Off Rte. 128', '44.61259', '-73.00808', '1873', '2001'], ['Wolcott', 'Fisher Railroad', 'S off Rte. 15', '44.53248', '-72.42783', '1908', 'Null'], ['Brandon', 'Sanderson', 'S off Rte. 7', '43.78966', '-73.11176', '1840', 'Null'], ['Chelsea', 'Moxley .', 'E next to Rte. 110', '43.95703', '-72.46340', '1883', 'Null'], ['Clarendon', 'Kingsley', 'SW off Rte. 103', '43.52386', '-72.94106', '1870', '2003'], ['Coventry', 'Black River', 'Town highway 36', '44.86098', '-72.27332', '1881', 'Null'], ['East Montpelier', 'Coburn', 'NW off Rte. 2', '44.28084', '-72.45419', '1851', '1970'], ['Hartland', 'Martins Mill', 'E off Rte. 5', '43.53250', '-72.39594', '1881', 'Null'], ['Hartland', 'Willard', 'E off Rte. 5', '43.59387', '-72.34945', '1870', 'Null'], ['Hartland', '(Willard) Twin', 'E off Rte. 5', '43.59368', '-72.35015', '2001', 'Null'], ['Marshfield', 'Orton Farm', 'S off Rte. 2', '44.28756', '-72.40818', '1890', 'Null'], ['Middlebury', 'Pulp Mill', 'Seymour St. Ext.', '44.02466', '-73.17748', '1820', '2002'], ['Middlebury', 'Halpin', 'E off Rte. 7', '44.05013', '-73.14084', '1850', 'Null'], ['Northfield', 'Stony Brook', 'W off Rte. 12A', '44.12040', '-72.68918', '1899', 'Null'], ['Northfield', 'Northfield Falls
[OSRM-talk] Some question/issue using the trips service
Hi, I'm trying to use the trips service and I'm confused by the waypoint_index. If I pass 50 coordinates on input, is this index back to the original list of coordinates? Because, the lon,lat of the waypoint in the result does not appear to map back to the index of the coordinate in the request. I can see these don't match exactly, but close enough to know something is amiss, probably my understanding. For example, Here is a list of covered bridges in Vermont with an arbitrary START location in the first location: [['', 'START', '', 43.16477, -72.8958, '', ''], ['Belvidere', 'Mill', 'W off Rte. 109', '44.74373', '-72.74143', '1895', 'Null'], ['Belvidere', 'Morgan', 'W off Rte. I 09', '44.74357', '-72.72809', '1887', 'Null'], ['Cambridge', 'Grist Mill', 'E off Rte. 108', '44.63668', '-72.82537', '1872', 'Null'], ['Cambridge', 'Poland', 'N off Rte. 15', '44.65134', '-72.81465', '1887', '2001'], ['Charlotte', 'Holmes Creek', 'W off Rte. 7', '44.33312', '-73.28230', '1870', 'Null'], ['Charlotte', 'Sequin', 'E off Quinlan Bridge', '44.28906', '-73.15032', '1850', '1994'], ['Charlotte', 'Quinlan\\u0027s', 'Off Rte. 7 to N Ferrisburgh', '44.27641', '-73.18388', '1849', '1950'], ['Danville', 'Greenbanks Hollow', 'S off Rte. 2', '44.37762', '-72.12200', '1886', '2002'], ['Enosburg', 'Hopkins', 'W next to Rte. 118', '44.92062', '-72.67309', '1875', 'Null'], ['Fairfax', 'Maple Street', 'S off Rte. 104', '44.66369', '-73.01043', '1865', '2002'], ['Fairfield', 'East Fairfield', 'Adjacent to Rte. 36', '44.78616', '-72.86229', '1865', 'Null'], ['Island Pond', 'Island Pond Footbridge', 'At railroad depot off Main St.', '44.81629', '-71.88105', '2003', 'Null'], ['Johnson', 'Power House', 'W next to Rte. lOOC', '44.63611', '-72.67041', '1872', '2002'], ['Johnson', 'Scribner', 'E off Rte. lOOC', '44.63822', '-72.64855', 'Null', 'Null'], ['Lemington', 'Columbia', 'Off Rte. 102 to NH', '44.85324', '-71.55169', '1912', 'Null'], ['Lunenburg', 'Mount Orne', 'South Lancaster Rd. to NH', '44.46024', '-71.65270', '1911', '1999'], ['Lyndon', 'Schoolhouse', 'Rte. 5 E, Lyndon Corner', '44.51617', '-72.01014', '1879', 'Null'], ['Lyndon', 'Chamberlin', 'W off Rte. 5', '44.51659', '-72.01647', '1881', 'Null'], ['Lyndon', 'Millers Run', 'On Rte. 122', '44.54220', '-72.00989', '1878', '1995'], ['Montgomery', 'Comstock', 'S next to Rte. 118', '44.89963', '-72.64484', '1883', 'Null'], ['Montgomery', 'West Hill', 'S off Rte. 118', '44.86762', '-72.64831', '1883', 'Null'], ['Montgomery', 'Fuller', 'N next to Rte. 118', '44.90334', '-72.64011', '1890', '2000'], ['Montgomery', 'Hectorville', 'W off Rte. 118', '44.85378', '-72.61359', '1883', '2003'], ['Montgomery', 'Hutchins', 'W off Rte. 118', '44.85862', '-72.61255', '1883', 'Null'], ['Montgomery', 'Longley', 'W next to Rte. 118', '44.90723', '-72.65594', '1863', 'Null'], ['Morristown', 'Red', 'W off Rte. l00', '44.51869', '-72.67772', '1896', 'Null'], ['Shelburne', 'Shelburne Museum', 'Facing Rte. 7', '44.37700', '-73.22952', '1845', '1950'], ['Stowe', 'Emily\\u0027s', 'E off Rte. 100', '44.44048', '-72.67983', '1844', '1969'], ['Waterville', 'Village', 'W off Rte. 109', '44.69015', '-72.77098', '1877', 'Null'], ['Waterville', 'Montgomery', 'E next to Rte. 109', '44.70573', '-72.76023', '1877', '1998'], ['Waterville', 'Codding Hollow', 'E next to Rte. l09', '44.71216', '-72.75624', '1877', 'Null'], ['Westford', 'Browns River', 'Off Rte. 128', '44.61259', '-73.00808', '1873', '2001'], ['Wolcott', 'Fisher Railroad', 'S off Rte. 15', '44.53248', '-72.42783', '1908', 'Null'], ['Brandon', 'Sanderson', 'S off Rte. 7', '43.78966', '-73.11176', '1840', 'Null'], ['Chelsea', 'Moxley .', 'E next to Rte. 110', '43.95703', '-72.46340', '1883', 'Null'], ['Clarendon', 'Kingsley', 'SW off Rte. 103', '43.52386', '-72.94106', '1870', '2003'], ['Coventry', 'Black River', 'Town highway 36', '44.86098', '-72.27332', '1881', 'Null'], ['East Montpelier', 'Coburn', 'NW off Rte. 2', '44.28084', '-72.45419', '1851', '1970'], ['Hartland', 'Martins Mill', 'E off Rte. 5', '43.53250', '-72.39594', '1881', 'Null'], ['Hartland', 'Willard', 'E off Rte. 5', '43.59387', '-72.34945', '1870', 'Null'], ['Hartland', '(Willard) Twin', 'E off Rte. 5', '43.59368', '-72.35015', '2001', 'Null'], ['Marshfield', 'Orton Farm', 'S off Rte. 2', '44.28756', '-72.40818', '1890', 'Null'], ['Middlebury', 'Pulp Mill', 'Seymour St. Ext.', '44.02466', '-73.17748', '1820', '2002'], ['Middlebury', 'Halpin', 'E off Rte. 7', '44.05013', '-73.14084', '1850', 'Null'], ['Northfield', 'Stony Brook', 'W off Rte. 12A', '44.12040', '-72.68918', '1899', 'Null'], ['Northfield', 'Northfield Falls', 'W off Rte. 12', '44.17249', '-72.65146', '1872', 'Null'], ['Northfield', 'Slaughter House', 'W off Rte. 12', '44.16856', '-72.65457', '1872', 'Null'], ['Northfield', 'Lower Cox Brook', 'W off Rte. 12', '44.17278', '-72.65303', '1872', 'Null'], ['Northfield', 'Upper Cox Brook', 'W off Rte. 12', '44.17380',
Re: [OSRM-talk] OSRM 5.0.0 Released
On 4/23/2016 6:46 PM, Daniel Patterson wrote: Hi Steve, You're probably remembering this ticket: https://github.com/Project-OSRM/osrm-backend/issues/483 Yes, that is the one, thanks! OK, I saw a bunch of comments about restrictions and turn penalties so I thought I would check on it as I would like to see it implemented. We still haven't added support for this kind of restriction. Support for the `via node` restrictions exists because it maps fairly directly to OSRM's internal data model. All we have to do is remove the corresponding edge-based-edge identified by the edge-node-edge values. For edge-edge-edge restrictions, we'll need a more complex graph to represent a restricted intersection. Instead of just removing the restricted edge-based-edge, we'll need to replace the existing simple edges with a set of synthetic edges that represent the permitted routes. This is quite a bit more involved than the current approach, and is the reason that nobody has tackled it yet. I confess some ignorance in how these gets implemented, but in the simple model in my mind when doing Dijkstra, I would think that that when you are at a node in the graph and you are deciding which of the next nodes to evaluate, that if you had a restriction(s) for arriving at following node, then the restriction could be defined based on the path of arrival to that node. So for example we can represent edge based restriction of say: AB,BC,CD as a path sequentially through nodes A, B, C, arriving at D. So if we are at node C and the next node in the graph to be evaluated is D and D's parent is C, and C's parent is B, and B's parent is A, then you can not go to D and you evaluate the other following nodes if any that follow on from C. But if you were at C and your parentage was say C, B, Z, then you could proceed to D, unless you also had a restriction for Z->B->C->D. But I suspect that it is not this simple because of the contracted graph and you probably have to transform the intersection into some analogous representation that removes restricted paths through the intersection or something like that. Regardless, thank you for your response, the issue number and explanation. It would be a nice feature to support if anyone is up for the challenge. Thanks, -Steve daniel On Sat, Apr 23, 2016 at 2:09 PM, Stephen Woodbridge <wood...@swoodbridge.com <mailto:wood...@swoodbridge.com>> wrote: Patrick, et al, Congrats! Nice job on a major effort to pull this release together. I have one question, regarding turn restrictions. Last I knew, OSRM only supported edge-node-edge restrictions and not a more complex edge-edge-edge-... style restriction. There used to be issue for this, but I can't find it now. Has this been implemented and I just missed it? I believe there are cases for this in the OSM data and I can provide some simple use cases if needed. Thanks, -Steve On 4/23/2016 7:12 AM, Patrick Niklaus wrote: Hey, I'm proud to announce the 5.0.0 OSRM release. This is by far one of our biggest releases in terms of lines of code changed, since... a while. It was a real team effort featuring contributions from: @akaiser, Aleksei Potov, Bas Couwenberg, Wilhelm Berg, Dane Springmeyer, Daniel J. Hofmann, Daniel Patterson, David Low, Johan Uhle, Jeff Miccolis, Karen Shea, Kerrick Staley, Lauren Budorick, Lev Dragunov, Mathias Gug, Michael Krasnyk, Moritz Kobitzsch, Mortada Mehyar and yours truly. This release features a completely revamped API (both HTTP API and C++ library), a completely new guidance generation engine, replacing the Ruby based cucumber tests suit with its Javascript sibling cucumber-js and various other improvements and fixes. See the detailed changelog at the end of this post. Checkout the new documentation: https://github.com/Project-OSRM/osrm-backend/wiki/New-Server-api https://github.com/Project-OSRM/osrm-backend/wiki/Library-API https://github.com/Project-OSRM/osrm-backend/wiki/node-osrm-api On the development side, we removed the two branch system of having a `develop` and master `branch`. `master` now is _unstable_ and `develop` was removed. This model is now implemented in both `osrm-backend` and `node-osrm`. If you have a look at the structure of the source tree, you will notice we also re-arranged it a lot to more match the module structure. Combined with consistent namespacing, this makes the code a lot easier to navigate now. Changes with regard 5.0.0 RC2: - API: - if `geometry=geojson` is passed the resulting geometry can be a LineString or Point depending on how many co
Re: [OSRM-talk] OSRM 5.0.0 Released
Patrick, et al, Congrats! Nice job on a major effort to pull this release together. I have one question, regarding turn restrictions. Last I knew, OSRM only supported edge-node-edge restrictions and not a more complex edge-edge-edge-... style restriction. There used to be issue for this, but I can't find it now. Has this been implemented and I just missed it? I believe there are cases for this in the OSM data and I can provide some simple use cases if needed. Thanks, -Steve On 4/23/2016 7:12 AM, Patrick Niklaus wrote: Hey, I'm proud to announce the 5.0.0 OSRM release. This is by far one of our biggest releases in terms of lines of code changed, since... a while. It was a real team effort featuring contributions from: @akaiser, Aleksei Potov, Bas Couwenberg, Wilhelm Berg, Dane Springmeyer, Daniel J. Hofmann, Daniel Patterson, David Low, Johan Uhle, Jeff Miccolis, Karen Shea, Kerrick Staley, Lauren Budorick, Lev Dragunov, Mathias Gug, Michael Krasnyk, Moritz Kobitzsch, Mortada Mehyar and yours truly. This release features a completely revamped API (both HTTP API and C++ library), a completely new guidance generation engine, replacing the Ruby based cucumber tests suit with its Javascript sibling cucumber-js and various other improvements and fixes. See the detailed changelog at the end of this post. Checkout the new documentation: https://github.com/Project-OSRM/osrm-backend/wiki/New-Server-api https://github.com/Project-OSRM/osrm-backend/wiki/Library-API https://github.com/Project-OSRM/osrm-backend/wiki/node-osrm-api On the development side, we removed the two branch system of having a `develop` and master `branch`. `master` now is _unstable_ and `develop` was removed. This model is now implemented in both `osrm-backend` and `node-osrm`. If you have a look at the structure of the source tree, you will notice we also re-arranged it a lot to more match the module structure. Combined with consistent namespacing, this makes the code a lot easier to navigate now. Changes with regard 5.0.0 RC2: - API: - if `geometry=geojson` is passed the resulting geometry can be a LineString or Point depending on how many coordinates are present. - the removal of the summary field was revered. for `steps=flase` the field will always be an empty string. Changes with regard to 4.9.1: - API: - BREAKING: Complete rewrite of the HTTP and library API. See detailed documentation in the wiki. - BREAKING: The default coordinate order is now `longitude, latidue`. Exception: Polyline geometry which follow the original Google specification of `latitdue, longitude`. - BREAKING: Polyline geometries now use precision 5, instead of previously 6 - BREAKING: Removed GPX support - New service `tile` which serves debug vector tiles of the road network - Completely new engine for guidance generation: - Support for highway ramps - Support for different intersection types (end of street, forks, merges) - Instruction post-processing to merge unimportant instructions - Improved handling of roundabouts - Tools: - BREAKING: Renamed osrm-prepare to osrm-contract - BREAKING: Removes profiles from osrm-contract, only needed in osrm-extract. - Abort processing in osrm-extract if there are no snappable edges remaining. - Added .properties file to osrm-extract ouput. - Enables the use of multiple segment-speed-files on the osrm-contract command line - Profile changes: - Remove movable bridge mode - Add `maxspeed=none` tag to car profile. - A `side_road` tag support for the OSRM car profile. - Fixes: - Issue #2150: Prevents routing over delivery ways and nodes - Issue #1972: Provide uninstall target - Issue #2072: Disable alternatives by default and if core factor < 1.0 - Issue #1999: Fix unpacking for self-loop nodes not in core. - Infrastructure: - Cucumber test suit is now based on cucumber-js, removes Ruby as dependency - Updated to mapbox/variant v1.1 - Updated to libosmium v2.6.1 - Remove GeoJSON based debugging output, replaced by debug tiles Obtain the new release of node-osrm via: npm install osrm Or obtain our C++ core engine here: https://github.com/Project-OSRM/osrm-backend/archive/v5.0.0.tar.gz Cheers, Patrick ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Route Optimizing
Emil, See this for a good explanation of the various algorithms: http://www2.isye.gatech.edu/~mgoetsch/cali/VEHICLE/TSP/TSP003__.HTM Click the right arrow at the bottom of the page to see more algorithms. The heuristics for the traveling salesman problem fall into the following classes: Tour construction algorithms (creating an initial sub-tour and / or inserting remaining points if necessary) Tour improvement algorithms (improving the existing tour) Most of the time, a combination of algorithms from both of these classes is used. These are called composite procedures. The farthest insertion is a Tour construction algorithm and it could benefit from adding a simulated annealing of a tabu search optimizer to improve the initial tour construction. -Steve On 7/6/2015 4:16 PM, Emil Tin wrote: Seems it the farthest insertion algorithm although i'm not familiar with it. Sendt fra min iPhone Den 06/07/2015 kl. 21.40 skrev Stephen Woodbridge wood...@swoodbridge.com: I have not looked at the round_trip feature code, but if it is solving a TSP problem it really need to solve it for an asymmetric matrix and not a symmetric matrix. Most of the algorithms you find around the web solve the symmetric case, but because we are dealing with oneway streets and turn restrictions, etc there can be some big differences in cost based on the direction of travel between nodes. There is a way to convert an asymmetric matrix into a symmetric matrix that might be an option. Or if you are only looking for an approximate optimization solution then the symmetric matrix might be good enough. You can look at pgRouting and see what we have done there. -Steve On 7/6/2015 2:03 PM, Jonas Plass wrote: Thank you guys for the Inspiration. I will have a Look at it tomorrow. Am 06.07.2015 um 18:03 schrieb Stephen Woodbridge wood...@swoodbridge.com: A simple trick if you want to do start-via-via-...-via-end optimization like TSP is to use TSP and artificially set the distance between end and start locations to zero and this will keep those two city bound together. Then in the result just remove that link. This will allow the same code that solves TSP round trip to also optimize the vis points between a start and end trip. -Steve On 7/6/2015 10:02 AM, Emil Tin wrote: The OSRM team is working on a new API call to compute a round-trip that visits a list of point: https://github.com/Project-OSRM/osrm-backend/tree/feature/round_trip Med venlig hilsen Emil Tin IT- og Processpecialist Trafik ___ KØBENHAVNS KOMMUNE Teknik- og Miljøforvaltningen Byens Anvendelse Njalsgade 13, 1035 Postboks 380 2300 København S Direkte 2369 5986 Mobil 2369 5986 Email z...@tmf.kk.dk EAN 5798009493149 -Oprindelig meddelelse- Fra: Hans Gregers Petersen [mailto:greg...@septima.dk] Sendt: 6. juli 2015 15:55 Til: Mailing list to discuss Project OSRM Emne: Re: [OSRM-talk] Route Optimizing Hi Jonas, Are there any ambitions to create something like an optimization for route stops? And when not do you know any other way to do this via rest request? If you are thinking of the TSP - https://en.wikipedia.org/wiki/Travelling_salesman_problem - then it is a simple problem of I need to visit X places after each other, please optimize the order to visit those places in. The solution is well-described in the WikiPedia link. You can write a little JavaScript, Python or whatever - and optimize on either distance or time. It is fairly simple, and all you use OSRM for is getting the costs. If you want as few calls to the OSRM-server as possible, then have a look at the table-service ( https://github.com/Project-OSRM/osrm-backend/wiki/Server-api#available-services ) Best regards, Greg Hans Gregers Petersen Partner, Seniorkonsulent greg...@septima.dk -- Septima P/S Frederiksberggade 28, 2. tv. 1459 København K www.septima.dk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org
Re: [OSRM-talk] Route Optimizing
I have not looked at the round_trip feature code, but if it is solving a TSP problem it really need to solve it for an asymmetric matrix and not a symmetric matrix. Most of the algorithms you find around the web solve the symmetric case, but because we are dealing with oneway streets and turn restrictions, etc there can be some big differences in cost based on the direction of travel between nodes. There is a way to convert an asymmetric matrix into a symmetric matrix that might be an option. Or if you are only looking for an approximate optimization solution then the symmetric matrix might be good enough. You can look at pgRouting and see what we have done there. -Steve On 7/6/2015 2:03 PM, Jonas Plass wrote: Thank you guys for the Inspiration. I will have a Look at it tomorrow. Am 06.07.2015 um 18:03 schrieb Stephen Woodbridge wood...@swoodbridge.com: A simple trick if you want to do start-via-via-...-via-end optimization like TSP is to use TSP and artificially set the distance between end and start locations to zero and this will keep those two city bound together. Then in the result just remove that link. This will allow the same code that solves TSP round trip to also optimize the vis points between a start and end trip. -Steve On 7/6/2015 10:02 AM, Emil Tin wrote: The OSRM team is working on a new API call to compute a round-trip that visits a list of point: https://github.com/Project-OSRM/osrm-backend/tree/feature/round_trip Med venlig hilsen Emil Tin IT- og Processpecialist Trafik ___ KØBENHAVNS KOMMUNE Teknik- og Miljøforvaltningen Byens Anvendelse Njalsgade 13, 1035 Postboks 380 2300 København S Direkte 2369 5986 Mobil 2369 5986 Email z...@tmf.kk.dk EAN 5798009493149 -Oprindelig meddelelse- Fra: Hans Gregers Petersen [mailto:greg...@septima.dk] Sendt: 6. juli 2015 15:55 Til: Mailing list to discuss Project OSRM Emne: Re: [OSRM-talk] Route Optimizing Hi Jonas, Are there any ambitions to create something like an optimization for route stops? And when not do you know any other way to do this via rest request? If you are thinking of the TSP - https://en.wikipedia.org/wiki/Travelling_salesman_problem - then it is a simple problem of I need to visit X places after each other, please optimize the order to visit those places in. The solution is well-described in the WikiPedia link. You can write a little JavaScript, Python or whatever - and optimize on either distance or time. It is fairly simple, and all you use OSRM for is getting the costs. If you want as few calls to the OSRM-server as possible, then have a look at the table-service ( https://github.com/Project-OSRM/osrm-backend/wiki/Server-api#available-services ) Best regards, Greg Hans Gregers Petersen Partner, Seniorkonsulent greg...@septima.dk -- Septima P/S Frederiksberggade 28, 2. tv. 1459 København K www.septima.dk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Route Optimizing
A simple trick if you want to do start-via-via-...-via-end optimization like TSP is to use TSP and artificially set the distance between end and start locations to zero and this will keep those two city bound together. Then in the result just remove that link. This will allow the same code that solves TSP round trip to also optimize the vis points between a start and end trip. -Steve On 7/6/2015 10:02 AM, Emil Tin wrote: The OSRM team is working on a new API call to compute a round-trip that visits a list of point: https://github.com/Project-OSRM/osrm-backend/tree/feature/round_trip Med venlig hilsen Emil Tin IT- og Processpecialist Trafik ___ KØBENHAVNS KOMMUNE Teknik- og Miljøforvaltningen Byens Anvendelse Njalsgade 13, 1035 Postboks 380 2300 København S Direkte 2369 5986 Mobil 2369 5986 Email z...@tmf.kk.dk EAN 5798009493149 -Oprindelig meddelelse- Fra: Hans Gregers Petersen [mailto:greg...@septima.dk] Sendt: 6. juli 2015 15:55 Til: Mailing list to discuss Project OSRM Emne: Re: [OSRM-talk] Route Optimizing Hi Jonas, Are there any ambitions to create something like an optimization for route stops? And when not do you know any other way to do this via rest request? If you are thinking of the TSP - https://en.wikipedia.org/wiki/Travelling_salesman_problem - then it is a simple problem of I need to visit X places after each other, please optimize the order to visit those places in. The solution is well-described in the WikiPedia link. You can write a little JavaScript, Python or whatever - and optimize on either distance or time. It is fairly simple, and all you use OSRM for is getting the costs. If you want as few calls to the OSRM-server as possible, then have a look at the table-service ( https://github.com/Project-OSRM/osrm-backend/wiki/Server-api#available-services ) Best regards, Greg Hans Gregers Petersen Partner, Seniorkonsulent greg...@septima.dk -- Septima P/S Frederiksberggade 28, 2. tv. 1459 København K www.septima.dk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Reverting 11c6713 fixed at least 5 open issues
Hi all, Over the weekend commit 11c673 was reverted and this appears to have fixed issues #1429, #1451, #1474, #1475, #1479 I have verified that the change fixes these tickets and comment on each ticket. Thank you to everyone involved with reporting, chasing the issue, and resolving it. -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] I'm getting a crazy route in Montevideo
This looks like it is OSRM weirdness, but it might be a data problem: This is very crazy with lots of weird loops, 12 turns. http://osrm.at/bYh But if I add one more position the looping goes away, Granted this changes the exit strategy. but the next route does not. http://osrm.at/bYi Even this would be better, it is about the same time but shorter distance and 9 turns compared to the first route. http://osrm.at/bYj The looping is what makes this stand out. I thought we had turn penalties (but maybe they are not all that impactful) and that we should get few turns for routes of about the same distance going through similar via points. -Steve --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Can hints be used to identify a common segment?
Hi Fernando, Thank you that is a useful observation but my problem is a little different. I have many positions and I would like to determine those that lie on the same segment so that I can group them together. If I remember correctly the hint is a base64 encoded pointer to the edge plus something like and offset along the edge. I could be useful to take advantage of this information to group and order locations along a path. I'll see if I can figure it out from the source code. -Steve On 4/21/2015 8:28 AM, Fernando Pacheco wrote: If both points are in the same segment, their position (third element in route_instructions) should be consecutive. Located on the same street but in different segments (3 and 5): [9, General José Esteban Brito del Pino 122,3,14, 121m, S, 193.1] [9, General José Esteban Brito del Pino, 57,5,21, 56m, S, 191.1] In the same segment (4 and 5): [9, General José Esteban Brito del Pino, 35,4,5, 34m, S, 191.1] [9, General José Esteban Brito del Pino, 56,5,21, 56m, S, 190.1] May be?. Fernando. El 21/04/15 a las 00:24, Stephen Woodbridge escribió: Hi, I have a need to be able to identify if two locations are on the same segment. We already have the hints for each location and there seems to be some commonality in the hints for location on the same segment. I'm wondering if there is a safe way to use the hints to determine if they are on the same segment. Thanks, -Steve --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Can hints be used to identify a common segment?
Hi, I have a need to be able to identify if two locations are on the same segment. We already have the hints for each location and there seems to be some commonality in the hints for location on the same segment. I'm wondering if there is a safe way to use the hints to determine if they are on the same segment. Thanks, -Steve --- This email has been checked for viruses by Avast antivirus software. http://www.avast.com ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] OSRM-talk Digest, Vol 28, Issue 4
Hi Francisco, I am a little confused by the concept of user preferences when taken in context of OSRM and CH (contraction hierarchy). Because of the extensive preprocessing required when using CH, there is not really an ability to offer user preferences at routing because the hierarchy would change based on user preferences. Everything is sort of pre-baked into the graph during osrm-prepare and there is little that you can do for preferences during routing. Maybe I'm confused and don't understand what you are try to accomplish. Are you proposing creating another routing engine that is not based on contraction hierarchy that would solve the bi- or multi-criteria problem? which would be cool to have an alternative solver as an option. Thanks, -Steve On 4/12/2015 5:06 PM, Francis wrote: -First of all, my apologies if you receive duplicated this message, I sent it 2 days ago, but due to the length of the text it was sent to the moderator of the list.- Hi Mohammed, I should have mentioned this option, the linear combination of criteria is naturally the obvious solution when the multiobjective problem is approached for the first time. To define a ranking procedure for the solutions can reduce the multiobjective problem to a single objective problem, but it has many drawbacks: - An arbitrary vector of weights must be defined. Who should define that weights? each user? the routing machine? How do you measure that travel time is for example w_i more or less important than the economic cost, or the CO2 emissions? In the majority of the cases these weights would be totally arbitrary and will not reflect the user preferences. - The linear combination function would add different measures, in our case we are adding euros + minutes, or euros + meters + minutes, and the resulting value would not have much sense. Moreover, if you are adding meters and euro cents for example, the first criterion (or metric) is going to be several orders of magnitude greater than the second, i.e. if you are adding a solution path with 15 meters that costs you 900 euro cents (150km and 9 euros), when you add both magnitudes the second one will become quite irrelevant, and solution paths that minimize the distance are always preferred. - To avoid the second problem, you can normalize all the criteria to be expressed in a scale from 0 to 100, from the minimum value for that criterion (optimum) to the worst possible one, but this would still be a bad solution to model the user preferences. Thanks anyway for you answer Mohammed, it helped me to further develop the purpose of multiobjective graph search algorithms. In summary, routing on road maps with a single criterion have been practically solved, it is enough to see that OSRM calculates an optimum route in the order of milliseconds, and my guess is that other variants like routing with multiple criteria or dynamic costs (like traffic information) will be soon included in routing services. (if they are not yet). Best regards, Francisco J. Message: 2 Date: Thu, 9 Apr 2015 19:16:18 +0100 From: Mohammed Ayoub NEGGAZ am_neg...@esi.dz mailto:am_neg...@esi.dz To: Mailing list to discuss Project OSRM osrm-talk@openstreetmap.org mailto:osrm-talk@openstreetmap.org Subject: Re: [OSRM-talk] OSRM-talk Digest, Vol 27, Issue 12 Message-ID: CAKxYid4dzUESz_a=vxvgrjgd7u0l9w7rs++2jz7rzqri808...@mail.gmail.com mailto:vxvgrjgd7u0l9w7rs%2b%2b2jz7rzqri808...@mail.gmail.com Content-Type: text/plain; charset=utf-8 Hi Francis, I'm working on the dynamic case of the problem but I think I might help you. What if you create a formula for your critera for example, for the costs that you gave you can define : weight = (Travel Time * TT_Weight) + (Cost * C_Weight) where TT_ and C_Weight are factors and weight will allways give a scalar. If we are interrested in travel time we define TT_Weight = 2 C_Weight = 4 and then A= (60 min, 30€) = 60*2 + 30*4 = 240 B= (65 min, 25€) = 65*2 + 25*4 = 230 C= (85 min, 20€) = 85*2 + 20*4 = 250 D= (90 min, 15€) = 90*2 = 15*4 = 240 The algorithm can respond with B which (he thinks) the best route if we prefere spending less than arriving in time. With this modelization, you need to define a static vector of weights (on for each property) and therefore you can use CH with Bi-Dijkstra provided by OSRM. I wish this can help you. Good luck in your work. With respect, 2015-04-09 17:28 GMT+01:00 Francis francisr...@gmail.com mailto:francisr...@gmail.com: Hi Romain, I would also be pretty interested in a bicriteria (or multicriteria) OSRM version. What bicriteria algorithm would you like to use?
Re: [OSRM-talk] arrive to destination on the adequate side in two-way streets
Fernando, I open this issue to document the use case: https://github.com/Project-OSRM/osrm-backend/issues/1413 You can watch this issue to get informed of updates and to add additional comments to it. -Steve On 3/10/2015 10:18 AM, Stephen Woodbridge wrote: Fernando, Here are some other oddities that are related to this problem: http://osrm.at/blp In this example the route should be from green, yellow, to red. But in this case it appears to not honor the implied direction and reorders the points or it does a uturn at the yellow that is not noted in the directions. What the route should look like is similar to this: http://osrm.at/a3p I have moved it over to the oneway streets to force the path. Or the path could like like this: http://osrm.at/blq where it bypasses the red going the wrong direction, around the block to yellow then back to red. This is a good application to see how OSRM responds to various route requests given different conditions. -Steve On 3/10/2015 5:31 AM, Fernando Pacheco wrote: I understand the idea. I had imagined something similar but with existing spatial information and using OSRM directly. The definition of that stretch of the road goes from node i (OSM-Ni) to node j (OSM-Nj). The street is two-way. A is an intermediate point of my route (client). a is the nearest point of A on the street. The vehicles driven on the right side of the street. A OSM-Ni + --- a --- + OSM-Nj If A is the left side of the street: 1 - find the point nearest the street (a); 2 - add the node OSM-Nj (before node a) to the route. OSM-Ni + --- a --- + OSM-Nj A If A is the right side of the street: 1 - find the point nearest the street (a); 2 - add the node OSM-Ni (before node a) to the route. [The first point of the route should be approached in a similar way but the subsequent node is added after a (in flow direction).] Is this a possible approach to solve the problem with OSRM ?. Thank you very much. Fernando. El 09/03/15 a las 16:34, Stephen Woodbridge escribió: On 3/9/2015 1:54 PM, Fernando Pacheco wrote: Hi all. We are developing an application for routing solid waste collection trucks. We use OSRM to calculate distances between collecting points (intermediate points). Some customers are located in two-way streets (simple, no separators) ... How can I say to OSRM that must arrive to the collecting points at the adequate side of the street? That is, on the side where the client is, without crossing the opposite direction of the street to access the load. Is it possible?. Thanks in advance. Fernando. Hi Fernando, When the street is represented as a single street center line, a via point is just located on the street segment and then the route is computed to that point. I do not believe there is any way to say arrive at this node from a given direction. One idea for a work around to this problem is to take advantage of the fact that OSRM does not make u-turns at via points so it might be possible if you know the location is on the right side of a street (in right side drive countries) and you can access the geometry to generate an additional via point before or after the collecting point and then include both points in order in the route. so for a simple street: B A --- C D with 4 collection points and you want to collect A and B only when traversing the street from right to left and C and D only when traversing the street from left to right. So if we add the lower case points a,b,c,d as: B b A a --- c C d D Then we can compute a route from say Z (not shown) to A we would compute the route Z-a-A and this should force OSRM to arrive at A after point a and this would put it in the correct direction. Say you want to route A to C then presumably you should be able to route a-A-c-C, this makes some presumptions that would need to be validated but the idea is that a-A forces the direction along the street for the start of the route, then it would have to exit the left end of the street and either make a u-turn or drive around a block and re-enter the street from the left to get to c then to C moving in the correct direction. But as you can see this greatly complicates the problem by potentially doubling the number of nodes and requiring a lot of pre-computation to assess the side and direction. -Steve W ___ OSRM-talk mailing list osrm-t...@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list osrm-t...@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list osrm-t...@openstreetmap.org https
Re: [OSRM-talk] osrm route directly from command line not via http possible?
On 3/11/2015 10:12 AM, Michael Leonard wrote: Hi Steve Thanks for the message! Yes I found simpleclient.cpp before... but unfortunately I don't know where to start to turn that into a command I can call from the command line to generate a route I'm afraid. If it's not too much trouble, would you be able to share any example code or tips on how to turn simpleclient.cpp into something I can call from the command line? I'd really really appreciate any help! Micheal, It is a command line the generates a route from start_coordinate to end_coordinate and prints out the result. Mimic this code and substitute your coordinates and do something with the output. You have not defined your requirements in enough detail for anyone to be more specific. Also you will need to do some C++ programming to implement whatever you need. -Steve PS: If you want some consulting help to build a specific tool contact me off-list. Thanks Kind regards Mike PS On 10 March 2015 at 16:26, Stephen Woodbridge wood...@swoodbridge.com mailto:wood...@swoodbridge.com wrote: Michael, Take a look at this. https://github.com/Project-__OSRM/osrm-backend/blob/master/__tools/simpleclient.cpp https://github.com/Project-OSRM/osrm-backend/blob/master/tools/simpleclient.cpp I have used it to wrap OSRM into some VRP code. I also recommend using osrm-datastore to host the data, then you can make multiple parallel request to that and we were getting 5-8ms responses to route queries. -Steve On 3/10/2015 12:14 PM, Michael Leonard wrote: Hi I'm new to OSRM and the mailing list but have managed to get it all working well on a 64bit Ubuntu 14.04 server. I’m writing looking for help to get OSRM’s routing algorithm working directly from the command line rather than via http. More info OSRM is already extremely fast when called using |curl http://localhost …| - however I’m looking to include it an algorithm to generate millions of realistic driving routes, and am hoping that the underlying C++ routing function might be accessible from the command line directly in some way. With a command line function that wraps around OSRM’s routing algorithm, I could call this from within R where I’m writing the rest of my code. My C++ skills are unfortunately at the extreme beginner level, but I have found where I might start if I were much more advanced: either the file simpleclient.cpp in the OSRM github repo or the node_osrm.cpp file in the node-osrm repo which looks like it's similar to simpleclient.cpp. Hopefully someone has already crossed this bridge and could I'm hoping provide some kind of simple command-line wrapper function to these underlying c++ functions so that I can avoid the http approach which I'm guessing slows things down. Thanks in advance... Thanks so much in advance I really appreciate any help or tips to get this working Love osrm by the way Mike PS I posted this on stackoverflow but found this mailing list and thought I might have more luck using it. http://stackoverflow.com/__questions/28958577/open-__source-routing-machine-via-__the-commandl http://stackoverflow.com/questions/28958577/open-source-routing-machine-via-the-commandl _ OSRM-talk mailing list OSRM-talk@openstreetmap.org mailto:OSRM-talk@openstreetmap.org https://lists.openstreetmap.__org/listinfo/osrm-talk https://lists.openstreetmap.org/listinfo/osrm-talk _ OSRM-talk mailing list OSRM-talk@openstreetmap.org mailto:OSRM-talk@openstreetmap.org https://lists.openstreetmap.__org/listinfo/osrm-talk https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Using OSRM linked into other code?
Hi, As a follow up to this, I wanted to report some progress. Using Tools/simpleclient.cpp as a model, it was pretty easy to create a class that mimic that code for our limited needs. We connect via shared memory and we are able to get response times from viaroute requests of 11-20 ms where they used to be about 500 ms. And this is still passing json documents in the response so some of this time is spend encoding and decoding the json. If you need to do something similar, I recommend looking at Tools/simpleclient.cpp and then wrapping those ideas into a class as you require. Thank you everyone that offered ideas and suggests. -Steve On 11/7/2014 2:42 PM, Stephen Woodbridge wrote: I'll answer a bunch of the reply's here: 1. we do pre-compute a distance matrix and use that already but if you have a situation like: o--t--uv- | || B C| | || A--x--yz---D and you want the route A-B-C-D if you use a precomputed distance matrix you get a path A-x-B-x-y-C-y-z-D (depending on where B and C are in those segments (ie: the vehicle makes a u-turn at B and C) when we want a route like A-x-B-t-u-C-y-z-D. OSRM will generate the later route if you ask for the route A,B,C,D with via points. So a simple distance matrix does not work well. 2. The performance issue is not with the C++, we get basically the same performance using Perl (GET) or curl at the command line, or curl from C or from c++. 3. I will look at the node-osrm code. I remember seeing that posted, but had forgotten. Thanks for the reminder of that. 4. I am some what stuck on an older version of the source code because I'm not in a position to upgrade my server OS and packages. :( So this is somewhat problematic for me at the moment. Anyway, lots of great ideas. I appreciate them all and will be digging into them over the weekend. Best regards, -Steve On 11/7/2014 12:46 PM, John Firebaugh wrote: Hi Steve, Recent versions of osrm-backend build a library which you can link against. See https://github.com/Project-OSRM/node-osrm/ for an example. cheers, John On Fri, Nov 7, 2014 at 7:13 AM, Stephen Woodbridge wood...@swoodbridge.com mailto:wood...@swoodbridge.com wrote: Hi, I seem to remember a while back that there was a discussion about the possibility to embed the OSRM routing engine at the code level rather than doing HTTP requests to a server. I now find myself in a position that this would be desirable to do. I have a small coverage area like a city, but I'm getting killed by the overhead of formatting requests as strings, making a socket connection to osrm-routed, parsing the responses, etc. Making local requests my server this is taking 4-500 ms per request. Basically, I'm doing viaroute requests with 2-100 via points. 99% of the time all I need to know is the travel time. Since I'm developing in C++, I thought it might be easy and much faster to instantiate the routing engine and then have a simple interface where I can pass a container of points and get back the travel time for that route and/or the path coordinates. But I could live without the coordinates if I had to. Has anyone done this already? Can you share? I have started digging through the source to see if I can do this, but working my way in from osrm-routed or Tools/simpleclient.cpp the code is very entangled with all the http request/response stuff that I would ideally like to avoid. So far the most promising path looks like using some variant of the simpleclient, but its not obvious if or how to untangle all the json stuff and simply return a struct or class to the caller without that. I spent most of yesterday, digging through this and made a lot of progress just understanding simpleclient and getting ti to compile and work and get it to actual return results using a shared memory connection. A little help in this direction would be appreciated. Thanks, -Steve _ OSRM-talk mailing list OSRM-talk@openstreetmap.org mailto:OSRM-talk@openstreetmap.org https://lists.openstreetmap.__org/listinfo/osrm-talk https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Using OSRM linked into other code?
Hi, I seem to remember a while back that there was a discussion about the possibility to embed the OSRM routing engine at the code level rather than doing HTTP requests to a server. I now find myself in a position that this would be desirable to do. I have a small coverage area like a city, but I'm getting killed by the overhead of formatting requests as strings, making a socket connection to osrm-routed, parsing the responses, etc. Making local requests my server this is taking 4-500 ms per request. Basically, I'm doing viaroute requests with 2-100 via points. 99% of the time all I need to know is the travel time. Since I'm developing in C++, I thought it might be easy and much faster to instantiate the routing engine and then have a simple interface where I can pass a container of points and get back the travel time for that route and/or the path coordinates. But I could live without the coordinates if I had to. Has anyone done this already? Can you share? I have started digging through the source to see if I can do this, but working my way in from osrm-routed or Tools/simpleclient.cpp the code is very entangled with all the http request/response stuff that I would ideally like to avoid. So far the most promising path looks like using some variant of the simpleclient, but its not obvious if or how to untangle all the json stuff and simply return a struct or class to the caller without that. I spent most of yesterday, digging through this and made a lot of progress just understanding simpleclient and getting ti to compile and work and get it to actual return results using a shared memory connection. A little help in this direction would be appreciated. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Using OSRM linked into other code?
Per, Thank your for responding. We are also doing the HTTP requests, but the performance is killing us. So just a quick update on what I have found out so far: time GET 'http://localhost:5000/viaroute?...' takes about 500 ms on my system. time ./simpleclient --sharedmemory takes about 44 ms with all the default options in the code turned on takes about 22 ms with all the options turned off And these numbers are based on still returning json output and parsing that. So if it were possible to turn simpleclient into object with a few simple options we could get something close to the performance improvement above. And if we can untangle the json encoding and parse and just pass back raw data we would probably see some additional improvement over those numbers. This seems like a worthy path to follow, hence my request for some help or pointers untangling json encoding and parsing. Thanks, -Steve On 11/7/2014 10:41 AM, Per Lindberg wrote: I guess that was me. We also would love to have a single sharable object file (.dll and .so) with a documented API. All we need is travel time from A to B. We currently do gazillions of HTTP calls to a separate process, so a more direct call would be wonderful. Keep me posted if you see any progress in this. Cheers, Per Lindberg Facility labs On 2014-11-07 16:13, Stephen Woodbridge wrote: Hi, I seem to remember a while back that there was a discussion about the possibility to embed the OSRM routing engine at the code level rather than doing HTTP requests to a server. I now find myself in a position that this would be desirable to do. I have a small coverage area like a city, but I'm getting killed by the overhead of formatting requests as strings, making a socket connection to osrm-routed, parsing the responses, etc. Making local requests my server this is taking 4-500 ms per request. Basically, I'm doing viaroute requests with 2-100 via points. 99% of the time all I need to know is the travel time. Since I'm developing in C++, I thought it might be easy and much faster to instantiate the routing engine and then have a simple interface where I can pass a container of points and get back the travel time for that route and/or the path coordinates. But I could live without the coordinates if I had to. Has anyone done this already? Can you share? I have started digging through the source to see if I can do this, but working my way in from osrm-routed or Tools/simpleclient.cpp the code is very entangled with all the http request/response stuff that I would ideally like to avoid. So far the most promising path looks like using some variant of the simpleclient, but its not obvious if or how to untangle all the json stuff and simply return a struct or class to the caller without that. I spent most of yesterday, digging through this and made a lot of progress just understanding simpleclient and getting ti to compile and work and get it to actual return results using a shared memory connection. A little help in this direction would be appreciated. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Selectively avoiding London Congestion Charge
So there might be a work around to this problem is you want to place high costs on the congested area. This would cause the routes to avoid this area but still allow you to start or end a route in the congested area. The side effect of doing this it the the cost might not reflect accurate drive times. -Steve W On 5/19/2014 9:43 AM, John Aherne wrote: Emil Thanks for the information. Not what I wanted to hear, but there we are. I shall have to go away and ponder on this. Anyway I got my routing via motorways working - not 100% - so I still need to do some tweaks. Thanks John Aherne ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Cannot find route between points
On 2/12/2014 8:48 AM, John Aherne wrote:
On Wed, Feb 12, 2014 at 1:19 PM, Rodolphe Quiédeville
rodol...@quiedeville.org mailto:rodol...@quiedeville.org wrote:
John Aherne jjahe...@gmail.com mailto:jjahe...@gmail.com writes:
[...]
Rodolphe
Thanks for the quick reply.
I tried your suggestion using the webconsole.
I then mimiced the url used by the javascript adding on the zoom
level and
instructions=true for good measure and rounded the coordinates to
match.
I still get the same result - status 207
I was already trying zoom levels from 1 to 18 without any luck.
If the map is using the same routing engine then I am baffled.
It's really weird, how do you call the router ie with which tool ? Do
you use a webrowser or a tool like curl or wget ? May be you can paste
some logs off your shell console on a pastie to share with us, may be
will find something.
Regards
--
Rodolphe Quiédeville
Expert Tsung - Consulting en performance des SI
Tel : 06 13 79 63 41
http://blog.rodolphe.quiedeville.org
Thanks for responding.
I use a simple python script for this test.
import urllib2
from datetime import datetime
import json
import urllib
URL =
'http://router.project-osrm.org/viaroute?z=14loc=51.502874,-0.119249loc=51.534750,-0.135230instructions=true'
req = urllib2.Request(URL)
try:
response = urllib2.urlopen(req)
except Exception, e:
print 'Error', str(e)
print 'Error msg', e.msg
print 'Error hdrs', e.hdrs
print 'Error fp', e.fp
res = response.read()
print 'the res', res
my_data = json.loads(res)
for item in my_data:
print 'ITEM',item, my_data[item],'\r\n'
I can send a wireshark capture of the actual wire data if anyone thinks
it is useful.
But the python call is about as simple as it can get.
I can reproduce this with:
$ GET
'http://router.project-osrm.org/viaroute?z=14loc=51.502874,-0.119249loc=51.534750,-0.135230instructions=true'
{status:207,status_message: Cannot find route between points}[]
And I get the same thing in a browser.
-Steve
___
OSRM-talk mailing list
OSRM-talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] problem OSRM compilying
On 2/6/2014 4:33 AM, marcos...@virgilio.it wrote: Hello everybody I want to compile the extractor file on my OSRM-Project, I follow the github guide for my ubuntu pc( I do this on my shell :/|sudo apt-get install build-essential git cmake pkg-config libprotoc-dev libprotobuf7 \ protobuf-compiler libprotobuf-dev libosmpbf-dev libpng12-dev \ libbz2-dev libstxxl-dev libstxxl-doc libstxxl1 libxml2-dev \ libzip-dev libboost-all-dev lua5.1 liblua5.1-0-dev libluabind-dev libluajit-5.1-dev|/) but when i try this command : */gcc -Wall -W -Werror extractor.cpp -o Extractor /* on my shell there are some errors: /*In file included from Extractor/ScriptingEnvironment.h:34:0, from extractor.cpp:30: Extractor/../Util/LuaUtil.h:32:21: fatal error: lua.h: File o directory non esistente compilation terminated.* / Why?thank you/ / I have these lua packages installed and have not problem compiling: liblua5.1-0/precise uptodate 5.1.4-12ubuntu1 liblua5.1-0-dev/precise uptodate 5.1.4-12ubuntu1 libluabind-dev/precise uptodate 0.9.1+dfsg-4 libluabind0.9.1/precise uptodate 0.9.1+dfsg-4 libluajit-5.1-2/precise uptodate 2.0.0~beta9+dfsg-1 libluajit-5.1-common/precise uptodate 2.0.0~beta9+dfsg-1 libluajit-5.1-dev/precise uptodate 2.0.0~beta9+dfsg-1 -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Run more than one instance on same host
On 2/5/2014 4:53 AM, Rodolphe Quiédeville wrote: Rodolphe Quiédeville rodol...@quiedeville.org writes: Hi, Is it possible to run more than one instance on the same host, obviously I'll use a different tcp port for listenning, but what about files in /dev/shm ? I had a look at the source code, and a solution could be to concatenate the tcp port number to filename in /dev/shm as is the filenames will for example for sem.query rename to : /dev/shm/sem.query.5000 What do you think about this solution ? May I implement it and propose a path in this way ? Regards, Rodolphe, I think it makes sense to open a ticket for this. Describe the problem. Your proposed fix as above. Ask what branch they would like a pull request for. I'm guessing that it should be a recent version of develop. -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Getting viaroute response where name if zero bytes
Dennis, I'm getting a viaroute response where the street name is a string of zero bytes. here is a hexdump of the response text: http://codepad.org/UJ6Nwoyh look at line 57-59. If it matters I'm using master [master] ~/work/Project-OSRM$ git show commit fbbd6caf97e3366f358df0a17c976d1d7f46c47a Merge: 992458a 58a5754 Author: Dennis Luxen i...@project-osrm.org Date: Fri Dec 6 04:14:59 2013 -0800 Merge pull request #818 from rodo/master Add missing option '-p' in help message Got any ideas? thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Initial release of OSRM-Tools has been pushed to github
I'm happy to announce that OSRM-Tools has been pushed to to github: https://github.com/woodbri/osrm-tools This project provides some basic tools to: 1. extract pgRouting tables from a database into OSRM normalized file format that can be used to load the data into and OSRM Router instance 2. provide a PostgreSQL extension so you can access the OSRM instance from within the database via postgresql functions. Sorry, it currently has very little documentation but I will add more over time. Since this project is a cross-over between pgRouting and Project-OSRM, and I'm on both email lists lets use these for now and if there is a lot of traffic then we can look at getting dedicated list. Thanks, -Steve W ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] via_points in output
On 11/18/2013 12:09 PM, Florian Lohoff wrote: On Mon, Nov 18, 2013 at 09:08:47AM -0500, Stephen Woodbridge wrote: Or would it be more feasible to extend the OSRM output a bit to dump these interesting nodes additionally? Why not take the route geometries and load them into postgis where you can intersect the routes and get the crossing points? You could alternatively probably do this in C, PHP, Perl or Python using a libgeos wrapper which is the geometry engine that is in PostGIS. I am talking about multiple million routes - My testset has 26k routes. And the point is that after discoverying the intersections i need to delete the routes and calculate 3 new routes. A --x B | | | C Calculating A - B Calculating C - B Finding x Calculating A - x Calculating x - B Calculating C - c Flo Flo, pgRouting has a function to node a network of lines. It basically does exactly this. It takes a table of linestring geometries, chops all the strings at the intersection and forms a correctly noded result in a new table, maintaining a back reference to the original table so you can move any attributes over to the new table. https://github.com/pgRouting/pgrouting/blob/master/src/common/sql/pgrouting_node_network.sql Let me know if you want to try this and need help getting started. -Steve ___ OSRM-talk mailing list osrm-t...@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] What is the maximum number of alternate routes?
Dennis, What is the maximum number of potential alternate routes I can safely assume will be returned? Or is there no fixed number? Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] What is the maximum number of alternate routes?
On 11/18/2013 12:05 PM, Florian Lohoff wrote: On Mon, Nov 18, 2013 at 11:55:56AM -0500, Stephen Woodbridge wrote: Dennis, What is the maximum number of potential alternate routes I can safely assume will be returned? Or is there no fixed number? Computer says no :) The code says currently only 1 alternative route - at least thats what i found in the code ... Thank you for looking at the code, I was confused by the fact that: route_geometry: string alternative_geometries:[string,...] So alternative_geometries is plural and it is and array object. And route_geometry is singular and returns an string object. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Need help understanding the json responses
Hi,
I'm parsing the json responses and the is going well, but my question is
more related what is all this stuff? :)
Most of it makes sense, but I'm a little confused by the following:
The instructions indicate road names: N20, N60, N2, N61, N28,
So the is on the arrive at destination maneuver, but not sure what
the rationale for not having a name there.
The route summary otherwise states the start point and end point names
correctly as N20, N28
How is route_name determined? The wiki page states Array of strings
giving the name of roads. There does not appear to be an
alternative_names item so this must be unique to the route regardless
of alternatives?
What are the via_points? How are they determined? These appear to be
the mapped locations for the start and end points like if I called
nearest?loc=lat,lon
Is the position field in the instructions an index into a point in the
route geometry array?
BTW, thanks for putting up with all my questions, this is all starting
to come together. Great job all around!
Thanks
-Steve
GET
'http://imaptools.com:5000/viaroute?instructions=truealt=truez=18loc=43.235198,-76.420898loc=43.709579,-76.286316'
{version: 0.3,
status:0,
status_message: Found route between points,
route_geometry:
{|zmqA`qjwpC?wyP?o{qBl{qB??m{qBm{qB?o{qB?m{qB?o{qB?m{qB?o{qB?m{qB?o{qB?m{qB??ufN,
route_instructions: [
[10,N20,5506,0,2915,5506m,E,90],
[3,N60,6542,2,562,6542m,S,180],
[7,N2,4771,3,146,4771m,E,90],
[7,N61,58884,4,1797,58884m,N,0],
[3,N28,627,13,153,627m,E,90],
[15,,0,14,0,,N,0.0]],
route_summary:{
total_distance:76332,
total_time:4676,
start_point:N20,
end_point:N28
},
alternative_geometries: [],
alternative_instructions:[],
alternative_summaries:[],
route_name:[N60,N61],
alternative_names:[[,]],
via_points:[[43.235294,-76.420897 ],[43.705882,-76.286315 ]],
hint_data: {
checksum:52824373,
locations: [Dw0AAA4AAADwOAAAdgoAAJqulHdeCOs_3reTAt_ocfs,
MRIAABYAAADkBgAAES0AAEeiWVys-sA_GuaaApX2c_s]
},
transactionId: OSRM Routing Engine JSON Descriptor (v0.3)
}
___
OSRM-talk mailing list
OSRM-talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Update on OSRM-pgRouting integration
Hi all, Here is a demo of using OSRM and pgrouting using php to glue things together. http://imaptools.com:8081/demo/osrm.html Sorry for the weird test grid. The roads have some randomly assigned speed_cat and direction of travel which can make for some unexpected routes. When I have a chance I'll build a real work graph. Read the HELP that comes up when you load the page. Here is a description of how it works. The [Route Path] function just collects the marker locations and makes a direct call to the OSRM server. No magic there, you can look at the Javascript to see how to make Leaflet do that. I was going to add dragging the route or markers, but it more interesting to get this working in pgrouting. For the [Optimize *] functions, I call a PHP script to glue everything together for the moment. The basic logic is: 1. get and array of points 2. loop through the points calling OSRM N*(N-1) times to build a distance matrix and cache the result in an NxN array 3. call pgrouting pgr_tsp(dmatrix, start, stop) to get an ordered list of points 4. Get the cached route_geometry for the ordered pairs of nodes 5. return that as jsonp object. Time vs Distance Currently there is no easy way to ask OSRM to compute shortest time vs shortest distance. I understand that I might be able to do this using profile.lua files but this would require having two OSRM servers running using different profiles. For the time being, I compute a routes for the distance matrix calling OSRM and build my distance matrix based on the route_distance vs the route_time values returned in the request. Then TSP orders the nodes to minimize the cost values used. I suspect we would get different results if we could compute the routes based on either time or distance. My next step is to move the php code into C code in the database. And at some point build out a real road network for a better demo. -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Shortest time versus shortest distance
Hi Dennis, I noticed that in the map.osrm for edges we define both edge length and edge weight that is in 1/10s units. Is is possible to get routes based on shortest path versus shortest time? When I was reading somewhere (maybe in the profile.lua) that you can apply a cost for traffic lights which implies to me that this is adding some time to the route. Maybe this is all defined in the profile.lua and I need to look more into these. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] All route requests report status: 207
Dennis,
I can't get osrm-routed to generate a route. I'm guessing I did
something wrong in the process of generating the graph or extracting it
into the intermediate file.
All I get is status:207
GET 'http://localhost:5000/viaroute?loc=42.5,-74.5loc=44.0,-75.5'
{version: 0.3,status:207,status_message: Cannot find route
between points,route_geometry: ,route_instructions:
[],route_summary:{total_distance:0,total_time:214748365,start_point:,end_point:},alternative_geometries:
[],alternative_instructions:[],alternative_summaries:[],route_name:[,],alternative_names:[[,]],via_points:[],hint_data:
{checksum:66848108, locations:
[pssrALsDfwAAXaRDAHCdiv8,
pssrALsDfwAAXaRDAHCdiv8]},transactionId: OSRM
Routing Engine JSON Descriptor (v0.3)}
Since it is late here, I'll check things in the morning. I'll build a
mapserver mapfile to display the graph and a web app to allow us to
interact with the map and the router.
Running osrm-prepare reported the following. Can you look this over and
see if anything looks off in it.
Thanks,
-Steve
$ osrm-prepare st.osrm
[info] Input file: st.osrm
[info] Restrictions file: st.osrm.restrictions
[info] Profile: profile.lua
[info] Threads: 8
[info] Importing n = 3069975 nodes
[info] and 3066984 edges
[info] Graph loaded ok and has 3066984 edges
[info] 0 restrictions, 0 bollard nodes, 0 traffic lights
[info] Generating edge-expanded graph representation
[info] Identifying components of the road network
. 10% . 20% . 30% . 40% . 50% . 60% . 70% . 80% . 90% . 100%
[info] identified: 284018 many components
[info] generating edge-expanded nodes
. 10% . 20% . 30% . 40% . 50% . 60% . 70% . 80% . 90% . 100%
[info] Generated 4632706 nodes in edge-expanded graph
[info] generating edge-expanded edges
. 10% . 20% . 30% . 40% . 50% . 60% . 70% . 80% . 90% . 100%
[info] Generated 4632706 edge based nodes
[info] Node-based graph contains 4632706 edges
[info] Edge-expanded graph ...
[info] contains 4631249 edges
[info] skips 0 turns, defined by 0 restrictions
[info] writing node map ...
[info] building r-tree ...
[info] constructing r-tree of 4632706 elements
[info] finished r-tree construction in 10.2177 seconds
[info] using software based CRC32 computation
[info] CRC32: 66848108
[info] initializing contractor
merged 0 edges out of 9262498
contractor finished initalization
Contractor is using 1 threads
initializing elimination PQ ...ok
preprocessing 4632706 nodes 10% . 20% . 30% . 40% . 50% . 60% .
[flush 3081249 nodes] 70% . 80% . 90% . 100%
100% [info] Contraction took 44.2668 sec
[info] Getting edges of minimized graph
. 10% . 20% . 30% . 40% . 50% . 60% . 70% . 80% . 90% . 100%
[info] Building Node Array
[info] Serializing compacted graph of 9196634 edges
[info] Preprocessing : 134.197 seconds
[info] Expansion : 90165.2 nodes/sec and 136063 edges/sec
[info] Contraction: 104654 nodes/sec and 207754 edges/sec
[info] finished preprocessing
___
OSRM-talk mailing list
OSRM-talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] osrm-extract on xml file generates weird file names
Hi, I just got a small map.osm.xml extract file from openstreetmap and ran it with: $ osrm-extract map.osm.xml $ ls map.osm.xml map.osrm.restrictionsxml map.osrmxml map.osrmxml.names It looks like there is a bug generating the restrictions file name. Shouldn't that be: map.osrmxml.restrictions rather than: map.osrm.restrictionsxml as this will be a problem when you run: $ osrm-prepare map.osrmxml it will not find map.osrmxml.restrictions! Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
[OSRM-talk] Questions on OSRM normalized file format
Hi all, I'm currently the lead developer for pgRouting and I'm interested in being able to create a local instance of OSRM that can be accessed via pgRouting. This all seems pretty straight forward to prototype up. At the moment, I have a few questions on OSRM normalized file format. I want to be able to write a utility program for postgresql to prep and dump a pgRouting topology to an OSRM normalized file. Looking at the edge section: It appears that edges can be entered as undirected (ie: twoway with same costs in both directions) or directed in the case of oneway streets. 1. if I have a edge with different to-from and from-to costs am I correct in assuming that the way to enter this is to split the edge and enter it as two oneway edges with the appropriate costs? 2. oneway edges must be entered where from-to is the direction of the edge, ie: the direction must be to the target node? 3. Can you expound on the rank of the speed profile? Is this still used? in the Speedprofile page it says that this is now handled as LUA scripts. Looking at the turn restrictions section: Is there a graphic that explains this better. It is a little bit hard to follow where the various items come into play. There are references to via-node, from-node, to-node, from-way, and to-way. Here is a common use case that I need to model. It is a majow road digitized as separated north and south bound lanes and intersected by a two way street digitized as a single line. e h | | | | | | a-b-c-d | | | | | | f g Traffic is allowed: ab-bc-cd - east bound cross traffic dc-cb-ba - west bound cross traffic eb-bf - one way south bound traffic gc-ch - one way north bound traffic ab-bc-ch - east bound onto north bound ab-bf - east bound onto south bound dc-ch - west bound onto north bound dc-cb-bf - west bound onto south bound eb-ba - south bound onto west bound gc-cd - north bound onto east bound U-Turns are not allowed: eb-bc-ch - south bound u-turn to north bound gc-cb-bf - north bound u-turn to south bound and more obviously any turn the wrong way down a oneway edge. Any help in understanding this would be greatly appreciated. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Questions on OSRM normalized file format
Yes, but I will read it again. I'm still trying to get up to speed on everything. Dennis, if you can answer my questions below that would be appreciated. By the way, I'm extremely pleased that you have redone this under the BSD license. I look forward to workign with you guys and integrating this into pgRouting. Thanks, -Steve On 11/9/2013 12:17 PM, Emil Tin wrote: Hi Stephen, Back in April I wrote a wiki page detailing the internal processing flow and data structures, did you see it? https://github.com/DennisOSRM/Project-OSRM/wiki/Processing-Flow Maybe there is some useful info there. Dennis can probably tell you what has changed since April. Best regards, Emil On 09 Nov 2013, at 17:38 , Stephen Woodbridge wood...@swoodbridge.com mailto:wood...@swoodbridge.com wrote: Hi all, I'm currently the lead developer for pgRouting and I'm interested in being able to create a local instance of OSRM that can be accessed via pgRouting. This all seems pretty straight forward to prototype up. At the moment, I have a few questions on OSRM normalized file format. I want to be able to write a utility program for postgresql to prep and dump a pgRouting topology to an OSRM normalized file. Looking at the edge section: It appears that edges can be entered as undirected (ie: twoway with same costs in both directions) or directed in the case of oneway streets. 1. if I have a edge with different to-from and from-to costs am I correct in assuming that the way to enter this is to split the edge and enter it as two oneway edges with the appropriate costs? 2. oneway edges must be entered where from-to is the direction of the edge, ie: the direction must be to the target node? 3. Can you expound on the rank of the speed profile? Is this still used? in the Speedprofile page it says that this is now handled as LUA scripts. Looking at the turn restrictions section: Is there a graphic that explains this better. It is a little bit hard to follow where the various items come into play. There are references to via-node, from-node, to-node, from-way, and to-way. Here is a common use case that I need to model. It is a majow road digitized as separated north and south bound lanes and intersected by a two way street digitized as a single line. e h | | | | | | a-b-c-d | | | | | | f g Traffic is allowed: ab-bc-cd - east bound cross traffic dc-cb-ba - west bound cross traffic eb-bf - one way south bound traffic gc-ch - one way north bound traffic ab-bc-ch - east bound onto north bound ab-bf - east bound onto south bound dc-ch - west bound onto north bound dc-cb-bf - west bound onto south bound eb-ba - south bound onto west bound gc-cd - north bound onto east bound U-Turns are not allowed: eb-bc-ch - south bound u-turn to north bound gc-cb-bf - north bound u-turn to south bound and more obviously any turn the wrong way down a oneway edge. Any help in understanding this would be greatly appreciated. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org mailto:OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
Re: [OSRM-talk] Questions on OSRM normalized file format
OK, I have read most all the wiki pages and I'm not getting how the restrictions are being defined. So a little help getting over the mind block I seem to be having for this. Everything else seems really straight forward, so I can't imagine that the restrictions are all the complicated, but it is not clicking with any confidence in my head yet. If I have the following: ab-c | | d ab, bc, and db are all two way and there is no right turn from db to bc Then is the restriction: b, d, c, forbidden, 7F 00 00 where: b is the via-node d is the from-node c is the to-node and the records should then be sorted by EDGE[via-node, to-node] ref. Is this correct? Thanks, -Steve On 11/9/2013 2:52 PM, Stephen Woodbridge wrote: Yes, but I will read it again. I'm still trying to get up to speed on everything. Dennis, if you can answer my questions below that would be appreciated. By the way, I'm extremely pleased that you have redone this under the BSD license. I look forward to workign with you guys and integrating this into pgRouting. Thanks, -Steve On 11/9/2013 12:17 PM, Emil Tin wrote: Hi Stephen, Back in April I wrote a wiki page detailing the internal processing flow and data structures, did you see it? https://github.com/DennisOSRM/Project-OSRM/wiki/Processing-Flow Maybe there is some useful info there. Dennis can probably tell you what has changed since April. Best regards, Emil On 09 Nov 2013, at 17:38 , Stephen Woodbridge wood...@swoodbridge.com mailto:wood...@swoodbridge.com wrote: Hi all, I'm currently the lead developer for pgRouting and I'm interested in being able to create a local instance of OSRM that can be accessed via pgRouting. This all seems pretty straight forward to prototype up. At the moment, I have a few questions on OSRM normalized file format. I want to be able to write a utility program for postgresql to prep and dump a pgRouting topology to an OSRM normalized file. Looking at the edge section: It appears that edges can be entered as undirected (ie: twoway with same costs in both directions) or directed in the case of oneway streets. 1. if I have a edge with different to-from and from-to costs am I correct in assuming that the way to enter this is to split the edge and enter it as two oneway edges with the appropriate costs? 2. oneway edges must be entered where from-to is the direction of the edge, ie: the direction must be to the target node? 3. Can you expound on the rank of the speed profile? Is this still used? in the Speedprofile page it says that this is now handled as LUA scripts. Looking at the turn restrictions section: Is there a graphic that explains this better. It is a little bit hard to follow where the various items come into play. There are references to via-node, from-node, to-node, from-way, and to-way. Here is a common use case that I need to model. It is a majow road digitized as separated north and south bound lanes and intersected by a two way street digitized as a single line. e h | | | | | | a-b-c-d | | | | | | f g Traffic is allowed: ab-bc-cd - east bound cross traffic dc-cb-ba - west bound cross traffic eb-bf - one way south bound traffic gc-ch - one way north bound traffic ab-bc-ch - east bound onto north bound ab-bf - east bound onto south bound dc-ch - west bound onto north bound dc-cb-bf - west bound onto south bound eb-ba - south bound onto west bound gc-cd - north bound onto east bound U-Turns are not allowed: eb-bc-ch - south bound u-turn to north bound gc-cb-bf - north bound u-turn to south bound and more obviously any turn the wrong way down a oneway edge. Any help in understanding this would be greatly appreciated. Thanks, -Steve ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org mailto:OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk ___ OSRM-talk mailing list OSRM-talk@openstreetmap.org https://lists.openstreetmap.org/listinfo/osrm-talk
