You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Make start an executable by running chmod +x start
Before running the executable ensure that you're in an environment with osmnx and networkx (apart from other dependencies in requirements.txt) installed.
./start To start up the application.
Head over to http://127.0.0.1:5000/mapbox_gl_new on preferably Google Chrome to use the web interface.
To run the tests run python test/test.py from the home directory
Behind the scenes
Our algorithm runs different weighting mechanism of Dijkstra's and chooses the one that it considers best
Our current system assumes that the user will only be querying in areas around San Fransisco. If a user queries outside this area, our algorithm will not work (the user will see a popup in this case). Even within San Fransisco, we have restricted ourselves to a radius of 20 km.
If a user wishes to try with a different start location they should delete the file graph.p and then try for a route within their required location. The system will download the new graph object and then work as expected. Note, downloading the graph may take time. The user would also have to generate their api key to get the elevation data and add that in elenav/model/config.py
UI Details
Most of the configurations are self explanatory. Before querying a new route, user must reset the graph. Even if they wish to update the location of a marker, the systems needs to be reset.
Below the reset button, we should several statistics. The row corresponding to elena shows the detailed statistics for the path returned by our system. Gain is total “gain-only” distance, Drop is total “drop-only” distance and Dist is the total distance for that particular route.
We show similar statistics for the shortest path to compare how good/bad our system performed.
We also show display the elevation changes in the graph in the form of an image. It helps get a clearer sense of how that route actually looks like.