Closed iamiranjbar closed 3 years ago
@iamiranjbar The problem is that your data contains reverse movement, as shown below.
FMM assume the object always move along the direction of an edge so that it cannot turn back. This issue has been reported previously. #98
stmatch --network edges.shp --network_id fid --source u --target v --gps traj.csv \
-k 16 -r 0.01 -e 0.01 --output mr.txt --output_fields opath,cpath,mgeom
You can set the parameter reverse_tolerance to correct this problem (which set the proportion of reverse movement on an edge)
stmatch --network edges.shp --network_id fid --source u --target v --gps traj.csv \
-k 16 -r 0.01 -e 0.01 --output mr_corrected.txt --output_fields opath,cpath,mgeom --reverse_tolerance 0.3
The current code contains this parameter but it is not printed for stmatch. I have updated the code in the master branch to print that parameter for stmatch.
Thanks a lot for your response. You helped us very much and i really appreciate you. @cyang-kth
Hello @cyang-kth Sorry we contact you again. We continued with the last version of code and saw another unusual result: Red is GPS track and blue is STMatch result. Here is parameters: K = 32 RADIUS = 0.003 GPS_ERROR = 0.0005 VMAX = 0.0003 FACTOR = 1.5 We use base map that is a little different from osm. You can find it in below link. base map: https://drive.google.com/file/d/1sobd93y9IFWi5S9ju1aI39dD6o92_z9j/view?usp=sharing Here is the GPS track linestring:
We play with parameters a lot but can't get good result. Can you help us again?