-
Notifications
You must be signed in to change notification settings - Fork 2
Generate IRIS Correspondence Table (Utilities) Tool
This tool compares MHN arterial links to IDOT's IRIS features and attempts find the best match for each. Matches, consisting of the MHN link's ABB value and the corresponding IRIS link's OBJECTID value, are saved in a timestamped DBF file in the MHN's output directory. This table is useful for joining IRIS data (such as the AADT counts used for travel demand model validation) to the MHN.
In order to find a suitable match, the script turns every MHN arterial link and IRIS match candidate into a series of points, spaced every x feet (default=25). For each MHN point, all IRIS points within y feet (default=60) are identified. For each link, the number of times each nearby IRIS link was matched is summed into a frequency table. The script then iterates over the frequency table, checking for the most likely match for each MHN link. A minimum of z matches (default=5) must have been made to a particular IRIS link for a match to even be considered, and the names of the two roads (using a combination of ROAD_NAME and the MARKED_RT fields in IRIS's case) are compared using the brilliant fuzzywuzzy module, which must be installed beforehand.
The final match for each MHN link (assuming any match was made) will be the IRIS link with the highest number of point-level matches that has an acceptable fuzzy string comparison score (or the best such score, if multiple IRIS links matched the same number of points).
This process is actually applied separately to four subsets of the MHN:
- Normal arterials
- Divided arterials/boulevards
- Expressways
- Ramps
Each of these uses slightly different distance parameters for the matching. For ramps and expressways, the road names are completely ignored; instead, a higher minimum number of point-matches (20) is required for a link-match to be considered.