model_data_dictionary.md

Feature Name	Description	Example Values	City	Comments (Optional)
segment_id	Id of segment in question	26	All
Surface_Tp	1 = Unimproved graded earth, or soil surface road 2 = Gravel or stone road 3 = Brick road 4 = Block road 5 = Surface-treated road 6 = Bituminous concrete road 7 = Portland cement concrete road 8 = Composite road; flexible over rigid 9 = Composite road; rigid over flexible or rigid over rigid ("white topping") 10=Stone dust	5	Boston	Categorical, a column is made for each unique value of this in the dataset
SPEEDLIMIT	Speed limit for that segment	20	Boston	Categorical, a column is made for each unique value of this in the dataset
volume_coalesced	Total Volume of vehicles passing that segment, inclusing heavyweight, lightweight and bikes. Final values are after processing ATR data in our pipeline to impute missing continuous values using the k-NN algorithm.	2045	Boston
speed_coalesced	Mean speed (in mph) of vehicles passing that segment. Final values are after processing ATR data in our pipeline to impute missing continuous values using the k-NN algorithm.	22	Boston
Lanes	Number of lanes in the segment	3	All
width	Width of the road segment, as given by open street map	15	All	Used as a log value in our model
Struct_Cnd	Structural Condition	'pre_month'	Boston	Categorical, a column is made for each unique value of this in the dataset
cycleway_type	0 = None 1 = Lane 2 = Shared Labe 3 = Track 4 = Shared busway 5 = Buffered Lane	0	All
F_F_Class	Federal Functional class for that segment. Codes includes: 1 = Interstate 2 = Other Freeways and Expressways 3 = Other Principal Arterial 4 = Minor Arterial 5 = Major Collector 6 = Minor Collector 7 = Local	5	Boston	Functional classification is the grouping of highways, roads and streets by the character of service they provide and was developed for transportation planning purposes. Categorical, a column is made for each unique value of this in the dataset
Conflict
oneway	Is the segment one way or not	0	All	Boolean
AADT	Annual average daily traffic value on that segment for that week	10.008883	Boston	Used in our model as a log transformed value
visionzero	Number of vision zero comments from seeclickfix for that segment	5	Boston
Seeclickfix	Number of concerns from seeclickfix for that [week, year]	10	Currently in use for Cambridge only
parking_tickets	Number of parking tickets for that segment	10	Currently in use for Cambridge only
intersection	Is this segment part of an intersection	0/1	All
log_width_per_lane	Log value of width		All
ACCIDENT_TYPE	Type of accident, basic description of what occured. 1 = Collision with vehicle 2 = Struck pedestrian 3 = Struck animal 4 = Collision with a fixed object 5 = Collision with some other object 6 = Vehicle overturned (no collision) 7 = Fall from or in moving vehicle 8 = No collision and no object stuck 9 = Other accident	2	Melbourne, Australia	ACCIDENT_TYPE is a character field and cannot contain NULL values.
HIT_RUN_FLAG	Binary indicates whether or not the crash was a hit-run accident. N = No Y = Yes - = Unknown	N	Melbourne, Australia	HIT_RUN_FLAG is a character field and cannot contain NULL values.
LIGHT_CONDITION	Indicates light condition or level of brightness at the time of the accident. 1 = Day 2 = Dusk/dawn 3 = Dark (street lights on) 4 = Dark (street lights off) 5 = Dark (no street lights) 6 = Dark (street lights unknown) 7 = Dusk 8 = Dawn 9 = Unknown	5	Melbourne, Australia	LIGHT_CONDITION is a character field and cannot contain NULL values.
ROAD_GEOMETRY	indicates the layout of the road where the accident occurred. 1 = Cross intersection 2 = T-intersection 3 = Y-intersection 4 = Multiple intersections 5 = Not at intersection 6 = Dead end 7 = Road closure 8 = Private property 9 = Unknown	4	Melbourne, Australia	ROAD_GEOMETRY is a character field and cannot contain NULL values.
SEVERITY	Indicates estimation of the severity or seriousness of the accident. 1 = Fatal 2 = Serious injury 3 = Other injury 4 = Non-injury	3	Melbourne, Australia	SEVERITY is a character field and cannot contain NULL values.
SPEED_ZONE	Indicates the speed zone at the location of the accident. ranges from 030 to 110 = 30km/hr to 110km/hr 777 = Other speen limit 888 = Camping grounds or off road 999 = Not known	030	Melbourne, Australia	SPEED_ZONE is a character field and cannot contain NULL values.
LOCATION_TYPE	Indicates the type of node location or position that corresponds to the accident I = Intersection N = Non-intersection O = off-road U = Unknown	I	Melbourne, Australia	LOCATION_TYPE is a character field and cannot contain NULL values. It may contain empty fields.
ROAD_USER_TYPE	Indicates what the role of the person was at the time of the accident. 1 = Pedestrian 2 = Driver 3 = passenger 4 = Motorcyclist 5 = Pillion Passenger 6 = Bicyclist 7 = Other driver 8 = Other passenger 9 = Not known	4	Melbourne, Australia	ROAD_USER_TYPE is a character field and cannot contain NULL values.