JHWu92/Cycling-Safe

Cycling-Safe

Base Map

data

1. http://opendata.dc.gov
2. Street segment (ST) - 13522 segments
3. Moving Violation (MV) - Most STREETSEGID in MV DATA is corresponding to ST DATA - Some points in MV DATA have no STREETSEGID - Some points’ STREETSEGID is different from ST DATA - The best buffer value for spatial match is still undecided

Tools

1. Geopandas (python package)

• geopandas.tools.sjoin: it can analyse points in polygon/intersections of polygons.

2. [SNAP][SNAP] (python package): Stanford Network Analysis Project

• snap.GetBetweennessCentr can calculate betweenness for nodes and edges.

Function

def build_load_idx(seg_idx_path, segs_with_ids, update=False):
    """
    params:
        seg_idx_path: the path of segment index if the index doesn't exist, the index will be built.
        segs_with_ids: dict: key=objid, value=shapely.geometry.LineString, segments that will be indexed
        update: True: rebuild the index no matter whether the index exists
    Output:
        rtree index of given segments.
    """

def pt2seg(pt, segs_with_ids, seg_idx):
    """
    params: 
        pt: shapely.geometry.Point
        segs_with_ids:
        seg_idx: rtree index of segments
    Process:
        1. load segment index and find segments(var:idx_match) whose bounds intersect with point's bounds
           if there isn't any matched segment, the point will be buffered until len of idx_match !=0
        2. find the segments that actually intersect with given point
           if there isn't any intersected segment, nearest segment will be returned
    Output:
       hint: whether the returned segid is "nearest" or "intersected"
       segid: assigned segment id
       cnt: len of candidates 
       candidates: the index of segments that are "equally near"/"intersected" to the point
       distance: the distance between point and segment
    """