def local_k(network, events, refs, scale_set, cache=None):
"""
Computes local K function
network: an undirected network data to which reference points are injected
refs: a set of reference points on the given network
points unprojected into the network
events: a set of event points on the given network
points projected into the network
scale_set: a tuple defining spatial scales to be examined
(min, max, interval)
"""
node2localK = {}
net_distances = {}
if cache: net_distances = cache
for node in refs:
node = node[1][0]
a_dest = network[node].keys()[0]
node_proj = (node, a_dest, 0, network[node][a_dest])
if node not in net_distances:
net_distances[node] = pynet.dijkstras(network, node, scale_set[1])
if a_dest not in net_distances:
net_distances[a_dest] = pynet.dijkstras(network, node,
scale_set[1])
distances = pynet.proj_distances_undirected(
network, node_proj, events, scale_set[1],
cache=net_distances).values()
node2localK[node] = pykfuncs.kt_values(scale_set, distances, 1)
return node2localK, net_distances
def local_k(network, events, refs, scale_set, cache=None):
"""
Computes local K function
network: an undirected network data to which reference points are injected
refs: a set of reference points on the given network
points unprojected into the network
events: a set of event points on the given network
points projected into the network
scale_set: a tuple defining spatial scales to be examined
(min, max, interval)
"""
node2localK = {}
net_distances = {}
if cache: net_distances = cache
for node in refs:
node = node[1][0]
a_dest = network[node].keys()[0]
node_proj = (node, a_dest, 0, network[node][a_dest])
if node not in net_distances:
net_distances[node] = pynet.dijkstras(network, node, scale_set[1])
if a_dest not in net_distances:
net_distances[a_dest] = pynet.dijkstras(network, node, scale_set[1])
distances = pynet.proj_distances_undirected(network, node_proj, events, scale_set[1], cache=net_distances).values()
node2localK[node] = pykfuncs.kt_values(scale_set, distances, 1)
return node2localK, net_distances
def dist_weights(network, id2linkpoints, link2id, bandwidth):
"""
Obtains a distance-based spatial weights matrix using network distance
Parameters:
network: an undirected network without additional attributes
id2linkpoints: a dictionary that includes a list of network-projected, midpoints of edges in the network
link2id: a dictionary that associates each edge to a unique id
bandwidth: a threshold distance for creating a spatial weights matrix
Returns:
w : a distance-based, binary spatial weights matrix
id2link: a dictionary that associates a unique id to each edge of the network
"""
linkpoints = id2linkpoints.values()
neighbors, id2link = {}, {}
net_distances = {}
for linkpoint in id2linkpoints:
if linkpoints[linkpoint] not in net_distances:
net_distances[linkpoints[linkpoint][0]] = pynet.dijkstras(
network, linkpoints[linkpoint][0], r=bandwidth)
net_distances[linkpoints[linkpoint][1]] = pynet.dijkstras(
network, linkpoints[linkpoint][1], r=bandwidth)
ngh = pynet.proj_distances_undirected(network,
linkpoints[linkpoint],
linkpoints,
r=bandwidth,
cache=net_distances)
#ngh = pynet.proj_distances_undirected(network, linkpoints[linkpoint], linkpoints, r=bandwidth)
if linkpoints[linkpoint] in ngh:
del ngh[linkpoints[linkpoint]]
if linkpoint not in neighbors:
neighbors[linkpoint] = []
for k in ngh.keys():
neighbor = link2id[k[:2]]
if neighbor not in neighbors[linkpoint]:
neighbors[linkpoint].append(neighbor)
if neighbor not in neighbors:
neighbors[neighbor] = []
if linkpoint not in neighbors[neighbor]:
neighbors[neighbor].append(linkpoint)
id2link[linkpoint] = id2linkpoints[linkpoint][:2]
weights = copy.copy(neighbors)
for ngh in weights:
weights[ngh] = [1.0] * len(weights[ngh])
return pysal.weights.W(neighbors, weights), id2link
def dist_weights(network, id2linkpoints, link2id, bandwidth):
"""
Obtains a distance-based spatial weights matrix using network distance
Parameters:
network: an undirected network without additional attributes
id2linkpoints: a dictionary that includes a list of network-projected, midpoints of edges in the network
link2id: a dictionary that associates each edge to a unique id
bandwidth: a threshold distance for creating a spatial weights matrix
Returns:
w : a distance-based, binary spatial weights matrix
id2link: a dictionary that associates a unique id to each edge of the network
"""
linkpoints = id2linkpoints.values()
neighbors, id2link = {}, {}
net_distances = {}
for linkpoint in id2linkpoints:
if linkpoints[linkpoint] not in net_distances:
net_distances[linkpoints[linkpoint][0]] = pynet.dijkstras(network, linkpoints[linkpoint][0], r=bandwidth)
net_distances[linkpoints[linkpoint][1]] = pynet.dijkstras(network, linkpoints[linkpoint][1], r=bandwidth)
ngh = pynet.proj_distances_undirected(network, linkpoints[linkpoint], linkpoints, r=bandwidth, cache=net_distances)
#ngh = pynet.proj_distances_undirected(network, linkpoints[linkpoint], linkpoints, r=bandwidth)
if linkpoints[linkpoint] in ngh:
del ngh[linkpoints[linkpoint]]
if linkpoint not in neighbors:
neighbors[linkpoint] = []
for k in ngh.keys():
neighbor = link2id[k[:2]]
if neighbor not in neighbors[linkpoint]:
neighbors[linkpoint].append(neighbor)
if neighbor not in neighbors:
neighbors[neighbor] = []
if linkpoint not in neighbors[neighbor]:
neighbors[neighbor].append(linkpoint)
id2link[linkpoint] = id2linkpoints[linkpoint][:2]
weights = copy.copy(neighbors)
for ngh in weights:
weights[ngh] = [1.0]*len(weights[ngh])
return pysal.weights.W(neighbors, weights), id2link