-
Notifications
You must be signed in to change notification settings - Fork 1
/
O-D estimation.py
257 lines (166 loc) · 6.79 KB
/
O-D estimation.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
# coding: utf-8
# In[203]:
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import folium
from sklearn.cluster import AgglomerativeClustering
from scipy.spatial.distance import cdist
# <h2>Remove estimated intersections in route:</h2>
# In[330]:
def remove_inter(df):
clean_df = []
for name, group in df.groupby(["id"]):#, df.time.dt.day]):
#print(name, group)
group["cluster"] = (group.time.diff() > pd.Timedelta(minutes=10)).cumsum()
clean_df += [group.groupby(group.cluster).first()]
#print(group.groupby(group.cluster).first())#.sort_values(group.time).head(1))
#for name, g in group.groupby(group.cluster):
#print(name, g)
#print()
#print()
clean = pd.concat(clean_df)
return clean.drop("cluster", axis=1)
# <h2>Estimate Destination:</h2>
# In[331]:
def get_OD(clean_df):
OD = []
for name, group in clean_df.groupby(clean_df.id):
group_dest = group.copy()
group.columns = columns_orig
OD += [pd.concat([group, group_dest.shift(-1)], axis=1).drop(["time", "id"], axis=1)]
return pd.concat(OD)
# <h2>Clustering Metric:</h2>
# In[332]:
def bundling_metric(X):
def pairwise(l1,l2):
return np.sum(np.dot(np.sum(np.square(l1 - l2)),
np.log(1 + np.abs((l1[:2] - l2[:2])**2 - (l1[2:] - l2[2:])**2))))
return cdist(X, X, metric=pairwise)
# <h2>Clustering</h2>
# In[333]:
def get_clusters(X, n_clusters, metric):
X = metric(X)
model = AgglomerativeClustering(n_clusters=n_clusters, linkage="average", affinity='precomputed')
model.fit(X)
#X[model.labels_ == l]
return model.labels_
# In[334]:
def cluster(df, n_clusters, freq=False):
df = df.dropna(axis=0, how='any')
clustered = []
if freq:
groups = []
for name, group in df.groupby([df.time_orig.dt.hour]):
if group.shape[0] == 1:
continue
group["cluster"] = get_clusters(group.loc[:,"lat_orig":"lon"].values, n_clusters, bundling_metric)
groups += [group]
grouped = group.groupby(group.cluster)
representative = []
for name, g in grouped:
representative += [[g.lat_orig.mean(), g.long_orig.mean(), g.lat.mean(), g.lon.mean(), g.shape[0]]]
#representative = grouped.first()
#representative["count"] = grouped.size()
clustered += [representative]
return clustered, groups
else:
df["cluster"] = get_clusters(df.loc[:, "lat_orig":"lon"].values, n_clusters, bundling_metric)
grouped = df.groupby(df.cluster)
representative = []
for name, g in grouped:
representative += [[g.lat_orig.mean(), g.long_orig.mean(), g.lat.mean(), g.lon.mean(), g.shape[0], name]]
clustered += [representative]
#return pd.concat(clustered)
return clustered, df
# In[335]:
def plot_cluster(cluster):
ax = cluster.plot.scatter("lat_orig", "long_orig")
cluster.plot.scatter("lat", "lon", ax=ax)
plt.show()
# In[336]:
#plot_cluster(clusters[0])
# In[337]:
def cluster_on_map(clustered, original, df=True):
print(clustered)
colors = sum([["green", "blue", "red", "orange", "brown", "yellow",
"white"] for i in range(1000)], [])
if df:
m = folium.Map(location=[clustered.lat.mean(), clustered.lon.mean()])
else:
m = folium.Map(location=[sum([cl[0] for cl in clustered]) / len(clustered), sum([cl[1] for cl in clustered])/len(clustered)])
i = 0
if df:
for index, row in clustered.iterrows():
folium.features.PolyLine([row.loc["lat_orig":"long_orig"].values.tolist(),
row.loc["lat":"lon"].values.tolist()], color=colors[i], weight=row["count"]).add_to(m)
i += 1
return m
for r in clustered:
folium.features.PolyLine([r[:2],
r[2:4]], color=colors[i], weight=r[4]).add_to(m)
#for name, group in original.groupby(original.cluster):
# if name == r[5]:
# for index, row in group.iterrows():
# folium.Marker([row.lat, row.lon], icon=folium.Icon(color=colors[i])).add_to(m)
# folium.Marker([row.lat_orig, row.long_orig], icon=folium.Icon(color=colors[i])).add_to(m)
#folium.features.PolyLine([row.loc["lat_orig":"long_orig"].values.tolist(),
# row.loc["lat":"lon"].values.tolist()], color=colors[i], weight=row["count"]).add_to(m)
i+=1
return m
# In[338]:
# In[326]:
import pymongo
import pandas as pd
def getDbMetrics():
client = pymongo.MongoClient()
db = client['hack']
collection = db.data
aux = collection.find()
col = ["id_orig", "time_orig", "lat_orig", "long_orig", "lat", "lon"]
listona =[]
for item in aux:
listona.append([item['id_orig'],item['time_orig'],item['lat_orig'],item['long_orig'],item['lat'], item['lon']])
df = pd.DataFrame(listona, columns = col)
return df
a = getDbMetrics()
toy_data = [[1, "1:3:12", -23.613, -46.0443],
[1, "1:3:13", -23.613, -46.6443],
[1, "1:3:50", -23.713, -46.6443],
[1, "1:4:12", -23.814, -46.6443],
[1, "1:4:13", -23.613, -46.0443],
[1, "1:4:50", -23.713, -46.0443],
[2, "1:3:12", -23.313, -46.9443],
[2, "1:3:13", -23.513, -46.9443],
[2, "1:3:50", -23.713, -46.9443],
[2, "1:4:12", -23.314, -46.9443],
[2, "1:4:13", -23.513, -46.9443],
[2, "1:4:50", -23.713, -46.9443]]
columns = ["id", "time", "lat", "lon"]
columns_orig = ["id_orig", "time_orig", "lat_orig", "long_orig"]
merged = pd.read_csv("./merged.csv")
merged.columns = ["nothing"] + columns
merged_new = pd.concat([merged.id, merged.lon, merged.time, merged.lat], axis=1)
merged_new.columns = columns
merged_new.time = pd.to_datetime(merged_new.time, format='%Y-%m-%d %H:%M:%S')
merged_new.lat = merged_new.lat.values + 0.1*np.random.random_sample((merged_new.shape[0],))
merged_new.lon = merged_new.lon.values + 0.1*np.random.random_sample((merged_new.shape[0],))
a = pd.concat([merged_new, a])
# In[327]:
df = pd.DataFrame(toy_data, columns=columns)
# In[328]:
a.time_orig = pd.to_datetime(a.time_orig, format='%Y-%m-%d %H:%M')#format='%Y-%m-%d %H:%M:%S')
df.time = pd.to_datetime(df.time, format='%d:%H:%M')#format='%Y-%m-%d %H:%M:%S')
#a.
# print(df)
# print(get_OD(remove_inter(df)))
# print(a)
# clusters2 = cluster(get_OD(remove_inter(df)), 2)
#print a
clusters, orig = cluster(a, 4)
m = cluster_on_map(clusters[0], orig, df=False)
m.save("map3.html")
a["count"] = [1 for i in range(a.shape[0])]
m = cluster_on_map(a, a)
m.save("map2.html")
# In[ ]: