def get_dismatrix(tasks):
'''
计算矩阵
:param tasks:字典格式{id：[x,y,start_time,end_time,q]}
:return:实际距离矩阵，画图矩阵
'''
plot_matrix = pd.DataFrame(index=tasks.keys(), columns=tasks.keys())
dis_matrix = pd.DataFrame(index=tasks.keys(), columns=tasks.keys())
for k1, v1 in tasks.items():
for k2, v2 in tasks.items():
if k1 == k2: # 同一节点距离为0
dis_matrix[k1][k2] = 0
plot_matrix[k1][k2] = 0
if dis_matrix[k2][k1] >= 0:
break
else:
new_plot = ''
start = str(v1[0]) + ',' + str(v1[1])
end = str(v2[0]) + ',' + str(v2[1])
dis, plot = request_dis(start, end)
#plot坐标转换
one_line = plot.split(';')
for lonlat in one_line:
ll = lonlat.split(',')
lon = float(ll[0])
lat = float(ll[1])
new_lonlat = ct.gcj02_to_wgs84(lon, lat)
point = str(new_lonlat[0]) + ',' + str(new_lonlat[1]) + ';'
new_plot += point
dis_matrix[k1][k2] = dis
plot_matrix[k1][k2] = new_plot[:-1]
dis_matrix[k2][k1] = dis
plot_matrix[k2][k1] = new_plot[:-1]
return dis_matrix, plot_matrix
def gd_convert_lbs_location(ak, imei, bts, **kwargs):
"""
bts:
1. ( mcc,mnc,lac,cellid,signal)
2.
[
[mcc,mnc,lac,cellid,signal],
..
]
return:
{"infocode":"10000",
"result":{"city":"惠州市","province":"广东省",
"poi":"惠州会展中心","adcode":"441302","street":"文昌一路",
"desc":"广东省 惠州市 惠城区 惠州大道 靠近惠州会展中心",
"country":"中国","type":"4","location":"114.4209817,23.1027075",
"road":"惠州大道","radius":"550","citycode":"0752"},
"info":"OK","status":"1"}
"""
url = "http://apilocate.amap.com/position?accesstype=0&imei={imei}&cdma=0&bts={bts}&output=json&key={key}"
main = ''
nears = []
if not isinstance(bts[0], (list, tuple)):
bts = map(lambda _: str(_), bts)
main = ','.join(bts)
else:
bts[0] = map(lambda _: str(_), bts[0])
main = ','.join(bts[0])
remains = bts[1:]
for p in remains:
nears.append(','.join(p))
nearbts = '|'.join(nears)
url = url.format(imei=imei, bts=main, key=ak)
r = requests.get(url)
result = r.json().get('result', {})
data = {}
lon, lat = result['location'].split(',')
data['lon'] = float(lon)
data['lat'] = float(lat)
data['radius'] = float(result.get('radius', 0))
data['address'] =result.get('country','') + result.get('province','')+ \
result.get('city','')+result.get('street','') + result.get('poi','')
data['desc'] = result.get('desc', '')
data['lon'], data['lat'] = gcj02_to_wgs84(data['lon'], data['lat'])
return data
def extract_hot_point(df, dst):
# df is a groupby object
line_num = 1
counter = 0
for key, values in df:
rows = values.itertuples()
last_row = next(rows)
hot_line = []
for row in rows:
counter += 1
if counter % 500000 == 0:
print("processed {},0000 lines".format(int(counter / 10000)))
if bool(row.status) ^ bool(last_row.status):
# 0 -> 1: leave
# 1 -> 0: arrive
hot_line.append(row[0])
line_num += 1
last_row = row
values = values.loc[hot_line]
lon_array = values['longitude'].as_matrix()
lat_array = values['latitude'].as_matrix()
[lon_array, lat_array] = gcj02_to_wgs84(lon_array, lat_array)
values[['longitude']] = lon_array
values[['latitude']] = lat_array
region_id = get_region_label(lon_array, lat_array)
values['region'] = region_id
mask = (values.region != -1) & (values.region != 0) & (values.region !=
1)
values = values[mask]
values.to_csv(dst, index=False, header=False, mode='a+')
#print(values.head(5))
stat = False
if stat:
print("total " + str(line_num) + " lines.")
def plot_poi():
poi_src = Path(r'/home/dlbox/Documents/func_region/Data/Point/CSV')
srcs = poi_src.glob('*')
df = []
for pth in srcs:
#print(pth)
#print(len(df.index))
df.append(pd.read_csv(pth))
c = pd.concat(df)
c.drop(c.columns[[0, 1]], axis=1, inplace=True)
# calibrate GPS
lon = c['lng'].as_matrix()
lat = c['lat'].as_matrix()
[lon_, lat_] = gcj02_to_wgs84(lon, lat)
c[['lng']] = lon_
c[['lat']] = lat_
region_labels = get_region_label(lon_, lat_)
c['region'] = region_labels
c = c[(c['region'] != 0) & (c['region'] != 1)]
c.to_csv(poi_src.parent / 'total.csv', index=False)