python使用数据帧进行地理编码的多进程处理

Question

我试图返回zipcode作为数据帧中的列。此代码可以工作，但不会在dataframe gps中创建新列。在

import geocoder
import multiprocessing as mp
import pandas as pd

google_key = 'key'

def reverse_gecode(coordinates):
    return geocoder.google(coordinates, key = google_key, method = 'reverse').postal

if __name__ == '__main__':            
    gps = pd.DataFrame({'lat': [27.950575, 40.6936488],
                        'lon': [-82.4571776, -89.5889864]})  # dataframe mehtod
    gps['gps'] = zip(gps.lat, gps.lon)
    x = list(gps['gps'])
    # multiprocessings                    
    pool = mp.Pool(processes = (mp.cpu_count() - 1))
    result_latlong = pool.map(reverse_gecode, x)
    pool.close()
    pool.join()

我试过了

1. gps['zip_code'] = gps.apply(lambda x: pool.map(reverse_gecode, list(x[2])), axis = 1)
2. gps['zip_code'] = gps.apply(lambda x: pool.map(reverse_gecode, x[2]), axis = 1)
3. gps['zip_code'] = gps.apply(lambda x: pool.map(reverse_gecode, [x[0], x[1]]), axis = 1)

但是我什么都做不到。我一直得到的错误是

ValueError: ('Unknown location: 27.950575', u'occurred at index 0')

Answer 1

尝试是：

import geocoder
import multiprocessing as mp
import pandas as pd

def reverse_gecode(coordinates):
    return geocoder.google(coordinates, method = 'reverse').postal

if __name__ == '__main__':            
    gps = pd.DataFrame({'lat': [27.950575, 40.6936488],
                        'lon': [-82.4571776, -89.5889864]})  # dataframe mehtod
    coords = gps[['lat','lon']].astype(str).apply(lambda x: (x[0],x[1]), axis=1).tolist()
    # multiprocessings                    
    pool = mp.Pool(processes = (mp.cpu_count() - 1))
    gps['zip_code'] = pool.map(reverse_gecode, coords)
    print(gps)
    pool.close()
    pool.join()

PS我在geocoder.google()调用中删除了key=google_key，因为它对我不起作用

输出：

^{pr2}$