根据Python中的lat long，groupby 2字段计算距离

Question

我的车辆跟踪系统有一组数据，需要我根据纬度和经度计算距离。 理解使用Haversine公式可以帮助获取行之间的距离，但由于需要基于2个字段（模型类型和模式）的距离，因此我有些困惑。

如下所示是我的代码：

def haversine(lat1,lon1,lat2,lon2, to_radians = True, earth_radius =6371):
    if to_radians:
        lat1,lon1,lat2,lon2 = np.radians([lat1,lon1,lat2,lon2])

    a = np.sin((lat2-lat1)/2.0)**2+ np.cos(lat1) * np.cos(lat2) * np.sin((lon2-lon1)/2.0)**2

    return earth_radius *2 * np.arcsin(np.sqrt(a))

mydataset = pd.read_csv(x + '.txt')
print (mydataset.shape)
mydataset = mydataset.sort_values(by=['Model','timestamp']) #sort
mydataset['dist'] = 
np.concatenate(mydataset.groupby(["Model"]).apply(lambda 
         x: haversine(x['Latitude'],x['Longitude'],
         x['Latitude'].shift(),x['Longitude'].shift())).values)

这样，我就能够基于模型（通过使用排序）计算行之间的距离。

但我想更进一步，根据“模式”和“模型”进行计算。 我的字段是“索引，模型，模式，纬度，经度，时间戳”

请指教！

Index, Model, Timestamp, Long, Lat, Mode(denote as 0 or 2), Distance Calculated
1, X, 2018-01-18 09:16:37.070, 103.87772815, 1.35653496, 0, 0.0
2, X, 2018-01-18 09:16:39.071, 103.87772815, 1.35653496, 0, 0.0
3, X, 2018-01-18 09:16:41.071, 103.87772815, 1.35653496, 0, 0.0
4, X, 2018-01-18-09:16:43.071, 103.87772052, 1.35653496, 0, 0.0008481795
5, X, 2018-01-18 09:16:45.071, 103.87770526, 1.35653329, 0, 0.0017064925312804799
6, X, 2018-01-18 09:16:51.070, 103.87770526, 1.35653329, 2, 0.0
7, X, 2018-01-18 09:16:53.071, 103.87770526, 1.35653329, 2, 0.0
8, X, 2018-01-18 09:59:55.072, 103.87770526, 1.35652828, 0, 0.0005570865824842293

无论哪种模式，我都需要它来计算模型的总行程距离以及模型的总行程距离

Answer 1

我认为需要添加DataFrame构造函数，然后通过列名将另一个列名添加到groupby例如["Model", "Mode(denote as 0 or 2)"]或["Model", "Mode"] 。

def haversine(lat1,lon1,lat2,lon2, to_radians = True, earth_radius =6371):
    if to_radians:
        lat1,lon1,lat2,lon2 = np.radians([lat1,lon1,lat2,lon2])

    a = np.sin((lat2-lat1)/2.0)**2+ np.cos(lat1) * np.cos(lat2) * np.sin((lon2- 
    lon1)/2.0)**2

    return pd.DataFrame(earth_radius *2 * np.arcsin(np.sqrt(a)))


mydataset['dist'] = (mydataset.groupby(["Model", "Mode(denote as 0 or 2)"])
                              .apply(lambda x: haversine(x['Lat'],
                                                         x['Long'], 
                                                         x['Lat'].shift(),
                                                         x['Long'].shift())).values)

#if need replace NaNs to 0  
mydataset['dist'] = mydataset['dist'].fillna(0)

---

print (mydataset)
   Index Model               Timestamp        Long       Lat  \
0      1     X 2018-01-18 09:16:37.070  103.877728  1.356535   
1      2     X 2018-01-18 09:16:39.071  103.877728  1.356535   
2      3     X 2018-01-18 09:16:41.071  103.877728  1.356535   
3      4     X 2018-01-18 09:16:43.071  103.877721  1.356535   
4      5     X 2018-01-18 09:16:45.071  103.877705  1.356533   
5      6     X 2018-01-18 09:16:51.070  103.877705  1.356533   
6      7     X 2018-01-18 09:16:53.071  103.877705  1.356533   
7      8     X 2018-01-18 09:59:55.072  103.877705  1.356528   

   Mode(denote as 0 or 2)  Distance Calculated      dist  
0                       0             0.000000  0.000000  
1                       0             0.000000  0.000000  
2                       0             0.000000  0.000000  
3                       0             0.000848  0.000848  
4                       0             0.001706  0.001706  
5                       2             0.000000  0.000557  
6                       2             0.000000  0.000000  
7                       0             0.000557  0.000000