Question

我有两个要合并在一起的数据框（“自行车骑行”数据框和“自行车站”数据框）。

我一直在使用pandas库，但是我似乎无法编写代码来完美地操纵联接。最初，我只是加入键“ station_id”，但是我发现了一个更新的站点数据集，其中包含更多的站点，问题是有些站点没有station_id。对于这些站点，我想加入对纬度和经度坐标的匹配。

我刚使用station_id加入数据帧时的初始代码

rides_df = rides_df.rename(columns = {'start_station_id': 'station_id'})
rides_df = rides_df.merge(stations_df[['station_id','station_name']],
                         on = 'station_id', how = 'left')
rides_df = rides_df.rename(columns = {'station_id':'start_station_id',
                                     'station_name':'station_name_start'})

#merge ending station name
rides_df = rides_df.rename(columns = {'end_station_id': 'station_id'})
rides_df = rides_df.merge(stations_df[['station_id', 'station_name']],
                         on = 'station_id', how = 'left')
rides_df = rides_df.rename(columns = {'station_id':'end_station_id',
                                     'station_name': 'station_name_end'})

游乐设施数据帧的结构如下（采样）：

rides_df = pd.DataFrame([[1912818,'Round Trip',3014,34.0566101,-118.23721,3014,34.0566101,-118.23721],
                 [1933383,'Round Trip',3016,34.0528984,-118.24156,3016,34.0528984,-118.24156],
                  [1944197,'Round Trip',3016,34.0528984,-118.24156,3016,34.0528984,-118.24156],
                  [1940317,'Round Trip','NaN',34.03352,-118.24184,'NaN',34.03352,-118.24184],
                  [1944075,'One Way',3021,34.0456085,-118.23703,3016,34.0566101,-118.23721]]
                 , columns = ['trip_id','trip_route_category','start_station_id','start_lat',
                              'start_lon','end_station_id','end_lat','end_lon'])

台站数据帧的结构如下（采样）：

stations_df = pd.DataFrame([['Union Station West Portal',34.05661,-118.23721,3014],
                            ['Los Angeles & Temple',34.0529,-118.24156,3016],
                            ['Grand & Olympic',34.04373,-118.26014,3018],
                            ['12th & Hill',34.03861,-118.26086,3019],
                            ['Hill & Washington',34.03105,-118.26709,3020],
                            ['Row DTLA',34.03352,-118.24184,'NaN']], 
                           columns = ['station_name', 'lat', 'lon','station_id'])

我想要的是在游乐设施数据帧上添加起点和终点位置的站点名称，这样我将在“ Start_Station_Name”和“ End_Station_Name”列中添加一个。我想加入“ station_id”，但如果station_id为NaN，则匹配lat＆lon作为开始和结束位置。

The data frame that I want as a result is structured as follows:
want_df = pd.DataFrame([[1912818,'Round Trip','Union Station West Portal',3014,34.0566101,-118.23721,'Union Station West Portal',3014,34.0566101,-118.23721],
                 [1933383,'Round Trip','Los Angeles & Temple',3016,34.0528984,-118.24156,'Los Angeles & Temple',3016,34.0528984,-118.24156],
                  [1944197,'Round Trip','Los Angeles & Temple',3016,34.0528984,-118.24156,'Los Angeles & Temple',3016,34.0528984,-118.24156],
                  [1940317,'Round Trip','Row DTLA','Nan',34.03352,-118.24184,'Row DTLA','Nan',34.03352,-118.24184],
                [1944075,'One Way','NaN',3021,34.0456085,-118.23703,'Los Angeles & Temple',3016,34.0566101,-118.23721]]
                 , columns = ['trip_id','trip_route_category','start_station_name','start_station_id','start_lat',
                              'start_lon','end_station_name','end_station_id','end_lat','end_lon'])

Answer 1

这是实现此目的的代码更新版本：

post_title

输出：

post_content

多列+条件联接

1 个答案: