索引到新列时出现TypeError

Question

我在df中生成了两列，它们由“Player”列中的值填充。因此，当一个值在这列中时，我想返回该索引点处的XY

下面的代码适用于示例数据集：

import pandas as pd
import numpy as np
import random 

k = 5
N = 10

df = pd.DataFrame({ 'Frame' : range(1, N + 1 ,1),
    'John Doe_X' : np.random.uniform(k, k + 100 , size=N),
    'John Doe_Y' : np.random.uniform(k, k + 100 , size=N),
    'Kevin Lee_X' : np.random.uniform(k, k + 100 , size=N),
    'Kevin Lee_Y' : np.random.uniform(k, k + 100 , size=N),   
    'Liam Smith_X' : np.random.uniform(k, k + -100 , size=N),
    'Liam Smith_Y' : np.random.uniform(k, k + 100 , size=N),
    'Henry Conr_X' : np.random.uniform(k, k + 100 , size=N),
    'Henry Conr_Y' : np.random.uniform(k, k + 100 , size=N),
    'Dan On_X' : np.random.uniform(k, k + 100 , size=N),
    'Dan On_Y' : np.random.uniform(k, k + 100 , size=N),
    'Player' : ['nan','John Doe','nan','Liam Smith','Henry john-Conr','John Doe','nan','nan','Kevin Lee','Kevin Lee']})


df['X'] = df.apply(lambda row: row.get(row['Player']+'_X'), axis=1)
df['Y'] = df.apply(lambda row: row.get(row['Player']+'_Y'), axis=1)

输出：

    Dan On_X   Dan On_Y  Henry john-Conr_X  Henry john-Conr_Y  John Doe_X  \
0  44.768998  84.473350          73.295603           9.432975   45.098229   
1  22.078255  83.308328          68.302211          96.222105   45.511418   
2   8.958403  32.589984          36.570954          11.388089   76.967684   
3  88.710772  39.136471          46.045362          11.080126   73.837371   
4  51.959727  95.460223           8.558665          57.413331   43.943096   
5  32.703784  50.699055          54.201836          81.972841   35.237445   
6  10.533631  53.555805          81.192208          63.985641   88.664571   
7  31.100324  76.484070          62.607465          59.920721   36.455336   
8  58.747071  16.464910         104.331561         102.078546   28.662882   
9  28.593463  15.175946          87.434676         101.343336   63.139973   

   John Doe_Y  Kevin Lee_X  Kevin Lee_Y  Liam Smith_X  Liam Smith_Y  \
0   32.444921    75.818689    56.684665      1.367917     67.193555   
1   48.496147    17.040589    13.027425    -16.125411     75.731070   
2  103.802791    24.569848    89.321086    -32.929826     28.512003   
3  103.249107    64.077139    81.212017    -11.560951     18.367170   
4   21.053600    65.848794    42.016315    -17.002813     67.181597   
5   38.703567    43.069560    74.596956    -34.210479     91.396546   
6   35.942260    22.064457    24.871318     -4.745078     81.645923   
7   78.317289     5.980761    74.924581    -72.922074     97.187079   
8   53.819660   102.053136    77.254865    -57.583882     24.007903   
9   20.381099    31.395646    27.660293    -20.299473     19.465666   

            Player           X          Y  
0              nan         NaN        NaN  
1         John Doe   45.511418  48.496147  
2              nan         NaN        NaN  
3       Liam Smith  -11.560951  18.367170  
4  Henry john-Conr    8.558665  57.413331  
5         John Doe   35.237445  38.703567  
6              nan         NaN        NaN  
7              nan         NaN        NaN  
8        Kevin Lee  102.053136  77.254865  
9        Kevin Lee   31.395646  27.660293 

但当我导入数据集时，它会返回一个TypeError：

TypeError: ("unsupported operand type(s) for +: 'float' and 'str'", 'occurred at index 0')

我检查了每个数据类型是否相同。布局也是如此。所有的值也是浮动的

Answer 1

有些值是NaN的，因此建议只在列中添加string：

df = pd.DataFrame({ 'Frame' : range(1, N + 1 ,1),
    'John Doe_X' : np.random.uniform(k, k + 100 , size=N),
    'John Doe_Y' : np.random.uniform(k, k + 100 , size=N),
    'Kevin Lee_X' : np.random.uniform(k, k + 100 , size=N),
    'Kevin Lee_Y' : np.random.uniform(k, k + 100 , size=N),   
    'Liam Smith_X' : np.random.uniform(k, k + -100 , size=N),
    'Liam Smith_Y' : np.random.uniform(k, k + 100 , size=N),
    'Henry Conr_X' : np.random.uniform(k, k + 100 , size=N),
    'Henry Conr_Y' : np.random.uniform(k, k + 100 , size=N),
    'Dan On_X' : np.random.uniform(k, k + 100 , size=N),
    'Dan On_Y' : np.random.uniform(k, k + 100 , size=N),
    'Player' : [np.nan,'John Doe','nan','Liam Smith','Henry Conr','John Doe','nan','nan','Kevin Lee','Kevin Lee']})

df['X'] = df['Player']+'_X'
df['Y'] = df['Player']+'_Y'
print (df)
   Frame  John Doe_X      ...                  X             Y
0      1   10.240240      ...                NaN           NaN
1      2   91.760964      ...         John Doe_X    John Doe_Y
2      3   76.286604      ...              nan_X         nan_Y
3      4   24.907703      ...       Liam Smith_X  Liam Smith_Y
4      5   57.319837      ...       Henry Conr_X  Henry Conr_Y
5      6   18.380069      ...         John Doe_X    John Doe_Y
6      7    6.361868      ...              nan_X         nan_Y
7      8   24.305204      ...              nan_X         nan_Y
8      9   22.669279      ...        Kevin Lee_X   Kevin Lee_Y
9     10   97.934949      ...        Kevin Lee_X   Kevin Lee_Y

如果需要apply添加if-else，但作为第一个解决方案，它比较慢：

df['X'] = df.apply(lambda row: row.get(row['Player']+'_X') if pd.notnull(row['Player']) else np.nan, axis=1)
df['Y'] = df.apply(lambda row: row.get(row['Player']+'_Y') if pd.notnull(row['Player']) else np.nan, axis=1)