Python Pandas 基于另一列的唯一列值计数

Question

我有一张桌子（下面的例子）

    kick_result kick_yards  kicker
50     MADE       28.0       X1
64     MADE       30.0       X2
75     MADE       27.0       X2
158    MADE       32.0       X2
259    MISS       46.0       X3

对于 kicker 的每个值 -

我想计算有多少投篮命中和错失（%）

对于 kicker 的每个值 -

我想找出每个码数范围 <20、21-30、31-40、41-50、51+ 的投篮命中数和未命中数

Answer 1

让我们用cut和crosstab链接

out = pd.crosstab([df.kicker,pd.cut(df.kick_yards,[20,30,40,50,np.Inf],include_lowest=True)]
                   ,df.kick_result,normalize='index')
 
out
Out[228]: 
kick_result            MADE  MISS
kicker kick_yards                
X1     (19.999, 30.0]   1.0   0.0
X2     (19.999, 30.0]   1.0   0.0
       (30.0, 40.0]     1.0   0.0
X3     (40.0, 50.0]     0.0   1.0

Answer 2

由于您的要求有两个部分：

1. 投篮命中率%，和
2. 码范围内的射门得分，

让我们一一解决。

第 1 部分：射门得分 %

我们可以使用df.groupby()和.value_counts(normalize=True)来得到它：

(df.groupby('kicker')['kick_result']
   .value_counts(normalize=True).mul(100).round(2)
   .sort_index()
   .to_frame(name='Result_%')
).reset_index()

测试运行

测试数据构建：

为了对各种需求进行完整的测试，我添加了如下测试数据：

    kick_result kick_yards  kicker
49     MADE       18.0       X1 
50     MADE       28.0       X1
51     MADE       38.0       X1
52     MISS       48.0       X1
53     MISS       58.0       X1
64     MADE       30.0       X2
75     MADE       27.0       X2
158    MADE       32.0       X2
159    MISS       32.0       X2
160    MISS       42.0       X2
259    MISS       46.0       X3
260    MISS       26.0       X3
261    MADE       56.0       X3

运行代码：

(df.groupby('kicker')['kick_result']
   .value_counts(normalize=True).mul(100).round(2)
   .sort_index()
   .to_frame(name='Result_%')
).reset_index()

结果：

  kicker kick_result  Result_%
0     X1        MADE     60.00
1     X1        MISS     40.00
2     X2        MADE     60.00
3     X2        MISS     40.00
4     X3        MADE     33.33
5     X3        MISS     66.67

第 2 部分：码范围内的射门得分

我们可以使用pd.crosstab()和pd.cut()来构建一个包含码范围的表格。

还包括所有范围的总尝试次数。

pd.crosstab(index=[df['kicker'], pd.cut(df['kick_yards'],[0, 20, 30, 40, 50, np.inf])], 
            columns=df['kick_result'], 
            margins=True, margins_name='Total_Attempts')

结果（使用丰富的测试数据）：

                        kick_result  MADE  MISS  Total_Attempts
        kicker           kick_yards                              
            X1          (0.0, 20.0]     1     0               1
                       (20.0, 30.0]     1     0               1
                       (30.0, 40.0]     1     0               1
                       (40.0, 50.0]     0     1               1
                        (50.0, inf]     0     1               1
            X2         (20.0, 30.0]     2     0               2
                       (30.0, 40.0]     1     1               2
                       (40.0, 50.0]     0     1               1
            X3         (20.0, 30.0]     0     1               1
                       (40.0, 50.0]     0     1               1
                        (50.0, inf]     1     0               1
Total_Attempts                          7     6              13

Answer 3

利用get_dummies ， cut并构建生成的DataFrame ：

df['Att'] = 1

dfmm = pd.get_dummies(df['kick_result'])

cols_A = ['A20','A21-30','A31-40','A41-50','A51+']
cols_M = [x.replace('A','M') for x in cols_A]

df_att = pd.DataFrame(pd.get_dummies(pd.cut(df.kick_yards,[0,20,30,40,50,np.Inf],include_lowest=True)))
df_att.columns = df_att.columns.to_list()
df_att.columns = cols_A

df_made = df_att.multiply(dfmm['MADE'], axis=0)
df_made.columns=cols_M
    
dff = pd.concat([df,dfmm,df_att,df_made], axis=1).drop(['kick_result','kick_yards'], axis=1)

结果 DataFrame：

  kicker  Att  MADE  MISS  A20  A21-30  A31-40  A41-50  A51+  M20  M21-30  \
0     X1    1     1     0    0       1       0       0     0    0       1   
1     X2    1     1     0    0       1       0       0     0    0       1   
2     X2    1     1     0    0       1       0       0     0    0       1   
3     X2    1     1     0    0       0       1       0     0    0       0   
4     X3    1     0     1    0       0       0       1     0    0       0   

   M31-40  M41-50  M51+  
0       0       0     0  
1       0       0     0  
2       0       0     0  
3       1       0     0  
4       0       0     0  

来自该 DataFrame 的聚合：

dff.groupby('kicker').agg(['sum'])

       Att MADE MISS A20 A21-30 A31-40 A41-50 A51+ M20 M21-30 M31-40 M41-50  \
       sum  sum  sum sum    sum    sum    sum  sum sum    sum    sum    sum   
kicker                                                                        
X1       1    1    0   0      1      0      0    0   0      1      0      0   
X2       3    3    0   0      2      1      0    0   0      2      1      0   
X3       1    0    1   0      0      0      1    0   0      0      0      0   

       M51+  
        sum  
kicker       
X1        0  
X2        0  
X3        0