Question

我喜欢：

public List<string> listGhe(string ten, int movie_id)
        {
            slotList.Clear();
            dbConnection.Open();
            try
            {
                string sql = "Select b.Ghe_num from Customers c inner JOIN bookings b ON b.cus_id = c.id and movie_id = " + movie_id + " WHERE c.`name` = '" + ten + "'";
                MySqlCommand cmd = new MySqlCommand();
                cmd.Connection = dbConnection;
                cmd.CommandText = sql;
                using (DbDataReader reader = cmd.ExecuteReader())
                {
                    if (reader.HasRows)
                    {
                        while (reader.Read())
                        {
                            int gheNumIndex = reader.GetOrdinal("Ghe_num");
                            string gheNum = reader.GetString(gheNumIndex);
                            slotList.Add(gheNum);
                        }
                    }
                }
            }
            catch
            {

            }
            finally
            {
                dbConnection.Close();
                dbConnection.Dispose();
                //dbConnection = null;
            }

            return slotList;
        }

和df1喜欢：

material plant  Order                
24990   89952   4568789,5098710     
24990   89952   9448609,1007081     
166621  3062    18364103            
166621  3062    78309139            
240758  3062    55146035            
276009  3062    38501581,857542

我想在df1中迭代Order，当df2中有Order匹配时，从m1到m5找到平均值。我想实现df2，如：

material plant   Order      m1     m2      m3   m4   m5
24990     89952 4568789     0.123  0.214   0.0  0.0  0.0
24990     89952 5098710     1.000  0.363   0.0  0.0  0.0
24990     89952 9448609     0.0    0.345   0.0  1.0  0.0
24990     89952 1007081     0.0    0.756   0.0  1.0  0.0
166621    3062  18364103    0.0    0.0     0.0  0.0  0.0
166621    3062  78309139    0.0    1.0     0.0  0.0  0.0
240758    3062  55146035    1.0    1.0     1.0  0.0  0.0
276009    3062  38501581    1.0    1.0     1.0  0.0  0.0
276009    3062  38575428    1.0    1.0     1.0  0.0  0.0

我正在尝试不同的方法来实现df2，例如：

material plant  Order              avg m1 avgm2 avgm3 avgm4 avgm5
24990   89952   4568789,5098710    0.5615 0.2885 0.0   0.0   0.0
24990   89952   9448609,1007081     
166621  3062    18364103            
166621  3062    78309139            
240758  3062    55146035            
276009  3062    38501581,857542

第二

df2 = (df.groupby(df1, sort=False)['Order'].apply(lambda x: ','.split(x.astype(str)))
   .mean() 
   .reset_index()
.reindex(columns=df.columns))
 print (df2)

但我不确定这是否是正确的方法。

Answer 1

这是通过numpy和映射字典的一种方式。

# map Order to values with a dictionary
mapper = dict(zip(df1['Order'], df1[['m'+str(i) for i in range(1, 6)]].values))

# map comma-separated numbers to list of integers
df2_orders = [list(map(int, i)) for i in df2['Order'].str.split(',')]

# calculate mean
res = [np.mean([mapper.get(o, [0]*5) for o in order], axis=0).tolist() \
       for order in df2_orders]

# join results to dataframe
df2 = df2.join(pd.DataFrame(res, columns=['avg_m'+str(i) for i in range(1, 6)]))

注意，如果缺少数据（例如，订单857542），您可以指定要包含的值，这里我使用0。

<强>结果

   material  plant            Order  avg_m1  avg_m2  avg_m3  avg_m4  avg_m5
0     24990  89952  4568789,5098710  0.5615  0.2885     0.0     0.0     0.0
1     24990  89952  9448609,1007081  0.0000  0.5505     0.0     1.0     0.0
2    166621   3062         18364103  0.0000  0.0000     0.0     0.0     0.0
3    166621   3062         78309139  0.0000  1.0000     0.0     0.0     0.0
4    240758   3062         55146035  1.0000  1.0000     1.0     0.0     0.0
5    276009   3062  38501581,857542  0.5000  0.5000     0.5     0.0     0.0

Answer 2

您可以使用：

df = (df1.join(df1.set_index(['material','plant'], append=True)['Order']
          .str.split(',', expand=True)
          .stack()
          .astype(int)
          .reset_index(name='Order')
          .merge(df2, on=['material','plant','Order'], how='left') 
          .drop(['material','plant','Order','level_3'], axis=1)
          .groupby('level_0')
          .mean())
          )
print (df)
   material  plant            Order      m1      m2   m3   m4   m5
0     24990  89952  4568789,5098710  0.5615  0.2885  0.0  0.0  0.0
1     24990  89952  9448609,1007081  0.0000  0.5505  0.0  1.0  0.0
2    166621   3062         18364103  0.0000  0.0000  0.0  0.0  0.0
3    166621   3062         78309139  0.0000  1.0000  0.0  0.0  0.0
4    240758   3062         55146035  1.0000  1.0000  1.0  0.0  0.0
5    276009   3062  38501581,857542  1.0000  1.0000  1.0  0.0  0.0

<强>解释：

首先按split和stack

DataFrame

使用merge加入第二个DataFrame并离开加入
使用drop
然后汇总mean
输出系列join为DataFrame

<强>详细：

df3 = (df1.set_index(['material','plant'], append=True)['Order']
          .str.split(',', expand=True)
          .stack()
          .astype(int)
          .reset_index(name='Order')
          .merge(df2, on=['material','plant','Order'], how='left'))
print (df3)
   level_0  material  plant  level_3     Order     m1     m2   m3   m4   m5
0        0     24990  89952        0   4568789  0.123  0.214  0.0  0.0  0.0
1        0     24990  89952        1   5098710  1.000  0.363  0.0  0.0  0.0
2        1     24990  89952        0   9448609  0.000  0.345  0.0  1.0  0.0
3        1     24990  89952        1   1007081  0.000  0.756  0.0  1.0  0.0
4        2    166621   3062        0  18364103  0.000  0.000  0.0  0.0  0.0
5        3    166621   3062        0  78309139  0.000  1.000  0.0  0.0  0.0
6        4    240758   3062        0  55146035  1.000  1.000  1.0  0.0  0.0
7        5    276009   3062        0  38501581  1.000  1.000  1.0  0.0  0.0
8        5    276009   3062        1    857542    NaN    NaN  NaN  NaN  NaN

两个不同数据帧的平均值

2 个答案: