Python 从变量中的字符串公式计算dataframe列？

我使用一个excel文件确定传感器的名称，并使用一个公式根据真实传感器创建一个新的“合成”传感器。我想把公式写成字符串，比如“y1+y2+y3”，而不是“df['y1']+df['y2']+df['y3]”，但我不知道该使用哪种方法

Excel文件示例：

因此，我的脚本必须为此excel文件的每一行创建一个新传感器。这个新的传感器将被上传到我的数据库。计算新值的传感器数量是可变的

以下是我的代码示例：

# From excel file
sensor_cell = '008253_sercit_sercit_batg_b_flr0_g_tctl_z1_tr_prval|officielles_darksky_bierset_temp|officielles_darksky_uccle_temp|005317_esa_001_hur_piece_030110'
formula_cell = "df['y1'] + df['y2'] + df['y3'] + df['y4']"
# formula_cell = 'y1+y2+y3+y4' --> what I would like to be able to write in my  excel file cell

list = sensor_cell.split('|')

df = []
for sensor in list:
    position = list.index(sensor) + 1
    df_y = search_ES(sensor) # Function that return a df with timestamp and value from my database
    df_y = df_y.rename(columns={'value': "y"+ str(position)})
    df.append(df_y)

df = pd.concat(df, axis=1, sort=True)
# I would like to have :
# y1 = df['y1']
# y2 = df['y2']
# y3 = df['y3']
# y4 = df['y4']
df = df.dropna()
print(df)
df['value'] = eval(formula_cell) # Formula from excel file
print(df)

df在应用公式之前：

                           y1    y2    y3  y4
2019-12-11 00:00:00  20.500000  5.62  6.03  29
2019-12-11 01:00:00  21.180000  5.54  6.15  30
2019-12-11 02:00:00  21.020000  5.28  6.29  30
2019-12-11 03:00:00  20.760000  4.99  6.36  29
2019-12-11 04:00:00  20.680000  4.80  6.26  30
2019-12-11 05:00:00  20.760000  4.63  6.07  30
2019-12-11 06:00:00  20.900000  4.49  5.91  30
2019-12-11 07:00:00  20.920000  4.20  6.05  30
2019-12-11 08:00:00  21.320000  4.15  5.95  30
2019-12-11 09:00:00  21.840000  4.42  5.81  30
2019-12-11 10:00:00  22.460000  4.24  5.81  30
2019-12-11 11:00:00  22.240000  4.11  5.89  31
2019-12-11 12:00:00  22.420000  4.43  6.15  32
2019-12-11 13:00:00  21.740000  4.37  6.14  32
2019-12-11 14:00:00  22.500000  4.48  6.24  31
2019-12-11 15:00:00  22.980000  4.87  6.46  32
2019-12-11 16:00:00  22.420000  4.56  6.21  32
2019-12-11 17:00:00  22.320000  4.40  5.92  32
2019-12-11 18:00:00  21.939999  4.52  6.19  32
2019-12-11 19:00:00  20.680000  4.30  5.35  32
2019-12-11 20:00:00  20.900000  4.28  4.94  32
2019-12-11 21:00:00  20.859999  4.55  5.21  32
2019-12-11 22:00:00  20.520000  4.28  4.73  32
2019-12-11 23:00:00  20.320000  4.24  4.90  32

                          y1    y2    y3  y4      value
2019-12-11 00:00:00  20.500000  5.62  6.03  29  61.150000
2019-12-11 01:00:00  21.180000  5.54  6.15  30  62.870000
2019-12-11 02:00:00  21.020000  5.28  6.29  30  62.590000
2019-12-11 03:00:00  20.760000  4.99  6.36  29  61.110000
2019-12-11 04:00:00  20.680000  4.80  6.26  30  61.740000
2019-12-11 05:00:00  20.760000  4.63  6.07  30  61.460000
2019-12-11 06:00:00  20.900000  4.49  5.91  30  61.300000
2019-12-11 07:00:00  20.920000  4.20  6.05  30  61.170000
2019-12-11 08:00:00  21.320000  4.15  5.95  30  61.420000
2019-12-11 09:00:00  21.840000  4.42  5.81  30  62.070000
2019-12-11 10:00:00  22.460000  4.24  5.81  30  62.510000
2019-12-11 11:00:00  22.240000  4.11  5.89  31  63.240000
2019-12-11 12:00:00  22.420000  4.43  6.15  32  65.000000
2019-12-11 13:00:00  21.740000  4.37  6.14  32  64.250000
2019-12-11 14:00:00  22.500000  4.48  6.24  31  64.220000
2019-12-11 15:00:00  22.980000  4.87  6.46  32  66.310000
2019-12-11 16:00:00  22.420000  4.56  6.21  32  65.190000
2019-12-11 17:00:00  22.320000  4.40  5.92  32  64.640000
2019-12-11 18:00:00  21.939999  4.52  6.19  32  64.649999
2019-12-11 19:00:00  20.680000  4.30  5.35  32  62.330000
2019-12-11 20:00:00  20.900000  4.28  4.94  32  62.120000
2019-12-11 21:00:00  20.859999  4.55  5.21  32  62.619999
2019-12-11 22:00:00  20.520000  4.28  4.73  32  61.530000
2019-12-11 23:00:00  20.320000  4.24  4.90  32  61.460000

应用公式后的df：

                           y1    y2    y3  y4
2019-12-11 00:00:00  20.500000  5.62  6.03  29
2019-12-11 01:00:00  21.180000  5.54  6.15  30
2019-12-11 02:00:00  21.020000  5.28  6.29  30
2019-12-11 03:00:00  20.760000  4.99  6.36  29
2019-12-11 04:00:00  20.680000  4.80  6.26  30
2019-12-11 05:00:00  20.760000  4.63  6.07  30
2019-12-11 06:00:00  20.900000  4.49  5.91  30
2019-12-11 07:00:00  20.920000  4.20  6.05  30
2019-12-11 08:00:00  21.320000  4.15  5.95  30
2019-12-11 09:00:00  21.840000  4.42  5.81  30
2019-12-11 10:00:00  22.460000  4.24  5.81  30
2019-12-11 11:00:00  22.240000  4.11  5.89  31
2019-12-11 12:00:00  22.420000  4.43  6.15  32
2019-12-11 13:00:00  21.740000  4.37  6.14  32
2019-12-11 14:00:00  22.500000  4.48  6.24  31
2019-12-11 15:00:00  22.980000  4.87  6.46  32
2019-12-11 16:00:00  22.420000  4.56  6.21  32
2019-12-11 17:00:00  22.320000  4.40  5.92  32
2019-12-11 18:00:00  21.939999  4.52  6.19  32
2019-12-11 19:00:00  20.680000  4.30  5.35  32
2019-12-11 20:00:00  20.900000  4.28  4.94  32
2019-12-11 21:00:00  20.859999  4.55  5.21  32
2019-12-11 22:00:00  20.520000  4.28  4.73  32
2019-12-11 23:00:00  20.320000  4.24  4.90  32

                          y1    y2    y3  y4      value
2019-12-11 00:00:00  20.500000  5.62  6.03  29  61.150000
2019-12-11 01:00:00  21.180000  5.54  6.15  30  62.870000
2019-12-11 02:00:00  21.020000  5.28  6.29  30  62.590000
2019-12-11 03:00:00  20.760000  4.99  6.36  29  61.110000
2019-12-11 04:00:00  20.680000  4.80  6.26  30  61.740000
2019-12-11 05:00:00  20.760000  4.63  6.07  30  61.460000
2019-12-11 06:00:00  20.900000  4.49  5.91  30  61.300000
2019-12-11 07:00:00  20.920000  4.20  6.05  30  61.170000
2019-12-11 08:00:00  21.320000  4.15  5.95  30  61.420000
2019-12-11 09:00:00  21.840000  4.42  5.81  30  62.070000
2019-12-11 10:00:00  22.460000  4.24  5.81  30  62.510000
2019-12-11 11:00:00  22.240000  4.11  5.89  31  63.240000
2019-12-11 12:00:00  22.420000  4.43  6.15  32  65.000000
2019-12-11 13:00:00  21.740000  4.37  6.14  32  64.250000
2019-12-11 14:00:00  22.500000  4.48  6.24  31  64.220000
2019-12-11 15:00:00  22.980000  4.87  6.46  32  66.310000
2019-12-11 16:00:00  22.420000  4.56  6.21  32  65.190000
2019-12-11 17:00:00  22.320000  4.40  5.92  32  64.640000
2019-12-11 18:00:00  21.939999  4.52  6.19  32  64.649999
2019-12-11 19:00:00  20.680000  4.30  5.35  32  62.330000
2019-12-11 20:00:00  20.900000  4.28  4.94  32  62.120000
2019-12-11 21:00:00  20.859999  4.55  5.21  32  62.619999
2019-12-11 22:00:00  20.520000  4.28  4.73  32  61.530000
2019-12-11 23:00:00  20.320000  4.24  4.90  32  61.460000

编辑-解决方案： 使用

df.eval

解决了该问题：

formula_cell = 'fictive = y1+y2+y3+y4' 
df.eval(formula_cell, inplace=True)

---

我想熊猫就是你想要的

例如：

formula_cell = "y1 + y2 + y3 + y4"
df['value'] = df.query(formula_cell)