Pandas: find the maximum value in each row of a column and determine the corresponding values in another column
I have two columns of a pandas dataframe for which the values are lists of tuples such as:
df[‘A’].values
(1.55, 2.07, 2.20, 2.23)
(0.67, 1.10, 1.73, 1.35)
(2.92, 1.98, 2.30, 2.66)
and
df[‘B’].values
(1.55, 0.0086, 0.078, 0.12)
(0.672, 0.142, 0.0166, 0.0173)
(1.97, 0.0094, 0.1648, 0.016)
I would like to select the largest value for each row df['A']
and find the value at the appropriate position df['B']
to generate a new column like:
> df[‘C’]
0.12
0.0166
1.97
+3
source to share
1 answer
import pandas as pd
import numpy as np
df_dict = dict(A=[(1.55, 2.07, 2.20, 2.23), (0.67, 1.10, 1.73, 1.35), (2.92, 1.98, 2.30, 2.66)],
B=[(1.55, 0.0086, 0.078, 0.12), (0.672, 0.142, 0.0166, 0.0173), (1.97, 0.0094, 0.1648, 0.016)])
df = pd.DataFrame(df_dict)
Out[180]:
A B
0 (1.55, 2.07, 2.2, 2.23) (1.55, 0.0086, 0.078, 0.12)
1 (0.67, 1.1, 1.73, 1.35) (0.672, 0.142, 0.0166, 0.0173)
2 (2.92, 1.98, 2.3, 2.66) (1.97, 0.0094, 0.1648, 0.016)
def apply_func(row):
return row.B[np.array(row.A).argmax()]
df['C'] = df.apply(apply_func, axis=1)
Out[182]:
A B C
0 (1.55, 2.07, 2.2, 2.23) (1.55, 0.0086, 0.078, 0.12) 0.1200
1 (0.67, 1.1, 1.73, 1.35) (0.672, 0.142, 0.0166, 0.0173) 0.0166
2 (2.92, 1.98, 2.3, 2.66) (1.97, 0.0094, 0.1648, 0.016) 1.9700
+7
source to share