how to normalize panda dataframe based on another dataframe-CodePudding

there is two panda dataframe . I used MinMaxScaler to normalize first dataframe in purpose of training a neural network . and for test dataset i need to do the same but how can I scale dataframe base on min and max of the first dataframe ?
and because test data should not effect training , can not merge two dataframes , then scale and split again
datasets have a lot of columns
example :
first dataframe :

|     | colA |
| --- |---- |
| 1   | 3   |
| 2   | 10  |
| 3   | 4   |
| 4   | 0   |

second dataframe:

|     | colA |
| --- |--- |
| 1   | 2  |
| 2   | 5  |

expected scaling :

|     | colA |
| --- | --- |
| 1   | 0.2 |
| 2   | 0.5 |

CodePudding user response：

import pandas as pd
from sklearn.preprocessing import MinMaxScaler

train_df = pd.DataFrame({'colA': [3, 10, 4, 0]})
test_df = pd.DataFrame({'colA': [2, 5]})

scaler = MinMaxScaler()
scaler.fit(train_df)

train_df = scaler.transform(train_df)
test_df = scaler.transform(test_df)