notebook.community

Edit and run

merge dataframe and split again. Useful for merging test and train data to create panel. Then apply transformations on panel in one go. Finally split the panel back into train and test dataframes.



In [1]:

    
import pandas as pd
import numpy as np



In [2]:

    
ts1 = [1,2,3,4]
ts2 = [6,7,8,9]
d = {'col_1': ts1, 'col_2': ts2}



In [3]:

    
d









    Out[3]:





{'col_1': [1, 2, 3, 4], 'col_2': [6, 7, 8, 9]}



In [4]:

    
df_1 = pd.DataFrame(data=d)



In [5]:

    
df_1



In [6]:

    
df_2 = pd.DataFrame(np.random.randn(3, 2), columns=['col_1', 'col_2'])



In [7]:

    
df_2



In [8]:

    
df_all = pd.concat((df_1, df_2), axis=0, ignore_index=True)



In [9]:

    
df_all



In [10]:

    
print(df_1.shape)
print(df_2.shape)
print(df_all.shape)









    



(4, 2)
(3, 2)
(7, 2)



In [11]:

    
df_train = df_all[:df_1.shape[0]]
df_test = df_all[df_1.shape[0]:]



In [12]:

    
print(df_train.shape)
print(df_test.shape)
print(df_all.shape)









    



(4, 2)
(3, 2)
(7, 2)



In [ ]:

	col_1	col_2
0	0.654547	-1.201099
1	-0.088006	-0.049599
2	0.609881	-1.003260

	col_1	col_2
0	1.000000	6.000000
1	2.000000	7.000000
2	3.000000	8.000000
3	4.000000	9.000000
4	0.654547	-1.201099
5	-0.088006	-0.049599
6	0.609881	-1.003260