Row wise and column wise binding of various data frames in Python

There are situations where we have to append the intermediate results of a loop operation and create a single data frame.In this blog, we will look at how we can do this using pandas¶

In [1]:

import pandas as pd
import numpy as np

Step 1: Creating dummy data for row wise binding¶

In [17]:

row_df1=pd.DataFrame([np.array(["Product_Category","Product_Category","Product_Category"]),
                  np.array(["PC","Print","Services"]),
                 np.array([50,30,20])]).T
row_df1.columns=['Feature','Levels','Perc_Share']
row_df1

Out[17]:

	Feature	Levels	Perc_Share
0	Product_Category	PC	50
1	Product_Category	Print	30
2	Product_Category	Services	20

In [18]:

row_df2=pd.DataFrame([np.array(["Region","Region","Region","Region"]),
                  np.array(["East","West","North","South"]),
                 np.array([50,25,15,10])]).T
row_df2.columns=['Feature','Levels','Perc_Share']
row_df2

Out[18]:

	Feature	Levels	Perc_Share
0	Region	East	50
1	Region	West	25
2	Region	North	15
3	Region	South	10

row_df1 and row_df2 are result of an operation where we are trying to understand percentage share of sales for Product Category and by Region.In order to consolidate results, we need to combined row_df1 and row_df2 into a single data frame¶

Step 2: Appending row_df1 and row_df2¶

In [19]:

row_wise_append=pd.concat([row_df1,row_df2],axis=0)
row_wise_append

Out[19]:

	Feature	Levels	Perc_Share
0	Product_Category	PC	50
1	Product_Category	Print	30
2	Product_Category	Services	20
0	Region	East	50
1	Region	West	25
2	Region	North	15
3	Region	South	10

Step 3: Creating dummy data for column wise binding¶

In [37]:

col_df1=pd.DataFrame([np.array(["Week1","Week1","Week1"]),
                      np.array(["PC","Laptop","Services"]),
                 np.random.normal(30,2,3)]).T
col_df1.columns=['Week','Product_Category','Sales_W1']
col_df1

Out[37]:

	Week	Product_Category	Sales_W1
0	Week1	PC	25.436445
1	Week1	Laptop	29.492091
2	Week1	Services	28.286619

In [38]:

col_df2=pd.DataFrame([np.array(["Week2","Week2","Week2"]),
                      np.array(["PC","Laptop","Services"]),
                 np.random.normal(30,2,3)]).T
col_df2.columns=['Week','Product_Category','Sales_W2']
col_df2

Out[38]:

	Week	Product_Category	Sales_W2
0	Week2	PC	29.962953
1	Week2	Laptop	27.875145
2	Week2	Services	33.107039

col_df1 and col_df2 are result of an operation where we are trying to understand Weekly sales for Product Category.In order to consolidate results, we need to combined col_df1 and col_df2 into a single data frame¶

In [42]:

col_wise_append = pd.concat([col_df1,col_df2['Sales_W2']],axis=1)
col_wise_append

Out[42]:

	Week	Product_Category	Sales_W1	Sales_W2
0	Week1	PC	25.436445	29.962953
1	Week1	Laptop	29.492091	27.875145
2	Week1	Services	28.286619	33.107039