Table of Contents
In Python, the equivalent of the R cbind function is the NumPy hstack function. This function takes two or more NumPy arrays as its arguments and combines them horizontally (in columns) to form a single 2D array. It is important to note that all of the arrays must have the same number of rows. This function is useful for combining data together into a single array for further analysis.
The cbind function in R, short for column-bind, can be used to combine data frames together by their columns.
We can use the function from pandas to perform the equivalent function in Python:
df3 = pd.concat([df1, df2], axis=1)
The following examples shows how to use this function in practice.
Example 1: Use cbind in Python with Equal Index Values
Suppose we have the following two pandas DataFrames:
import pandas as pd #define DataFrames df1 = pd.DataFrame({'team': ['A', 'B', 'C', 'D', 'E'], 'points': [99, 91, 104, 88, 108]}) print(df1) team points 0 A 99 1 B 91 2 C 104 3 D 88 4 E 108 df2 = pd.DataFrame({'assists': ['A', 'B', 'C', 'D', 'E'], 'rebounds': [22, 19, 25, 33, 29]}) print(df2) assists rebounds 0 A 22 1 B 19 2 C 25 3 D 33 4 E 29
We can use the concat() function to quickly bind these two DataFrames together by their columns:
#column-bind two DataFrames into new DataFrame
df3 = pd.concat([df1, df2], axis=1)
#view resulting DataFrame
df3
team points assists rebounds
0 A 99 A 22
1 B 91 B 19
2 C 104 C 25
3 D 88 D 33
4 E 108 E 29
Example 2: Use cbind in Python with Unequal Index Values
Suppose we have the following two pandas DataFrames:
import pandas as pd #define DataFrames df1 = pd.DataFrame({'team': ['A', 'B', 'C', 'D', 'E'], 'points': [99, 91, 104, 88, 108]}) print(df1) team points 0 A 99 1 B 91 2 C 104 3 D 88 4 E 108 df2 = pd.DataFrame({'assists': ['A', 'B', 'C', 'D', 'E'], 'rebounds': [22, 19, 25, 33, 29]}) df2.index = [6, 7, 8, 9, 10] print(df2) assists rebounds 6 A 22 7 B 19 8 C 25 9 D 33 10 E 29
Notice that the two DataFrames do not have the same index values.
If we attempt to use the concat() function to cbind them together, we’ll get the following result:
#attempt to column-bind two DataFrames
df3 = pd.concat([df1, df2], axis=1)
#view resulting DataFrame
df3
team points assists rebounds
0 A 99.0 NaN NaN
1 B 91.0 NaN NaN
2 C 104.0 NaN NaN
3 D 88.0 NaN NaN
4 E 108.0 NaN NaN
6 NaN NaN A 22.0
7 NaN NaN B 19.0
8 NaN NaN C 25.0
9 NaN NaN D 33.0
10 NaN NaN E 29.0
This is not the result we wanted.
To fix this, we need to first reset the index of each DataFrame before concatenating them together:
import pandas as pd #define DataFrames df1 = pd.DataFrame({'team': ['A', 'B', 'C', 'D', 'E'], 'points': [99, 91, 104, 88, 108]}) df2 = pd.DataFrame({'assists': ['A', 'B', 'C', 'D', 'E'], 'rebounds': [22, 19, 25, 33, 29]}) df2.index = [6, 7, 8, 9, 10] #reset index of each DataFrame df1.reset_index(drop=True, inplace=True) df2.reset_index(drop=True, inplace=True) #column-bind two DataFrames df3 = pd.concat([df1, df2], axis=1) #view resulting DataFrame df3 team points assists rebounds 0 A 99 A 22 1 B 91 B 19 2 C 104 C 25 3 D 88 D 33 4 E 108 E 29
Notice that this DataFrame matches the one we got in the previous example.
The following tutorials explain how to perform other common operations in Python: