JIYIK CN >

In Pandas, two DataFrames Seriescan be easily joined or combined DataFrameusing various operations such as join and joinmerge merge. These operations combine two DataFrames based on the index and column names. Both join joinand mergemerge methods can combine two DataFrames. The main difference between join and merge operations is that jointhe join method combines two DataFrames based on their indices, whereas in mergethe join method, we need to specify the columns to combine the two DataFrames.

This article will discuss the difference between joinand mergemethods in pandas python.

Pandas DataFrame .joinMethods

joinThe method joins two dataframes on their indices DataFrame. Let's take an example to show joinhow the method works. We took two DataFrames: left_dfand right_df. Using left_df.join(right_df)the code, we combined the two DataFrames.

Sample code:

import pandas as pd

# create two dataframe
df_left = pd.DataFrame({"Name": ["X", "Y", "Z"], "Score": [10, 8, 9]}).set_index("Name")
df_right = pd.DataFrame({"Name": ["X", "Y", "Z"], "Steals": [4, 5, 2]}).set_index(
    "Name"
)
print(df_left)
print(df_right)

# join two dataframes
df_left.join(df_right)

Output:

      Score
Name       
X        10
Y         8
Z         9
      Steals
Name        
X          4
Y          5
Z          2
Score	Steals
Name		
X	10	4
Y	8	5
Z	9	2

If we have overlapping columns in both the DataFrames, in that case, join will expect you to add suffix to the overlapping or common column names from the left DataFrame. In the following DataFrames, the overlapping column names are C.

Sample code:

import pandas as pd

# Creating the two dataframes
df_left = pd.DataFrame([["x", 1], ["y", 2]], list("AB"), list("CD"))
df_right = pd.DataFrame([["u", 3], ["v", 4]], list("AB"), list("CF"))
print(df_left)
print(df_right)
# join two dataframes
joined_df = df_left.join(df_right, lsuffix="_")
print(joined_df)

Output:

   C  D
A  x  1
B  y  2
   C  F
A  u  3
B  v  4
  C_  D  C  F
A  x  1  u  3
B  y  2  v  4

As you can see in the above output, the index retains four columns. We can also single out specific columns on the left DataFrame by using the on parameter as the join key.

Pandas DataFrame .mergeMethods

mergeThe merge method is also used to merge two DataFrames. However, the merge method requires column names as the merge key to merge the two DataFrames. In the following example, we have implemented a simple merge function to merge two DataFrames without using any parameters.

Sample code:

import pandas as pd

# create two dataframe
df_left = pd.DataFrame({"Name": ["X", "Y", "Z"], "Score": [10, 8, 9]}).set_index("Name")
df_right = pd.DataFrame({"Name": ["X", "Y", "Z"], "Steals": [4, 5, 2]}).set_index(
    "Name"
)
print(df_left)
print(df_right)

# merge two dataframes
df_left.merge(df_right, on="Name")

Output:

      Score
Name       
X        10
Y         8
Z         9
      Steals
Name        
X          4
Y          5
Z          2
Score	Steals
Name		
X	10	4
Y	8	5
Z	9	2

We can specify overlapping column names mergeusing onthe parameter in the merge method. In the following example, we specify the overlapping column names Cto perform a merge operation on two DataFrames.

Sample code:

import pandas as pd

# Creating the two dataframes
df_left = pd.DataFrame([["x", 1], ["y", 2]], list("AB"), list("CD"))
df_right = pd.DataFrame([["u", 3], ["v", 4]], list("AB"), list("CF"))
print(df_left)
print(df_right)

# merge dataframes
merged_df = df_left.merge(df_right, on="C", how="outer")
print(merged_df)

Output:

   C  D
A  x  1
B  y  2
   C  F
A  u  3
B  v  4
   C    D    F
0  x  1.0  NaN
1  y  2.0  NaN
2  u  NaN  3.0
3  v  NaN  4.0

It is specified using right_onthe and left_onparameters respectively. See the following example where we have used different parameters like on, left_on, right_onfor better understanding.

Sample code:

import pandas as pd

# Creating the two dataframes
df_left = pd.DataFrame([["x", 1], ["y", 2]], list("AB"), list("CD"))
df_right = pd.DataFrame([["u", 3], ["v", 4]], list("AB"), list("CF"))
print(df_left)
print(df_right)
merged_df = df_left.merge(
    df_right, left_index=True, right_index=True, suffixes=["_", ""]
)
print(merged_df)

Output:

   C  D
A  x  1
B  y  2
   C  F
A  u  3
B  v  4
  C_  D  C  F
A  x  1  u  3
B  y  2  v  4

in conclusion

We have demonstrated the difference between joinand in pandas with some examples merge. We have seen these two methods, joinand mergeare used for similar purpose, to combine DataFrames in pandas. However, the difference is that the method combines two DataFrames on jointheir , whereas in the method, we specify the column names to combine two DataFrames.indexedmerge