You can use the following basic syntax to get the rows in one pandas DataFrame which are not in another DataFrame:

#merge two DataFrames and create indicator column
df_all = df1.merge(df2.drop_duplicates(), on=['col1','col2'],
                   how='left', indicator=True)

#create DataFrame with rows that exist in first DataFrame only
df1_only = df_all[df_all['_merge'] == 'left_only']

The following example shows how to use this syntax in practice.

Example: Get Rows in Pandas DataFrame Which Are Not in Another DataFrame

Suppose we have the following two pandas DataFrames:

import pandas as pd

#create first DataFrame
df1 = pd.DataFrame({'team' : ['A', 'B', 'C', 'D', 'E'], 
                    'points' : [12, 15, 22, 29, 24]}) 

print(df1)

  team  points
0    A      12
1    B      15
2    C      22
3    D      29
4    E      24

#create second DataFrame
df2 = pd.DataFrame({'team' : ['A', 'D', 'F', 'G', 'H'],
                    'points' : [12, 29, 15, 19, 10]})

print(df2)

  team  points
0    A      12
1    D      29
2    F      15
3    G      19
4    H      10

We can use the following syntax to merge the two DataFrames and create an indicator column to indicate which rows belong in each DataFrame:

#merge two DataFrames and create indicator column
df_all = df1.merge(df2.drop_duplicates(), on=['team','points'],
                   how='left', indicator=True)

#view result
print(df_all)

We can then use the following syntax to only get the rows in the first DataFrame that are not in the second DataFrame:

#create DataFrame with rows that exist in first DataFrame only
df1_only = df_all[df_all['_merge'] == 'left_only']

#view DataFrame
print(df1_only)

  team  points     _merge
1    B      15  left_only
2    C      22  left_only
4    E      24  left_only

Lastly, we can drop the _merge column if we’d like:

#drop '_merge' column
df1_only = df1_only.drop('_merge', axis=1)

#view DataFrame
print(df1_only)

  team  points
1    B      15
2    C      22
4    E      24

The result is a DataFrame in which all of the rows exist in the first DataFrame but not in the second DataFrame.

The following tutorials explain how to perform other common tasks in pandas: