Pandas Replace Blank Values (empty) with NaN

In pandas, you can replace blank values (empty strings) with NaN using the replace() method. In this article, I will explain the replacing blank values or empty strings with NaN in a pandas DataFrame and select columns by using either replace(), apply(), or mask() functions.

Related: You can also replace NaN values with blank/empty string.

Quick Examples of Replace Blank or Empty Values With NAN

Following are quick examples of replacing blank values or an empty string with NAN.

# Quick examples of replace blank or empty values with nAn

# Replace Blank values with DataFrame.replace() methods
df2 = df.replace(r'^\s*$', np.nan, regex=True)

# Using DataFrame.mask() method
df2=df.mask(df == '')

# Replace on single column
df2 = df.Courses.replace('',np.nan,regex = True)

# Replace on all selected columns
df2 = df[['Courses','Duration']].apply(lambda x: x.str.strip()).replace('', np.nan)

To run some examples of replacing blank values or an empty string with NAN, let’s create a pandas DataFrame.

# Create a Pandas DataFrame.
import pandas as pd
import numpy as np
technologies= {
    'Courses':["Spark","","Spark","","PySpark"],
    'Fee' :[22000,25000,23000,24000,26000],
    'Duration':['30days','','30days','','35days']
          }
df = pd.DataFrame(technologies)
print("Create DataFrame:\n", df)

Yields below output.

Pandas Replace Blank Values with NaN using replace()

You can replace blank/empty values with DataFrame.replace() methods. This method replaces the specified value with another specified value on a specified column or on all columns of a DataFrame; replaces every case of the specified value.

# Replace Blank values with DataFrame.replace() methods.
df2 = df.replace(r'^\s*$', np.nan, regex=True)
print("After replacing blank values with NaN:\n", df2)

Yields below output.

Pandas Replace Blank Values with NaN using mask()

You can also replace blank values with NAN with DataFrame.mask() methods. The mask() method replaces the values of the rows where the condition evaluates to True.

# Using DataFrame.mask() method.
df2=df.mask(df == '')
print("After replacing blank values with NaN:\n", df2)

Yields below output.

# Output:
# After replacing blank values with NaN:
   Courses    Fee Duration
0    Spark  22000   30days
1      NaN  25000      NaN
2    Spark  23000   30days
3      NaN  24000      NaN
4  PySpark  26000   35days

Pandas Replace Empty String with NaN on Single Column

Using replace() method you can also replace empty string or blank values to a NaN on a single selected column.

# Replace on single column
df2 = df.Courses.replace('',np.nan,regex = True)
print("After replacing blank values with NaN:\n", df2)

Yields below output

# Output:
# After replacing blank values with NaN:
0      Spark
1        NaN
2      Spark
3        NaN
4    PySpark
Name: Courses, dtype: object

Replace Blank Values with NAN by Using DataFrame.apply()

Another method to replace blank values with NAN is by using the DataFrame.apply() method along with lambda method. The apply() method enables the application of a function along one of the DataFrame’s axes, with the default being 0, representing the index (row) axis.

In order to use this, you need to have all columns as String type. If you have any non-string column this gives an error. Since I have a non-string column, I have selected only string columns and used the apply function.

# Replace on all selected columns
df2 = df[['Courses','Duration']].apply(lambda x: x.str.strip()).replace('', np.nan)
print("After replacing blank values with NaN:\n", df2)

Yields below output

# Output:
# After replacing blank values with NaN:
   Courses Duration
0    Spark   30days
1      NaN      NaN
2    Spark   30days
3      NaN      NaN
4  PySpark   35days

Complete Example of Replace Blank values (Empty String) with NaN

# Create a Pandas DataFrame.
import pandas as pd
import numpy as np
technologies= {
    'Courses':["Spark","","Spark","","PySpark"],
    'Fee' :[22000,25000,23000,24000,26000],
    'Duration':['30days','','30days','','35days']
          }
df = pd.DataFrame(technologies)
print(df)


# Replace Blank values with DataFrame.replace() methods.
df2 = df.replace(r'^\s*$', np.nan, regex=True)
print(df2)

# Using DataFrame.mask() method.
df2=df.mask(df == '')
print(df2)

# Replace on single column
df2 = df.Courses.replace('',np.nan,regex = True)
print(df2)

# Replace on all selected columns
df2 = df[['Courses','Duration']].apply(lambda x: x.str.strip()).replace('', np.nan)
print(df2)

Conclusion

In this article, I have explained the replacement blank values with NAN of pandas DataFrame by using replace(), apply(), mask() methods with the examples.

Related Articles

• Rename Index of Pandas DataFrame
• Pandas Drop Columns with NaN or None Values
• Pandas Drop Rows with NaN Values in DataFram
• Remove NaN From Pandas Series
• Pandas – Replace NaN Values with Zero in a Column
• Count NaN Values in Pandas DataFrame
• Check Any Value is NaN in DataFrame
• How to Replace String in pandas DataFrame
• Pandas DataFrame.fillna() function explained
  Pandas Series.fillna() function explained

References

• https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.replace.html