Pandas – Replace Values in a DataFrame

When working with pandas dataframes, it might be handy to know how to quickly replace values. In this tutorial, we’ll look at how to replace values in a pandas dataframe through some examples.

The pandas dataframe replace() function is used to replace values in a pandas dataframe. It allows you the flexibility to replace a single value, multiple values, or even use regular expressions for regex substitutions. The following is its syntax:

df_rep = df.replace(to_replace, value)

Here, to_replace is the value or values to be replaced and value is the value to replace with. By default, the pandas dataframe replace() function returns a copy of the dataframe with the values replaced. If you want to replace the values in-place pass inplace=True

Let’s look at some of the different use-cases of the replace() function through some examples.

The replace() function replaces all occurrences of the value with the desired value. See the example below:

import pandas as pd

# sample dataframe
df = pd.DataFrame({'A': ['a','b','c'], 'B':['b','c','d']})
print("Original DataFrame:\n", df)

# replace b with e
df_rep = df.replace('b', 'e')
print("\nAfter replacing:\n", df_rep)

Output:

Original DataFrame:
    A  B
0  a  b
1  b  c
2  c  d

After replacing:
    A  B
0  a  e
1  e  c
2  c  d

In the above example, the replace() function is used to replace all the occurrences of b in the dataframe with e.

The pandas dataframe replace() function allows you the flexibility to replace values in specific columns without affecting values in other columns.

import pandas as pd

# sample dataframe
df = pd.DataFrame({'A': ['a','b','c'], 'B':['b','c','d']})
print("Original DataFrame:\n", df)

# replace b with e
df_rep = df.replace({'A': 'b'}, 'e')

print("\nAfter replacing:\n", df_rep)

Output:

Original DataFrame:
    A  B
0  a  b
1  b  c
2  c  d

After replacing:
    A  B
0  a  b
1  e  c
2  c  d

In the above example, we replace the occurrences of b in just the column A of the dataframe. For this, we pass a dictionary to the to_replace parameter. Here the dictionary {'A': 'b'} tells the replace function that we want to replace the value b in the column A. And the 'e' passed to the value parameter use is used to replace all relevant matches.

You can also have multiple replacements together. For example, if you want to replace a with b, b with c and c with d in the above dataframe, you can pass just a single dictionary to the replace function.

import pandas as pd

# sample dataframe
df = pd.DataFrame({'A': ['a','b','c'], 'B':['b','c','d']})
print("Original DataFrame:\n", df)

# replace a with b, b with c, and c with d
df_rep = df.replace({'a':'b', 'b':'c', 'c':'d'})

print("\nAfter replacing:\n", df_rep)

Output:

Original DataFrame:
    A  B
0  a  b
1  b  c
2  c  d

After replacing:
    A  B
0  b  c
1  c  d
2  d  d

In the above example, we only pass a single dictionary to the replace() function. The function infers the dictionary keys as values to replace in the dataframe and the corresponding keys as the values to replace them with.

To replace values within a dataframe via a regular expression match, pass regex=True to the replace function. Keep in mind that you pass the regular expression string to the to_replace parameter and the value to replace the matches with to the value parameter. Also, note that regular expressions will only substitute on strings

import pandas as pd

# sample dataframe
df = pd.DataFrame({'A': ['tap','cap','map'], 'B':['cap','map', 'tap']})
print("Original DataFrame:\n", df)

# replace ap with op
df_rep = df.replace(to_replace='ap', value='op', regex=True)

print("\nAfter replacing:\n", df_rep)

Output:

Original DataFrame:
      A    B
0  tap  cap
1  cap  map
2  map  tap

After replacing:
      A    B
0  top  cop
1  cop  mop
2  mop  top

In the above example, the regular expression matches for the occurrences of ap and replaces them with op.

For more on the pandas dataframe replace function, refer to its official documentation.

With this, we come to the end of this tutorial. The code examples and results presented in this tutorial have been implemented in a Jupyter Notebook with a python (version 3.8.3) kernel having pandas version 1.0.5


Subscribe to our newsletter for more informative guides and tutorials.
We do not spam and you can opt out any time.