How to Handle Exceptions With ast.literal_eval in a Pandas

John D K -

To handle exceptions and use ast.literal_eval in Pandas we can define new function:

import ast
import pandas as pd

def parse_eval(value):
	try:
    	return ast.literal_eval(value)
	except (ValueError, SyntaxError):
    	return value

df = pd.DataFrame({'col': ['1', '2', '{"a": 3}', '[[']})
df['col'].apply(parse_eval)

result:

0       	1
1       	2
2	{'a': 3}
3      	[[
Name: col, dtype: object

Otherwise error will be raised:

df['col'].apply(ast.literal_eval)

The error is:

SyntaxError: unexpected EOF while parsing

ast.literal_eval + exception

To use ast.literal_eval in Pandas we need the .apply() method. This will apply the ast.literal_eval function to each value of a column in a Pandas DataFrame.

To handle exceptions in Python we use the try-except block - to surround the ast.literal_eval call. Finally we return a default value in the except block in case of error.

In the example above, the parse_eval function takes a string value and:

  • returns parsed value from using ast.literal_eval - when possible
  • returns the original string - if an exception occurs during the call to ast.literal_eval

The apply method applies the parse_eval function to each element of the 'col' column in the DataFrame df.