Pandas DataFrame: Convert all numeric strings to numbers

Updated: February 20, 2024 By: Guest Contributor Post a comment


Handling data often involves dealing with various types of formats and ensuring consistency among your dataset’s types is crucial for analysis and machine learning models. In this tutorial, we will explore how to convert all numeric strings to numbers within a Pandas DataFrame. This conversion is vital when your dataset contains numeric values as strings and you need to perform arithmetic operations or analyses. We will start from basic methods and gradually proceed to more advanced techniques.


Before we get started, make sure you have installed Pandas in your environment:

pip install pandas

And import Pandas in your Python script:

import pandas as pd

Basic Conversion with pd.to_numeric()

The simplest way to convert string to numbers is using the pd.to_numeric() function. Here’s a basic example:

df = pd.DataFrame({'A':['1', '2', '3'], 'B':['4.0', '5.6', '7.1']})
df = df.apply(pd.to_numeric)

The output will be:

   A    B
0  1  4.0
1  2  5.6
2  3  7.1

Error Handling in pd.to_numeric()

The to_numeric method can deal with errors in several ways. By default, it raises an error if it cannot convert a string to a number. However, you can specify how to handle errors using the errors parameter:

  • coerce will convert the unconvertable strings to NaN.
  • ignore will ignore the conversion and keep the string as is.
  • raise (default) will raise an exception if conversion fails.
df['C'] = ['invalid', '9.5', '3']
df = df.apply(lambda x: pd.to_numeric(x, errors='coerce'))

The output will be:

   A    B    C
0  1  4.0  NaN
1  2  5.6  9.5
2  3  7.1  3.0

Advanced Conversion Using applymap()

For a more refined control over conversion, especially when dealing with mixed-type DataFrames, one can use the applymap function. This function applies a specified function to each element of the DataFrame,

def convert_numeric(x):
        return pd.to_numeric(x)
    except ValueError:
        return x


Dealing with Large and Complex DataFrames

When working with large DataFrames, it may be more efficient to selectively convert columns that you know contain numeric strings. This can be done using the astype function for columns that don’t contain any string that cannot be converted to numbers.

df['D'] = df['D'].astype(float)

Optimizations for Performance

For large-scale data, performance can become an issue. Utilizing vectorized operations provided by Pandas and NumPy can result in significant performance improvements. A combination of apply, applymap, and datatype conversion using astype or to_numeric with appropriate error handling can provide optimal results.


Converting numeric strings to numbers in a Pandas DataFrame is a common data preprocessing step. Starting with basic conversion using pd.to_numeric() for whole dataframes, handling errors appropriately, and advancing to selective conversion using applymap() or astype for improving performance, this tutorial has covered essential methods for dealing with numeric strings. Implementing these techniques effectively can lead to more accurate data analysis and model performance.