How to change a DataFrame column type to float, int and string with Python?

by

In today’s short tutorial we’ll learn how to easily convert DataFrame columns to different types.

Let’s start by defining a very simple Data Frame made from a list of lists. You might want follow along by running the code in your Jupyter Notebook.

import pandas as pd

employee = ['Eric', 'Tom', 'John', 'Dean']
sales = [140.3, 140.2, 180.2, 167.1]
salary = [87, 92, 76, 98]

'create the dataframe

revenue = pd.DataFrame(dict(emp=employee, sales=sales, sal=salary))

Let’s look into the column data types:

revenue.dtypes

We’ll receive the following output:

emp       object
sales    float64
sal        int64
dtype: object

Change column to integer

We’ll start by using the astype method to convert a column to the int data type. Run the following code:

# convert to int
revenue['sales'].astype('int')

Change column to float in Pandas

Next example is to set the column type to float. 

revenue['sal'].astype('float')

Convert column to string type

Third example is the conversion to string. In a nutshell, you can accomplish that by using the following very simple snippet below:

revenue['emp'].astype('string')

Converting multiple columns to float, int and string

You can easily change the type for multiple columns, simply by passing a dictionary with the corresponding column index and target type to the astype method. We’ll persist the changes to the column types by assigning the result into a new DataFrame.

# putting everything together

revenue_2 = revenue.astype({'emp':'string', 'sales':'int', 'sal':'float'})

Now we can easily check the dtypes:

revenue_2.dtypes

And the result will be as expected:

emp       string
sales      int32
sal      float64
dtype: object