How to Convert Strings to Floats in Pandas DataFrame?

In this article, we’ll look at different ways in which we can convert a string to a float in a pandas dataframe. Now, let’s create a Dataframe with ‘Year’ and ‘Inflation Rate’ as a column.

Python3

# importing pandas library
import pandas as pd

# dictionary
Data = {'Year': ['2016', '2017',
                 '2018', '2019'],
        'Inflation Rate': ['4.47', '5',
                           '5.98', '4.1']}
# create a dataframe
df = pd.DataFrame(Data)

# show the dataframe
print (df)

# show the datatypes
print(df.dtypes)

Output:

Method 1: Using DataFrame.astype().

The method is used to cast a pandas object to a specified dtype. 

Syntax: DataFrame.astype(self: ~ FrameOrSeries, dtype, copy: bool = True, errors: str = ‘raise’)

Returns: casted: type of caller

Example: In this example, we’ll convert each value of ‘Inflation Rate’ column to float.

Python3

# importing pandas library
import pandas as pd

# dictionary
Data = {'Year': ['2016', '2017',
                 '2018', '2019'],
        'Inflation Rate': ['4.47', '5',
                           '5.98', '4.1']}
# create a dataframe
df = pd.DataFrame(Data)

# converting each value
# of column to a string
df['Inflation Rate'] = df['Inflation Rate'].astype(float)

# show the dataframe
print(df)

# show the datatypes
print (df.dtypes)

Output:

Method 2: Using pandas.to_numeric() function.

The function is used to convert the argument to a numeric type.

Syntax: pandas.to_numeric(arg, errors=’raise’, downcast=None)

Returns: numeric if parsing succeeded. Note that the return type depends on the input. Series if Series, otherwise ndarray. 

Example 1: In this example, we’ll convert each value of ‘Inflation Rate’ column to float.

Code:

Python3

# importing pandas library
import pandas as pd

# creating a dictionary
Data = {'Year': ['2016', '2017',
                 '2018', '2019'],
          'Inflation Rate': ['4.47', '5',
                             '5.98', '4.1']}
# create a dataframe
df = pd.DataFrame(Data)

# converting each value of column to a string
df['Inflation Rate'] = pd.to_numeric(df['Inflation Rate'])

# show the dataframe
print(df)

# show the data types
print (df.dtypes)

Output:

Example 2: Sometimes, we may not have a float value represented as a string. So, pd.to_numeric() function will show an error. To remove this error, we can use errors=’coerce’, to convert the value at this position to be converted to NaN. 

Code: 

Python3

# importing pandas as pd
import pandas as pd

# dictionary
Data = {'Year': ['2016', '2017',
                 '2018', '2019'],
         'Inflation Rate': ['4.47', '5',
                           'No data', '4.1']}

# create a dataframe
df = pd.DataFrame(Data)

# converting each value of column to a string
df['Inflation Rate'] = pd.to_numeric(df['Inflation Rate'],
                                     errors = 'coerce')

# show the dataframe
print(df)

# show the data types
print (df.dtypes)

Output:

Note: String data type shows as an object.