‘Dataframe’ object has no attribute ‘as_matrix’

If you are working with pandas dataframes in Python, you may encounter an error that says the ‘dataframe’ object has no attribute ‘as_matrix’. This error occurs when you try to use the as_matrix() method on a dataframe object, which was deprecated in pandas version 0.23.0 and removed in pandas version 1.0.0.

The as_matrix() method was used to convert a dataframe into a numpy array, which is useful for various operations such as linear algebra, machine learning, etc.

What causes the ‘dataframe’ object to have no attribute ‘as_matrix’ error?

The main cause of the ‘dataframe’ object has no attribute ‘as_matrix’ error is that you are using an outdated version of pandas that still has the as_matrix() method or you are following an outdated tutorial or code snippet that uses the as_matrix() method.

The as_matrix() method was deprecated in favor of the values attribute or the to_numpy() method, which are more consistent and flexible ways to convert a dataframe into a numpy array.

For example, if you have a dataframe like this:

import pandas as pd
df = pd.DataFrame({'name': ['Alice', 'Bob', 'Charlie'], 'age': [25, 30, 35]})
print(df)

Output:

     name  age
0   Alice   25
1     Bob   30
2  Charlie   35   

If you try to use the as_matrix() method on it, you will get the error:

import pandas as pd
df = pd.DataFrame({'name': ['Alice', 'Bob', 'Charlie'], 'age': [25, 30, 35]})
df.as_matrix()
print(df)

Output:

AttributeError: 'DataFrame' object has no attribute 'as_matrix.'

How do we resolve the ‘dataframe’ object has no attribute ‘as_matrix’ error?

The easiest way to resolve the ‘dataframe’ object has no attribute ‘as_matrix’ error is to replace the as_matrix() method with the values attribute or the to_numpy() method, depending on your preference and use case. These options will return a numpy array with the same shape and data type as the original dataframe.

Using the values attribute

The values attribute is a simple and fast way to access the underlying numpy array of a dataframe. It does not require any parentheses or arguments, and it preserves the data type of the dataframe columns. However, it may not always return a copy of the data, which means modifying the array may also modify the dataframe.

For example, you can use the values attribute like this:

import pandas as pd
df = pd.DataFrame({'name': ['Alice', 'Bob', 'Charlie'], 'age': [25, 30, 35]})
print(df.values)

Output:

[['Alice' 25]
 ['Bob' 30]
 ['Charlie' 35]]

Using the to_numpy() method

The to_numpy() method is a more explicit and flexible to convert a dataframe into a numpy array. It requires parentheses and accepts optional arguments such as dtype, copy, and na_value to control the data type, copying behavior, and missing value handling of the resulting array. It always returns a copy of the data, meaning modifying the array will not affect the dataframe.

For example, you can use the to_numpy() method like this:

import pandas as pd
df = pd.DataFrame({'name': ['Alice', 'Bob', 'Charlie'], 'age': [25, 30, 35]})
print(df.to_numpy())

Output:

[['Alice' 25]
 ['Bob' 30]
 ['Charlie' 35]]

FAQs

Here are some frequently asked questions related to the ‘dataframe’ object has no attribute ‘as_matrix’ error:

What is the difference between the values attribute and the to_numpy() method?

The main difference is that the values attribute is simpler and faster but may not always return a copy of the data. At the same time, the to_numpy() method is more explicit and flexible, but it always returns a copy of the data. The to_numpy() method also allows you to specify the data type, copying behavior, and missing value handling of the resulting array, while the values attribute does not.

How can I check the version of Pandas that I am using?

You can check the version of pandas that you are using by importing pandas and printing its __version__ attribute. For example:

import pandas as pd
print(pd.__version__)

How can I update pandas to the latest version?

Depending on how you installed pandas, you can update pandas to the latest version by using the pip or conda commands. For example, if you are using pip, you can run:

pip install --upgrade pandas

Or, for conda, you can run:

conda update pandas

Conclusion

In this article, we learned what causes the ‘dataframe’ object has no attribute ‘as_matrix’ error in Python and how to resolve it using the values attribute or the to_numpy() method. We have also answered some common questions related to this error. We hope this article has helped you understand the error and help you fix this error in your pandas code. 

References

1. as_matrix
2. to_numpy()

To learn more about fixes for common mistakes when writing in Python, follow us at Python Clear.