Adding Data to Empty DataFrames in pandas

One of the most popular 3rd party Python packages is called pandas. The pandas package “is a fast, powerful, flexible and easy to use open source data analysis and manipulation tool, built on top of the Python programming language.” It is used by data scientists and software engineers around the world.

In this tutorial, you will learn a little about creating different kinds of empty or partially empty DataFrames with pandas. Then you will learn a couple of different ways to add data to that DataFrame.

Specifically, you will learn about the following:

Creating an empty DataFrame and adding data
Creating an empty DataFrame with columns and adding data
Creating an empty DataFrame with columns and indices and adding data

Creating an Empty DataFrame in pandas

Sometimes you just need to create an empty DataFrame, much like you sometimes need to create an empty Python dictionary or list.

Here is an example of creating a completely empty DataFrame with pandas:

>>> import pandas as pd
>>> df = pd.DataFrame()
>>> df
Empty DataFrame
Columns: []
Index: []

Of course, an empty DataFrame isn’t especially useful. So let’s add a little data to the DataFrame!

>>> import pandas as pd
>>> df = pd.DataFrame()
>>> df
Empty DataFrame
Columns: []
Index: []

>>> df["Name"] = ["Mike", "Steve", "Rodrigo"]
>>> df["Jobs"] = ["Engineer", "Core Dev", "Content Creator"]
>>> df
      Name             Jobs
0     Mike         Engineer
1    Steve         Core Dev
2  Rodrigo  Content Creator

This example demonstrates how you can specify columns in pandas and add data to those columns.

Now let’s learn how to create an empty DataFrame that includes columns, but no data!

Creating an Empty DataFrame with Columns

This next example will show you how to create a pandas DataFrame that includes columns, but no indices or column data.

Let’s take a look:

>>> import pandas as pd

>>> df = pd.DataFrame(columns=["Name", "Job"])
>>> df
Empty DataFrame
Columns: [Name, Job]
Index: []

# Add some data using append()
>>> df = df.append({"Name": "Mike", "Job": "Blogger"}, ignore_index=True)
>>> df
   Name      Job
0  Mike  Blogger
>>> df = df.append({"Name": "Luciano", "Job": "Author"}, ignore_index=True)
>>> df
      Name      Job
0     Mike  Blogger
1  Luciano   Author

Well, that was better than a completely empty DataFrame! In this example, you also learn how to use the DataFrame’s append() method to add data to each column.

When you use append() it takes in a dictionary of column name and value. You also set ignore_index to True, which lets pandas update the index for you automatically.

Now let’s look at how to create one more type of empty DataFrame with pandas!

Creating an Empty DataFrame with Columns and Indices

For this example, you will learn how to create a pandas DataFrame that has two columns and three named rows or indices.

Here’s how it is done:

>>> import pandas as pd
>>> df = pd.DataFrame(columns = ["Name", "Job"], index = ["a", "b", "c"])
>>> df
  Name  Job
a  NaN  NaN
b  NaN  NaN
c  NaN  NaN

When you print out the DataFrame, you can see that the columns all contain NaN, which stands for “Not a Number”. A NaN is kind of like a None in Python.

One way to add data to this DataFrame in pandas is to use the loc attribute:

>>> df.loc["a"] = ["Mike", "Engineer"]
>>> df
   Name       Job
a  Mike  Engineer
b   NaN       NaN
c   NaN       NaN
>>> df.loc["b"] = ["Steve", "Core Dev"]
>>> df
    Name       Job
a   Mike  Engineer
b  Steve  Core Dev
c    NaN       NaN
>>> df.loc["c"] = ["Rodrigo", "Content Creator"]
>>> df
      Name              Job
a     Mike         Engineer
b    Steve         Core Dev
c  Rodrigo  Content Creator

When you use the loc attribute, you use dictionary-like syntax to set a specific index to a list of values. The above example shows how to add three rows of data.

Wrapping Up

This tutorial doesn’t even begin to scratch the surface of what you can do with pandas. But it isn’t supposed to. That’s what books are for. Here you learned how to create three different types of empty DataFrames.

To be specific, you learned the following: