How to Create Empty Dataframe in Pandas And Add Rows

Pandas Dataframe is a two-dimensional data structure that can be used to store the data in rows and columns format. Dataframes are very useful in data science and machine learning use cases.

You can create an empty dataframe in pandas using the pd.DataFrame() method.

In this tutorial, you’ll learn how to create an empty dataframe in Pandas.

If you’re in Hurry

You can use the below code snippet to create an empty dataframe in pandas

import pandas as pd

# create an empty dataframe
df  = pd.DataFrame()

df

Dataframe Looks Like

Empty DataFrame
Columns: []
Index: []

If You Want to Understand Details, Read on…

In this tutorial, you’ll learn the different methods available to create an empty dataframe in pandas and additional options available while creating an empty dataframe. Read on…

Create Empty Dataframe

First, you’ll learn how to just create an empty dataframe using the Dataframe() class available in the pandas library. It is similar to a constructor which can be used to construct the class.

Snippet

# import pandas library
import pandas as pd

# create an empty dataframe
df  = pd.DataFrame()

df

Empty Dataframe Looks Like

Empty DataFrame
Columns: []
Index: []

The Dataframe() class supports the below parameters. All the parameters are optional. If you dint pass any parameter, then a simple empty dataframe object will be created.

• data – Used to pass the inital values to the dataframe
• index – Used to create index in the resulting dataframe
• columns – Column labels to be used in the resulting dataframe
• dtypes – Used to mention the datatypes for thenewly created columns in the dataframe
• copy – Used to mention if the data should be copied from the inputs. By default, its False.

This is how you can create an empty dataframe.

Next, you’ll learn about creating a dataframe with just column names.

Create Empty Dataframe With column names

In this section, you’ll learn how to create an empty dataframe with column names.

You can define the column names as a list and pass the list to the parameter columns while calling the DataFrame() as shown below.

column_names =  ['Column_1', 'Column_2', 'Column_3']

df  = pd.DataFrame(columns = column_names)

df

An empty dataframe will be created with headers as shown below.

Dataframe Looks Like

	Column_1	Column_2	Column_3

This is how you can create an empty dataframe with the defined column names as headers.

Next, you’ll create an empty dataframe with dtypes.

Create Empty Dataframe With column names And Datatypes

In this section, you’ll learn how to create an empty dataframe with column names and data types defined for each column.

You’ll need to create an empty pandas series for each column and specify the data type for that column using the dtype parameter.

Creating a series

pd.Series([], dtype='int')

You can create a number of series with the different data types available in python. You can assign the series to each column while creating the dataframe as shown below.

Snippet

You can use the below snippet to create an empty dataframe with column headers and data types defined for it.

df = pd.DataFrame({'Column_1': pd.Series([], dtype='int'),
                   'Column_2': pd.Series([], dtype='str'),
                   'Column_3': pd.Series([], dtype='float')})

df.dtypes

When you print the dataframe column types using the df.dtypes, you’ll see the below output.

Output

    Column_1      int32
    Column_2     object
    Column_3    float64
    dtype: object

This is how you can create an empty dataframe with column headers and data types defined for each column.

Next, you’ll learn how to create an empty dataframe with size.

Create Empty Dataframe With Size

In this section, you’ll learn how to create an empty dataframe with size.

You can create a dataframe with a specified size for both columns and rows.

Use the range function to create a sequence of numbers and pass it to the index range or the columns range specify column and row sizes.

To specify the size of the rows, you can use the index parameter with range(). For example, index=range(no_of_Rows)

To specify the size of the columns, you can use the columns parameter with range(). For example, columns=range(no_of_Cols)

Snippet

Use the below snippet to create an empty dataframe with 2 rows and 5 columns.

no_of_Rows = 2

no_of_Cols = 5

df = pd.DataFrame(index=range(no_of_Rows),columns=range(no_of_Cols))

df

You’ll see the empty dataframe created with 2 rows and 5 columns and all the cells will have the value NaN which means the missing data.

Dataframe Looks Like

	0	1	2	3	4
0	NaN	NaN	NaN	NaN	NaN
1	NaN	NaN	NaN	NaN	NaN

To create an empty Dataframe only with a specified number of rows, use the below snippet.

nRows= 2

df = pd.DataFrame(index=range(nRows))

df

Dataframe Looks Like

0
1

To create a dataframe with only a specified number of columns, use the below snippet.

nCols = 5

df = pd.DataFrame(columns=range(nCols))

df

Dataframe Looks Like

	0	1	2	3	4

This is how you can create an empty dataframe with size.

Next, you’ll learn about appending columns to empty dataframe.

Create Empty Dataframe and Append Columns

In this section, you’ll learn how to create an empty dataframe and append columns to the empty dataframe.

First, create an empty dataframe using pd.Dataframe().

Next, you can append a column to the created dataframe using the insert() method. To know more about other methods available to add columns to the dataframe, refer to add column to dataframe tutorial.

Dataframe’s Insert() method accepts the following parameters.

• loc – Index position when the new column to be inserted
• column – Name of the new column to be appended
• values – List of values for the new column. It must be int, series or an array.
• allow_duplicates – To mention if the duplicate column names are allowed. By default, its False. If there is a column already available in the dataframe with the same name, then an error will be raised. If this parameter is True, then error will not be raised and a duplicate column will be created.

Snippet

Use the below code to append a column at the 0 th position of the dataframe.

df  = pd.DataFrame()


# Using DataFrame.insert() to add a column
df.insert(0, "Column_1", [5,10,10,5,10], True)

df

Where,

• 0– Index position
• Column_1 – Name for the new column
• [5,10,10,5,10] – List of values to pass to the dataframe
• True – To allow the duplicate column headers.

Column_1 will be inserted into the dataframe as shown below.

Dataframe Looks Like

	Column_1
0	5
1	10
2	10
3	5
4	10

To append multiple columns to the empty dataframe, then you can use the below code.

Snippet

df['Column_2'], df['Column_3'] = [pd.NaT, 3]

df

Then column_2 and column_3 will be inserted into the dataframe.

Dataframe Looks Like

	Column_1	Column_2	Column_3
0	5	NaT	3
1	10	NaT	3
2	10	NaT	3
3	5	NaT	3
4	10	NaT	3

This is how you can create an empty dataframe and add columns to it.

Next, you’ll learn about adding rows.

Create Empty Dataframe and Append Rows

In this section, you’ll learn how to create an empty dataframe and append rows to it.

First, create an empty dataframe using pd.DataFrame() and with the headers by using the columns parameter.

Next, append rows to it by using a dictionary. Each row needs to be created as a dictionary.

Dictionary’s key should be the column name and the Value should be the value of the cell. Create a dictionary with values for all the columns available in the dataframe and use the append() method to append the dictionary as a row.

For Example, a dictionary for each row should look like {'Name' : 'CPU', 'Quantity' : 5, 'Price' : 20000} for the dataframe with columns Name Quantity and Price.

df = pd.DataFrame(columns = ['Name', 'Quantity', 'Price'])

print(df)

# append rows to an empty DataFrame
df = df.append({'Name' : 'CPU', 'Quantity' : 5, 'Price' : 20000}, 
                ignore_index = True)
df = df.append({'Name' : 'Monitor', 'Quantity' : 10, 'Price' : 10000},
                ignore_index = True)
df = df.append({'Name' : 'Keyboard', 'Quantity' : 10, 'Price' : 550},
               ignore_index = True)

df

Where

• df.append() method invokes the append method on the dataframe.
• {'Name' : 'CPU', 'Quantity' : 5, 'Price' : 20000} – Dictionary with values for each column
• ignore_index = True – To label the index columns as 0 or 1 or n. Other words, it means, the dictionary doesn’t contain values for the index columns. So the default index value will be used.

Output

    Empty DataFrame
    Columns: [Name, Quantity, Price]
    Index: []

Dataframe Looks Like

	Name	Quantity	Price
0	CPU	5	20000
1	Monitor	10	10000
2	Keyboard	10	550

This is how you can create an empty dataframe and append rows to it.

Next, you’ll learn about creating a dataframe from another dataframe.

Create Empty Dataframe from Another Dataframe

In this section, you’ll create an empty dataframe from another dataframe which is already existing.

For example, assume the existing data df with the following columns and data.

Dataframe Looks Like

	Name	Quantity	Price	Column_2	Column_3
0	CPU	5	20000	NaT	3
1	Monitor	10	10000	NaT	3
2	Keyboard	10	550	NaT	3

Now, you’ll create a dataframe df2 using the dataframe df and its column but without copying the data.

First, you need to get the list of columns from the dataframe df using df.columns.

Then, you can create an empty dataframe by passing this column list to columns parameter.

Use the below snippet to create an empty dataframe from other dataframe columns.

columns_list = df.columns

df2 = pd.DataFrame(columns = columns_list)

print(df2)

Printing the new dataframe df2 will show the output where you can see the columns from the dataframe df is used to create the dataframe.

Output

Empty DataFrame
Columns: [Name, Quantity, Price, Column_2, Column_3]
Index: []

This is how you can create a dataframe using other dataframe columns.

Conclusion

To summarize, you’ve learned how to create an empty dataframe and also learned the various options available in the create dataframe operation. You’ve also appended columns and rows to the newly created dataframe.

If you have any questions, comment below.

You May Also Like

• How to Iterate over Rows in Pandas Dataframe
• How To Select Rows From Pandas Dataframe Based On Condition