Sort pandas DataFrame by Date in Python (Example)

In this tutorial, I’ll show how to order the rows of a pandas DataFrame by dates in Python programming.

The post will contain these content blocks:

2) Example: Order pandas DataFrame by Dates Using to_datetime() & sort_values() Functions

3) Video & Further Resources on this Topic

Let’s start right away…

Example Data & Software Libraries

In order to use the functions of the pandas library, we first have to import pandas.

import pandas as pd                                                      # Import pandas

As a next step, we’ll also have to create some data that we can use in the example syntax below:

data = pd.DataFrame({'dates':['02/25/2023','01/01/2024','17/11/2019','17/10/2022','02/02/2022'],  # Create example DataFrame
                     'values':range(1, 6)})
print(data)                                                              # Print example DataFrame

table 1 DataFrame sort pandas dataframe date python

Have a look at the previous table. It shows that our example data consists of five rows and the two columns “dates” and “values”.

Example: Order pandas DataFrame by Dates Using to_datetime() & sort_values() Functions

The following code illustrates how to reorder the rows of a pandas DataFrame by a date column.

For this, we first should create a copy of our example data, so that we can keep an original version of our data:

data_new = data.copy()                                                   # Create duplicate

In the next step, we have to convert the class of our date column to the datetime class using the to_datetime function. This is important for the sorting process later on, because otherwise Python would sort our data alphabetically.

data_new['dates'] = pd.to_datetime(data_new.dates)                       # Convert to date

Next, we can apply the sort_values function to order our pandas DataFrame according to our date variable:

data_new = data_new.sort_values(by = ['dates'])                          # Sort rows of data

Finally, let’s print our data to the console:

print(data_new)                                                          # Print new data set

table 2 DataFrame sort pandas dataframe date python

The output of the previous Python programming code is shown in Table 2: As you can see, we have sorted our pandas DataFrame by its date column.

Video & Further Resources on this Topic

Do you need more explanations on how to order and rearrange rows by date? Then you should have a look at the following YouTube video of the Statistics Globe YouTube channel. I show and eyplain the examples of this tutorial in some more detail.

For further examples, check out the following video of Corey Schafer’s YouTube channel. In the video, he explains how to sort pandas DataFrames using the Python programming language. You can find the video below:

Furthermore, you may want to read the other Python programming tutorials on my homepage:

This article has illustrated how to sort and rearrange a pandas DataFrame by year, month, and day in a date column in Python programming. Tell me about it in the comments below, in case you have any additional questions. Besides that, please subscribe to my email newsletter in order to receive updates on new tutorials.

2 Comments. Leave new

Guilherme Silva
March 30, 2022 6:31 pm

Hi, thanks for this excellent tutorial!

I repeated steps using your example and I worked well.

But, then I tried with a dataframe built from reading a excel spreadsheet and it did not work so well.

Dates from the spreadsheet are in ‘dd/mm/YYYY’ format.

Below is a shortened version of the Dataframe:
df = {‘Data’: {0: ’29/09/2021′, 1: ’14/10/2021′, 2: ’18/10/2021′, 3: ’26/10/2021′, 4: ’27/10/2021′, 5: ’05/11/2021′, 6: ’02/12/2021′, 7: ’01/01/2022′, 8: ’25/01/2022′, 9: ’03/02/2022′, 10: ’29/03/2022′, 11: ’29/09/2021′, 12: ’14/10/2021′, 13: ’18/10/2021′, 14: ’26/10/2021′, 15: ’27/10/2021′, 16: ’05/11/2021′, 17: ’02/12/2021′, 18: ’01/01/2022′, 19: ’25/01/2022′, 20: ’03/02/2022′}, ‘Hora’: {0: ’11:23:45′, 1: ’09:15:38′, 2: ’15:21:09′, 3: ’15:24:23′, 4: ’15:24:33′, 5: ’15:25:44′, 6: ’15:25:54′, 7: ’15:26:20′, 8: ’15:26:30′, 9: ’15:27:00′, 10: ’15:27:07′, 11: ’15:51:11′, 12: ’10:36:19′, 13: ’10:39:19′, 14: ’10:42:18′, 15: ’09:09:32′, 16: ’09:09:59′, 17: ’09:12:25′, 18: ’09:13:17′, 19: ’11:45:14′, 20: ’13:15:11′}, ‘Status’: {0: ‘Alarme’, 1: ‘Alarme’, 2: ‘Alarme’, 3: ‘Alarme’, 4: ‘Alarme’, 5: ‘Alarme’, 6: ‘Falha’, 7: ‘Falha’, 8: ‘Falha’, 9: ‘Falha’, 10: ‘Falha’, 11: ‘Falha’, 12: ‘Falha’, 13: ‘Falha’, 14: ‘Falha’, 15: ‘Notificação’, 16: ‘Notificação’, 17: ‘Notificação’, 18: ‘Notificação’, 19: ‘Notificação’, 20: ‘Notificação’}}

When I run to_datetime() function data from ‘Data’ column is converted to datatime64[ns].

But when I try to sort data based on ‘Data’ column values it mess up the data completely.

Follows below full code for reference:
>import os, sys, pandas
>header_list = [‘Data’, ‘Hora’, ‘Status’]
>df = {‘Data’: {0: ’29/09/2021′, 1: ’14/10/2021′, 2: ’18/10/2021′, 3: ’26/10/2021′, 4: ’27/10/2021′, 5: ’05/11/2021′, 6: ’02/12/2021′, 7: ’01/01/2022′, 8: ’25/01/2022′, 9: ’03/02/2022′, 10: ’29/03/2022′, 11: ’29/09/2021′, 12: ’14/10/2021′, 13: ’18/10/2021′, 14: ’26/10/2021′, 15: ’27/10/2021′, 16: ’05/11/2021′, 17: ’02/12/2021′, 18: ’01/01/2022′, 19: ’25/01/2022′, 20: ’03/02/2022′}, ‘Hora’: {0: ’11:23:45′, 1: ’09:15:38′, 2: ’15:21:09′, 3: ’15:24:23′, 4: ’15:24:33′, 5: ’15:25:44′, 6: ’15:25:54′, 7: ’15:26:20′, 8: ’15:26:30′, 9: ’15:27:00′, 10: ’15:27:07′, 11: ’15:51:11′, 12: ’10:36:19′, 13: ’10:39:19′, 14: ’10:42:18′, 15: ’09:09:32′, 16: ’09:09:59′, 17: ’09:12:25′, 18: ’09:13:17′, 19: ’11:45:14′, 20: ’13:15:11′}, ‘Status’: {0: ‘Alarme’, 1: ‘Alarme’, 2: ‘Alarme’, 3: ‘Alarme’, 4: ‘Alarme’, 5: ‘Alarme’, 6: ‘Falha’, 7: ‘Falha’, 8: ‘Falha’, 9: ‘Falha’, 10: ‘Falha’, 11: ‘Falha’, 12: ‘Falha’, 13: ‘Falha’, 14: ‘Falha’, 15: ‘Notificação’, 16: ‘Notificação’, 17: ‘Notificação’, 18: ‘Notificação’, 19: ‘Notificação’, 20: ‘Notificação’}}
>df = pandas.DataFrame(df)
>df[‘Data’] = pandas.to_datetime(df.Data)
>df[‘Data’] = df[‘Data’].dt.strftime(‘%Y-%d-%m’)
>df = df.sort_values(by = [‘Data’])
>print(df)

Would you know why does that happen?
Would that be due to datetime format from my dataset?
Yours looks like a mixture of ‘mm/dd/YYYY’ and ‘dd/mm/YYYY’

Thanks!

Reply
- Joachim
  March 31, 2022 9:14 am
  Hey Guilherme,
  
  Thanks a lot for the kind words, glad you like the tutorial!
  
  It seems like the problem in your case is that your data is not formatted as a pandas DataFrame. Please have a look at the simplified example below:
  import pandas as pd df = pd.DataFrame({'Data': {0: '29/11/2021', 1: '14/05/2021', 2: '18/10/2021', 3: '26/12/2021', 4: '27/10/2021', 5: '05/11/2021'}, 'Hora': {0: '11:23:45', 1: '09:15:38', 2: '15:21:09', 3: '15:24:23', 4: '15:24:33', 5: '15:25:44'}}) print(df) # Data Hora # 0 29/11/2021 11:23:45 # 1 14/05/2021 09:15:38 # 2 18/10/2021 15:21:09 # 3 26/12/2021 15:24:23 # 4 27/10/2021 15:24:33 # 5 05/11/2021 15:25:44 df['Data'] = pd.to_datetime(df.Data, format = '%d/%m/%Y') print(df) # Data Hora # 0 2021-11-29 11:23:45 # 1 2021-05-14 09:15:38 # 2 2021-10-18 15:21:09 # 3 2021-12-26 15:24:23 # 4 2021-10-27 15:24:33 # 5 2021-11-05 15:25:44 df = df.sort_values(by = ['Data']) print(df) # Data Hora # 1 2021-05-14 09:15:38 # 2 2021-10-18 15:21:09 # 4 2021-10-27 15:24:33 # 5 2021-11-05 15:25:44 # 0 2021-11-29 11:23:45 # 3 2021-12-26 15:24:23
  Note that I have created a pandas DataFrame at the beginning of this example.
  
  Regards,
  Joachim
  Reply