Hello and welcome! Today we’re going to be going over ChatGPT’s Advanced Data Analysis features. My name is Chuck Downing, and I’m a PhD student here at MIT Sloan. Today, we’re going to be covering how to enable access and use ChatGPT’s Advanced Data Analysis features and cover some common use cases and potential things to be concerned about along the way. Let’s get started!
ChatGPT’s Advanced Data Analysis feature is a very useful tool provided within the premium subscription feature of ChatGPT for paying members. What it allows you to do is upload files directly to ChatGPT’s system, where it can then interface directly with the data and make changes. This allows for more efficient and accurate test cases for your code and data analysis. The big advantage here is that ChatGPT can now test and verify that its code is correct as it goes along. With this being said, we’re going to show exactly how to enable this and how you can use it for your everyday tasks.
To enable ChatGPT’s Advanced Data Analysis feature, you need to have a premium paying subscription account. Once you have this enabled, you can go to the bottom left and click the three buttons next to your name. From here, click ‘Settings’ and navigate on the left tab bar to ‘Beta Features’. Then, activate Advanced Data Analysis by clicking the button next to the ‘Advanced Data Analysis’ tab.
Now that you’ve enabled that on your account, you can go to the bottom left and click the three buttons next to your name. From here, click ‘Settings’ and navigate on the left tab bar to ‘Beta Features’. Then, activate Advanced Data Analysis by clicking the button next to the ‘Advanced Data Analysis’ tab.
Once you have enabled Advanced Data Analysis, you can start using it in your sessions. To activate a specific Advanced Data Analysis section, go to the bottom left and click the three buttons next to your name. From here, click ‘Settings’ and navigate on the left tab bar to ‘Beta Features’. Then, activate Advanced Data Analysis by clicking the button next to the ‘Advanced Data Analysis’ tab.
Now that you have enabled Advanced Data Analysis, you can start using it in your sessions. To activate a specific Advanced Data Analysis section, go to the bottom left and click the three buttons next to your name. From here, click ‘Settings’ and navigate on the left tab bar to ‘Beta Features’. Then, activate Advanced Data Analysis by clicking the button next to the ‘Advanced Data Analysis’ tab.
Once you have activated a session within Advanced Data Analysis, you can upload files directly to the system. Click the plus button next to the ‘Send a message’ option and upload your file. This allows you to interface directly with ChatGPT, where it can test its errors live as you go. It provides a more accurate and seamless experience for you to write, test, and upload code.
For this demonstration, we’re going to be showing you how to use this feature and some common use cases associated with it. For the purposes of this demonstration, I’m going to be using the World Bank’s carbon emissions database. This data covers carbon emissions per capita from 1960 to 2020 for every country with data available in the World Bank. An important feature of this dataset is that it has no observations from 1960 to 1989, despite there being columns available for those years. So, we’re going to read the data, do some cleaning, get a better sense of the data with some summary statistics, and then visualize the data before moving to a regression analysis.
Now that we have our data uploaded, we can start working with it. First, let’s read in the data and describe its contents. This is a good first step to ensure that ChatGPT has a common understanding of the data. We can do this by sending a message that says ‘Read in this dataset and describe its contents’. ChatGPT will then analyze the data and provide a description of what it knows about the data.
Once we have a better understanding of the data, we can proceed with cleaning and transforming it. We can remove any years where all the values are null and provide summary statistics on the average emissions each year. We can do this by sending a message that says ‘Remove any years where all the values are null and provide summary statistics on the average emissions each year’. ChatGPT will then clean the data and provide summary statistics.
After cleaning and transforming the data, we can move on to visualizing it. We can create a graph of emissions over time for a specific country. For example, if we want to create a graph of emissions for the United States from 1990 to 2020, we can send a message that says ‘Provide a graph of emissions for the United States across time’. ChatGPT will then generate a graph of emissions for the United States over time.
Finally, we can perform a regression analysis on the data. For example, we can run a regression of developing versus developed nations on their emissions over time. We can ask ChatGPT to make its own classifications for developing and developed countries and perform the regression analysis. We can do this by sending a message that says ‘Provide a regression and interpretation for the emissions over time of developing and developed nations. Make your own classifications for developing and developed’. ChatGPT will then perform the regression analysis and provide the results.
Once we have completed our analysis, we can save our work. We can ask ChatGPT to provide download links for the new CSV dataset and the .py code written during the session. We can do this by sending a message that says ‘Provide a download link to the new CSV dataset and the .py code written during this session’. ChatGPT will then provide download links for the dataset and the code.
We hope this article has helped you understand how to enable and use ChatGPT’s Advanced Data Analysis features. It is a powerful tool for data analysis, cleaning, visualization, and regression analysis. We encourage you to explore and experiment with this feature in your own work. Thank you for your time, and we look forward to seeing you use ChatGPT’s Advanced Data Analysis in your own projects!
