ChatGPT Data Analysis With Noteable

ChatGPT in Data Analysis

Note that the plugin has the name “Noteable”, not “Notable”.

In a tutorial video by Chad Skelton, a former data journalist and current instructor in journalism and data visualization, he demonstrates the capabilities of ChatGPT in data analysis and visualization.

Using a dataset on bike thefts from the Vancouver Police Department, Skelton showcases how ChatGPT, with the help of a plug-in from Noteable, a Jupyter notebooks website, can analyze and visualize data with minimal input.

ChatGPT’s ability to identify patterns, generate insights, and even catch potential data errors is impressive. It provides an effortless solution for data analysis, reducing the need for complex formulas and models.

This makes data analysis more accessible to a wider audience, including those without a background in data science or programming.

Noteable is “Jupyter notebooks on steroids”

Another video about Reuven Lerner, coming from the perspective of a Python trainer.

The video is quite technical and assumes familiarity with Python, data analysis, and Jupyter notebooks. Certainly something that stretched my knowledgebase to its limits, since I’m mostly a MicroPython user.

Reuven demonstrates the capabilities of ChatGPT-4 model and its integration with Noteable.

Reuven’s video has three main sections.

The introduction talks to you about GPT-4, and how to activate it (pay $20 for ChatGPT Plus).

Then, he moves into an introduction to Noteable while demonstrating how to create a project in Noteable.

Finally, bam! He puts one and one together.

ChatGPT and Noteable cross paths as Reuven demonstrates how to use ChatGPT to retrieve data and create Jupyter notebooks in Noteable.

He shows how to instruct ChatGPT to download data from specific URLs, turn it into a Jupyter notebook in Noteable, join data frames together, and create a line plot using Seaborn.

A seasoned user like himself, he warns that you should have very specific queries and careful with their instructions to ChatGPT. Remember, you only have 25 queries in three hours with GPT-4.

Setting Up ChatGPT for Data Analysis

Setting up ChatGPT for data analysis is a straightforward process. With the Noteable plug-in, users can easily integrate ChatGPT into their data analysis workflow.

Once set up, users can simply ask ChatGPT questions about their dataset, and the model will provide answers in an understandable format. This allows users to explore their data in a conversational manner, making the process of data analysis more intuitive and engaging.

Data Literacy reviews Noteable

This comes from Ben Jones who runs the Data Literacy channel on YouTube.

Ben’s got this dataset from data.gov about EVs in Washington state. and he’s going to use it for the Noteable demo.

He creates a prompt for the Notable plugin, telling it to analyze the electric vehicle data.

He’s pretending he’s a Tesla sales rep who wants to convince people to buy a Tesla over other electric vehicles.

He asks the plugin to include a description of what it’s doing, do some machine learning, visual analytics, and exploratory data analysis, and give some data-driven suggestions.

Once he sends the prompt, ChatGPT starts working. It requests the project he gave it, and then creates a notebook within it called “EV analysis”.

He points out that this is a cool feature of the Noteable plugin – it automatically puts the Python code it generates into a notebook.

As he watches, ChatGPT loads the dataset, creates a header, and gives a preview of the variables in the dataset. It also gives a printout of the table, showing the first few rows.

ChatGPT then gives some basic stats about the dataset – how many rows and columns it has, the type of data in each column, and how many null or missing values there are. It also gives some basic descriptive stats of the numerical columns.

Next, it plots a histogram of the model year and a breakdown of the top 10 makes of electric vehicles. It then splits the data into a training version and a test version and does some machine learning, using a random forest regressor.

Finally, it gives some recommendations based on the findings. For example, a Tesla sales rep could emphasize the popularity of Tesla vehicles, the high average electric range, and the superior performance of Tesla models in terms of electric range.

Ben wraps up by saying how amazing he finds this technology. He points out that there could be errors, as the language model is interpreting his prompt and converting it into Python and markdown.

But he’s really impressed with the Noteable team for creating this plugin and thinks it’s going to change the way data analysts work.

Getting Started With Noteable

But thanks to Dave from Noteable, who demonstrated how to get started with the Noteable plugin within ChatGPT.

He guides the viewer through the process of logging into ChatGPT, switching to plugins, searching for the Noteable plugin, and installing it.

After signing into a Noteable account, the plugin is enabled in ChatGPT.

Dave then instructs ChatGPT to create a first notebook that prints ‘Hello World’. The model creates the notebook, adds content to it, and executes the cells within the notebook.

Once completed, a link is provided to view the notebook on Noteable. This tutorial showcases how ChatGPT can be used for data analysis with the help of Noteable.

Uploading datasets to Noteable

While Chad’s video is pretty comprehensive, I had to find other resources on how managed to upload his bikethefts.csv to Noteable.

There are several ways to get data into ChatGPT for analysis. These include using Excel, CSV, raw text, JSON, PostgreSQL, Snowflake, and BigQuery databases.

The video embed above has timestamps. Go to CSV if that’s what you’re trying to upload. Whatever data you have can be loaded into a Notable notebook, which then allows access to it through ChatGPT.

The Notable plugin can be used to install necessary toolkits, load data, and perform analysis. This makes the process of data analysis with ChatGPT completely effortless.

The Future of Data Analysis with ChatGPT

The potential of ChatGPT in data analysis is vast. As the model continues to improve and evolve, we can expect to see even more advanced features and capabilities.

For instance, future versions of ChatGPT might be able to provide more detailed insights, identify more complex patterns, and even predict future trends based on historical data.