🔥 Matt Dancho (Business Science) 🔥 Profile picture
On a mission to grow your business data science skills and accelerate your career | Get my free 5-day business for data scientists course 👇

Apr 20, 2023, 25 tweets

BIG NEWS: #ChatGPT breaks #Python vs #R Barriers in Data Science!

Data science teams everywhere rejoice.

A mind-blowing thread (with a FULL chatgpt prompt walkthrough). 🧵

#datascience #rstats

It's NOT R VS Python ANYMORE!

This is 1 example of how ChatGPT can speed up data science & GET R & PYTHON people working together.

(it blew my mind)

This example combines #R, #Python, and #Docker.

I created this example in under 10 minutes from start to finish.

I’m an R guy.

And I prefer doing my business research & analysis in R.

It's awesome. It has:

1. Tidyverse - data wrangling + visualization
2. Tidymodels - Machine Learning
3. Shiny - Apps

But the rest of my team prefers Python.

And they don't like R... it's just weird to them.

So I wanted to see if I could show them how we could work together...

Let’s start with a prompt.

I asked chatgpt to find a data set that I used for this example.

...ChatGPT found it...

... And gave me this code to read the data...

I prefer the tidyverse, so I asked Chatgpt to update the code.

That looks better.

With the data in hand, it’s time for some Data Science.

I asked this simple question.

ChatGPT's response was impressive.

But, even though I’m an R guy, my team uses Python for Deployment…

In the past, that’s a huge problem.

(resulting in days of translations from R to Python with Google and StackOverflow)

But now, that’s 1 minute of effort with chatGPT.

Can I show you?

I asked chatgpt to convert the R script to python...

And in 10 seconds chatgpt made this python code with pandas and scikit learn.

ChatGPT did in 10 seconds something that would have taken me 2 hours.

But let’s continue.

The reason we had to convert to Python is for “deployment”

Deployment is just a fancy word for allowing others to access my model so they can use it on-demand.

So I asked chatGPT this:

And ChatGPT made me a Python API using FastAPI.

But this code is useless…

… Without a docker environment.

So I asked chatGPT to make one:

And chatGPT delivered my Docker Environment's Dockerfile:

So in under 10 minutes, I had ChatGPT:

1. Make my research script in R.

2. Create my production script in Python for my Team

3. And create the API + Docker File to deploy it.

But when I showed my Python team, instead of excited...

...They were worried.

And I said, "Listen. There's nothing to be afraid of."

"ChatGPT is a productivity enhancer."

They didn't believe me.

My Conclusion:

You have a choice. You can rule AI.

Or, you can let AI rule you.

What do you think the better choice is?

If you want help, I'd like you to join me on a free #ChatGPT for #DataScientists Workshop on April 26th. And I will help you Rule AI.

What's the next step?

👉Register Here: us02web.zoom.us/webinar/regist…

Share this Scrolly Tale with your friends.

A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.

Keep scrolling