AI Analysis Tutorial: Part 1

We're going to use Devra AI to perform a two-part data analysis. Here we are on a Kaggle dataset for heart disease prediction. This is my project directory where I conduct all my Kaggle data analyses. I've already downloaded the heart disease data, so let's open Devra in the Kaggle project that I've set up.

First, we'll create a basic analysis of the heart disease data. Let's create and start the task and see what happens. The idea here is that the first analysis will be basic, and then we will use that analysis to get an idea for what our second analysis will be.

Devra has generated a proposal. Let's say yes, that plan looks good. It has created a placeholder notebook, and we see that placeholder notebook pop up here. Now, it's got our full notebook that it just created. Let's say yes, let's create that file and run it to see what happens.

Alright, here's the notebook. It's come up with a heart disease data analysis. It's loading in our libraries and the dataset, displaying the first few rows of it, performing some basic data cleaning, and conducting an Exploratory Data Analysis (EDA). Here are some histograms of the numeric columns and, finally, a correlation heat map.

This is where we start getting into some of the good stuff. This basic analysis can give us an idea of the next steps we want to take. Looking at this correlation heat map, we see that among the different variables, there's a relatively strong correlation between age, cholesterol, and heart disease. This could be valuable for isolating variables for the next step.

This basic analysis also includes prevalence data. This is a solid first step. We didn't have to do any coding; we just input the data, and Devra went through it and figured out a basic analysis. This is a great starting point for us to then conduct a secondary analysis, which we'll do in part two of this series.