US Medical insurance project review + specific questions

hlmattil · February 17, 2022, 1:53pm

I found the Medical Insurance project to be very open, so I was able to make it as challenging as I wanted it to be.

I decided to answer the following:

Calculate the average age of the people in the data set
Rank the areas by representation in the data set
Calculate how much more the average smoker pays in insurance than the average non-smoker
Find average age for someone with at least one child in the data set
Find average number of children men have vs. average number of children that women have
Answer if women or men pay more on average for health insurance and how much
Find a line of best fit for predicting a woman’s BMI based on her age

My questions:

How is my presentation of information? Did I provide enough information about my goals, or should I have given more explanations / used more markdown to make it readable?
Do the questions I chose to answer make sense/are they worthwhile to ask?
I was able to calculate a rough line of best fit, but I didn’t figure out how to actually plot that specific line on the graph I made. Instead, I followed a guide online for finding and displaying a line of best fit using matplotlib and numpy, which wasn’t exactly the same as the line if best fit I calculated (mine was an approximation). How can I get my line of best fit to display on my graph similar to how it’s done in the section Plot the points and line of best fit ?
I relied a lot on creating variables set to zero and updating those variables. For example, in the calculation of the avg number of children for men and women, I have the following:
num_men = 0
num_men_children = 0
men_child_count = 0

num_women = 0
num_women_children = 0
women_child_count = 0
Is there some better way to do this, or is creating and updating variables the most efficient method for calculating averages with a dataset like this?

My project can be found below. Any feedback is welcome—I’m a total beginner!

github.com

hlmattil/Solo-Medical-Insurance-Project/blob/12095a93c4767c7db0c6ec4cac4cfcbbc1c4d2a0/us-medical-insurance-costs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# **US Medical Insurance Costs**\n",
    "This project finds trends in US patient medical data. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import the data, organize into a list\n",
    "import csv\n",
    "with open('Insurance.csv', newline='') as insurance_csv:\n",
    "    fieldnames = ['age','sex','bmi','children','smoker','region','charges']\n",

This file has been truncated. show original

chang-joonlee · February 21, 2022, 7:39am

I think your code is tidy and neat!

Slightly off-topic, I thought of doing a similar thing with scatterplot and finding the best fit for age vs bmi, but the plot was too scattered and so I thought the line of best fit was probably meaningless. Instead, what I did was divide the bmi group into subgroups (underweight, healthy, overweight, and obese), calculate the average age for each subgroup and then plot as a line graph. I think that made more sense than the scatterplot. Maybe something to consider!

hlmattil · March 4, 2022, 6:12pm

Awesome feedback! I appreciate the idea

Silly question perhaps, but did you plot the graph with age on the x axis or the y axis?