Final Review on medical insurance cost project

afroleus · September 23, 2023, 6:04am

Hello. I have finished my first project. Since the scope was user-defined, I decided on getting a bunch of baseline analysis to couple with more data or to see trends!
Working up to an hour a day after work and family time, this took a couple weeks to fully implement and make sure everything worked. I didn’t end up using BMI as a variable to study
Happy Coding!

Afroleus/Codec_jects/blob/main/us-medical-insurance-costs main project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###### U.S. Medical Insurance Costs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Welcome to my Insurance Project! The goal of this project is to extrapolate findings from data on medical insurance. It will be focused in 3 areas after finding the average of all charges. BMI is reletively unreliable to use as a metric so we will be avoiding it here. This project just has base analysis and is open to more analysis.\n"
     ]

This file has been truncated. show original

We hope you enjoyed this project!

boudwin · September 29, 2023, 1:00pm

Hi! I have not seen the map function before. That looks like a useful function!
I am curious, how did you come to the conclusion that 80% of patients didn’t have children? I used very different calculations for my project, but I came to the conclusion that 57% of the patients did have children. Though comparing age to the number of children was an interesting idea and one that I now want to explore further.
I also saw that the Southeast had higher costs, but it also had quite a few more smokers than the other regions - one-third of the whole country.

Congratulations on finishing this project!

afroleus · September 30, 2023, 4:34am

it took a while to research the different functions I wanted to use as a baseline. As for the percentages, there are 1337 rows of data that are given to you. the percentage comes from dividing how many out or 1337, or in this case, half of that after splitting the data in 2.
For the children, most of the age ranges didnt have children. It was mainly focused in the older age groups, and it was a minority of people who had it
I still have to improve upon it, as it took me a while to do, but there is always room for improvement

boudwin · October 2, 2023, 8:07am

Thank you for the response! I appreciate being able to go over differences like this.
When I do print(children.count(‘0’)), the number that is printed is 574. Out of 1338 patients, 574 would be 43% of the total, meaning that about 57% of the patients have children. I am not sure, but is your code at the end, when you do your sums, is it counting and dividing the number of children rather than the number of patients who have children?

afroleus · October 4, 2023, 4:16am

I did number of children by age group, and split everything into fifths. I think I did do that type of math, although my original purpose was to see which age groups had the most children, then devolved into which percentage had kids because of the staggering number of people who didn’t have children, and which age group had it concentrated the most in. I hope this answer your question! I put my steps and form on thinking in # comments throughout the project