Central Limit Theorem using Shiny

Nishant Upadhyay
25 Sep 2015

Introduction

This shiny pitch is intended to display the Central Limit Theorem (CLT) in the Coursera's Developing Data Products module.

My Approach to App creation:

CLT basically says that the sampling distribution of the parent distribution will be tending towards normality irrespective of the population(parent) distribution as the sample size increases

Building on this idea i have considered many differnt types of probability distributions namely:

  • Beta,Chi.squared,Exponential,F,Gamma,Lognormal,Normal,t,Uniform,Weibull)

I will taking the sampling distributions of the means to demonstrate CLT

The Application

The purpose of this app is to simulate data and test the central limit theorem. To do that, you only need to follow the following steps:

  1. Choose one of the Eleven kinds of predefined population distributions;
  2. Specify a sample size on the slider ranging from 10 to 500 in steps of 10;
  3. Voila!—You can see the parent distribution as well as the sampling distribution displayed on the right panel.

Demo Screen shot of the app

The data selection (Probability distribution & sample size)

png1

The plots on the App

(On the right side panel)

download_1 download

The app was hosted at shinyapps.io which can be accessed here

CLT_Shiny