Random number generation and Simulation
google play linkA lot of real world measurements are first study by collecting samples. When we identify the distribution they belong to, we can simulate the data with random number generated from that known distribution.
This tool allows you to configure the parameters of a selected distribution. Click the triggering icon to generate the data and display the histogram. By repeating the generating runs, we can get a better visual feel of the variation of sampling from the same distribution as if you collect real life sample measurements. This process is the well known Monte Carlo Simulation.
Discrete Distribution
- Uniform Typical toss of coin, rolling dice, roulette are of this distribution.
Method used - java util's internal library.
- Binomial Dichotomous outcome
Method used - compare with sum of logarithms
- Poisson Counts
Method used - Inverse transform sampling for lamda < 50, else use normal
approximation
Continuous Distribution
- Normal A lot of measurement like weight, height, temperature, pressure especially sample mean.
Method used - accept or reject on logarithms
- Exponential Arrival time, service turn around time.
Method used - logarithms transformation
- Logistic Growth model
Method used - logarithms of the reciprocal conjugate
- Laplacian Also known as Double Exponential like difference on time
Method used - subtracting two independent exponential
- Cauchy Earthquake and the like
Method used - tangent of angle
There is a menu icon to allow you to send generated data to another app.
The final icon is to display this document
Have fun