A nodejs server and an Angular 8 frontend to simulate daily assignments of unitn-informatica statistic course
link to the site: https://unitn-statistica.herokuapp.com/index
These is a project consisting on a server where simulation scripts to get aproximated solutions of unitn-informatica statistic course's daily assignments are published. The server serves a site where you can choose the exercise, insert your own exercise data/parameters and run the simulation
Let's say that a new assignment asks to solve a statistic problem and the problem is to calculate the probability of getting number three after a six-faced dice is tossed. There are two possibilities to solve the problem. The first is solving the problem, by applying math rules. This is the best way, but you must be good and you are never 100% sure of doing the exercise right. The second is to execute by hand the experiment and get an aproximate result. The more iterations you do, the better the result is. In this case, let's say you should toss a dice ten million times and count how many times the number three is came up. You can't mistake it, but it is quite long. So, I made a third way, wich is computer simulation. I make a script simulate a dice toss millions of times and count how many times number three is came up. So you get a good approximated solution and you don't have to lose tons of your time.
This project is based on computer simulations. Neither gives it perfect solutions nor provides it you the steps needed to solve it mathematically. You still have to solve your exercises by yourself - the only advantage is that you will never be unsure again, you only will need to check if the results are similar to the simulated ones.
- Go to the website: https://unitn-statistica.herokuapp.com/index
- Type the credentials (lagrange, lagragna) and select the exercise date. If you are offline or there was not an exercise in that date, an error toast will be desplayed.
- Insert your exercises' values/parameters, by carefully following the fields instructions. Put a big number, one million or something, in the test iterations field. The bigger you put, the better you'll get but the longer you'll wait.
- Wait till the script has finished to execute. There is a pretty loading spinner wich shows you the current progress.
- Check the solutions.
The server was made with Javascript, specifically by using NodeJs. It sends to the client js scripts to run the simulations and serves the frontend. The frontend was made with Angular, by using Angular Material and Angular Flexbox for the style and Angular Universal to have an efficient server-side rendering. There is also the Angular Service Worker, wich makes the site a pwa.
If you are a developer and want to build it locally, download the project, install node, run npm i in the project directory and then you shall run node main.local to start a local server (out of heroku) serving the site and listening to port 8000. The frontend http domain should change to 'http://localhost:8000' to comunicate with the local server and be recompiled by moving to its directory, cd angular and by running npm run build:ssr.
Made with dree
unitn-statistica
├─> angular
│ ├── angular.json
│ ├── ngsw-config.json
│ ├── package-lock.json
│ ├── package.json
│ ├─> src
│ │ ├─> app
│ │ ├─> assets
│ │ │ └─> icons
│ │ ├── favicon.ico
│ │ ├── index.html
│ │ ├── main.server.ts
│ │ ├── main.ts
│ │ ├── manifest.webmanifest
│ │ ├── robots.txt
│ │ ├── styles.scss
│ │ ├── tsconfig.app.json
│ │ ├── tsconfig.server.json
│ │ ├── tsconfig.spec.json
│ │ └── tslint.json
│ ├── tsconfig.json
│ └── tslint.json
├── app.json
├─> exercises
│ ├── 2019-03-14.js
│ ├── 2019-03-15.js
│ └── exercises.js
├─> frontend
│ ├─> browser
│ └─> server
├── logo.svg
├── main.js
├── main.local.js
├── package-lock.json
├── package.json
├── Procfile
└─> utilities
├── redirect.js
└── serializer.js