We have a ball, which moves to a random position in the screen when clicked and also changes its color.

Lets divide that into steps:

1. First we need to build the ball.
2. Next we need to move the ball to a new position when pressed on it.
3. Finally, we have to also change the color of the ball when pressed.

Building the Ball

Or is it really that simple?

So I was working on a project where I needed to implement a simple stopwatch to let the user know about the time interval spent doing a particular task. Okay, how difficult will that be right? Dart already has a Stopwatch() class that I can use directly.

But as I started working on it, I realized how awfully wrong I was and how such a simple feature is not that easy to implement and how the Stopwatch() class in Dart really sucks! (That is only one person’s opinion. It may not suck for you)…

By asking myself just one simple question

Lets Face it we don’t need focus all the time. I don’t need to be focused all damn day. I don’t need focus while chatting with my loved ones or hanging out with my friends or watching some random YouTube videos or while just lying around like a potato chips packet on bed.

But when the time comes in my schedule to work, that's when I need all the focus I can gather (maybe more).

But the problem is I can’t lift myself off the damn sofa and leave the smartphone. It feels…

The Problem:

Given a gold mine of n*m dimensions. Each field in this mine contains a positive integer which is the amount of gold in tons. Initially the miner is at first column but can be at any row. He can move only (right, right up ,right down) that is from a given cell, the miner can move to the cell diagonally up towards the right or right or diagonally down towards the right. Find out maximum amount of gold he can collect. (source: geeksforgeeks.org)

So I was trying to learn about Reinforcement Learning, and then I came across this thing called ‘Value Iteration’. I really couldn’t wrap my head around Value Iteration. It was very difficult for me to understand how it worked and how it could help an agent to find the optimal policy. Then I got an idea.

What better way to understand “Value Iteration” than to use it to solve some game or environment. Thus I began my journey to find some game easy enough problem to solve. And then I stumbled upon this fairy from OpenAI.

Coding the GridWorld Example from DeepMind’s Reinforcement Learning Course in Python

Here I calculate the state value functions for all states in the GridWorld example from the well renowned David Silver’s Reinforcement Learning Course.