jlworkshop.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Julia\n",
    "\n",
    "University of Waterloo Data Science Club  \n",
    "February 25, 2020\n",
    "\n",
    "(Follow along with these slides at https://wafy.me/jlworkshop.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Objective\n",
    "\n",
    "Better title: **Being Productive with Julia**\n",
    "\n",
    "- Julia is a (fairly new) general purpose programming language which is particularly well-suited for *technical computing*\n",
    "- Will not talk about too many specific details of the language (documentation is better for that)\n",
    "  - limited syntax discussion; will cover some useful tricks\n",
    "- Workshop more focused on: practical advice and as examples\n",
    "\n",
    "Please **always feel free to ask questions**.\n",
    "\n",
    "(Follow along with these slides at https://wafy.me/jlworkshop.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Outline\n",
    "\n",
    "1. Setting up basic tools\n",
    "2. Solving a math problem\n",
    "3. Some discussion on types and efficiency\n",
    "4. Neat non-programming application: vector graphics\n",
    "5. A simple data science/programming project\n",
    "\n",
    "(Follow along with these slides at https://wafy.me/jlworkshop.html)"
   ]
  },
  {
   "attachments": {
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Part 1: Installation and Tools\n",
    "\n",
    "Download and install [Julia 1.3.1](https://julialang.org/downloads/) (recommended) from the official website. Windows, Mac, Linux binaries available, or compile it yourself. (I usually compile the latest `master` branch, but this is only recommended if you are risk-seeking.)\n",
    "\n",
    "Open a terminal emulator or command prompt and run the `julia` executable. You should see this (version and shell may vary):\n",
    "\n",
    "![image.png](attachment:image.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Tools\n",
    "\n",
    "- REPL: great for calculations, installing packages, and rapid iteration\n",
    "- IJulia: great for visualizations, multimedia, and notebook-style iteration\n",
    "- Text editors: for projects (more permanent code)\n",
    "- IDEs: for big projects; a few options available, e.g. Juno https://junolab.org/\n",
    "\n",
    "Personally: I use [`nvim`](https://neovim.io/) and [TabNine](https://tabnine.com/) for projects. I don’t use IDEs because they are not necessary for the size and scope of the projects I work on.\n",
    "\n",
    "Today: We will use REPL and IJulia. REPL is immediately available upon downloading/installing Julia. We need to install `IJulia` using the package manager.\n",
    "\n",
    "(If you must: You can also run Julia on https://repl.it/. The advantage is you don’t need to install anything. The disadvantage is that this REPL is quite bad compared to the normal Julia REPL. No promises all this workshop will work... Let me know if you need help with this.)"
   ]
  },
  {
   "attachments": {
    "image2.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Installing IJulia\n",
    "\n",
    "- To use the package manager, type `]` at the REPL prompt. This activate the package manager REPL mode.\n",
    "- Then type `add IJulia`. Hit enter/return.\n",
    "\n",
    "Hopefully you see this:\n",
    "\n",
    "![image2.png](attachment:image2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Running IJulia\n",
    "\n",
    "- Now hit Backspace to exit the package manager REPL mode.\n",
    "- Type `using IJulia`, then hit enter/return, to import the `IJulia` package.\n",
    "- Type `notebook()`, then hit enter/return, to run a notebook.\n",
    "\n",
    "If everything goes well, this should run a jupyter notebook server and you should see a browser window come up.\n",
    "\n",
    "These slides are a Jupyter notebook. You can download it at https://wafy.me/jlworkshop.ipynb and open it through the Jupyter server running. (Optional; everything we will do can be done in REPL too, or typed by hand.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## REPL vs. IJulia\n",
    "\n",
    "- `notebook()` blocks your REPL session\n",
    "- If you hit Ctrl-C, you can interrupt it, but this will shut down the Jupyter notebook server\n",
    "- Up to you whether you want to use REPL for the rest of this workshop\n",
    "- You can also open a new REPL window and try both"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Quick Tips for REPL\n",
    "\n",
    "Most important:\n",
    "\n",
    "- Up/Down: Cycle through history. You can also type some letters, then Up/Down to cycle through a filtered history.\n",
    "- TAB: completion (for identifiers, special characters, etc.)\n",
    "\n",
    "Useful:\n",
    "\n",
    "- Ctrl-A: Go to start of line\n",
    "- Ctrl-E: Go to end of line\n",
    "- Ctrl-R: Reverse-i-search (if you want to find something in history)\n",
    "- Ctrl-L: Clear screen (if you want to hide some noise)\n",
    "- `\\alpha<TAB>`: Type special characters\n",
    "\n",
    "Special variables:\n",
    "\n",
    "- `ans`: result of last expression entered into REPL"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Part 2: Using Julia as a Calculator\n",
    "\n",
    "By this I don’t just mean computing $\\sum_{n=1}^9 \\sin\\left(\\frac{3\\pi}{4n}\\right)$, although of course you can do that too.\n",
    "\n",
    "We will focus on numeric inputs and results for now. No side effects, no file input/output, no UI, no API, no CSV files, no Excel spreadsheets, etc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "You have two fair coins, labeled A and B. When you flip coin A, you get 1 point if it comes up heads, but you lose 1 point if it comes up tails. Coin B is worth twice as much — when you flip coin B, you get 2 points if it comes up heads, but you lose 2 points if it comes up tails.\n",
    "\n",
    "To play the game, you make a total of 100 flips. For each flip, you can choose either coin, and you know the outcomes of all the previous flips. In order to win, you must finish with a positive total score. In your eyes, finishing with 2 points is just as good as finishing with 200 points — any positive score is a win. (By the same token, finishing with 0 or −2 points is just as bad as finishing with −200 points.)\n",
    "\n",
    "If you optimize your strategy, what percentage of games will you win? (Remember, one game consists of 100 coin flips.)\n",
    "\n",
    "Extra credit: What if coin A isn’t fair (but coin B is still fair)? That is, if coin A comes up heads with probability p and you optimize your strategy, what percentage of games will you win?\n",
    "\n",
    "*Citation: Can You Flip Your Way To Victory? by Zach Wissner-Gross, via FiveThirtyEight https://fivethirtyeight.com/features/can-you-flip-your-way-to-victory/, originally submitted by Abijith Krishnan*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Talking about the problem first...\n",
    "\n",
    "- Why can we do better than 50%?\n",
    "- By linearity of expectation, our expected score is going to be $0$, regardless of our choices."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "- Because if we have a lead, we can play it safer and flip coin A.\n",
    "- If we are running a deficit, we can play it riskier and flip coin B.\n",
    "- When we lose we lose big, when we win we win small.\n",
    "- So on average our score is $0$, but we can do better than $0$ more than half the time."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## A simpler version of the problem\n",
    "\n",
    "- Still two coins: A is ±1, B is ±2\n",
    "- Only flip $5$ times\n",
    "- Try a few strategies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Some setup..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "# Random is a standard library (we do not need to install it using `]`)\n",
    "using Random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "false"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Simple example of `Random`: Pick a random boolean.\n",
    "# This is true with 1/2 probability, false with 1/2 probability.\n",
    "# Try this a few times to convince yourself.\n",
    "rand(Bool)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "true"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rand(Bool)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "false"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rand(Bool)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Actually the score from coin A is not `true` or `false`, it’s 1 or -1.\n",
    "# We can use `rand([-1, 1])` instead.\n",
    "# `[-1, 1]` is a vector containing two `Int` values.\n",
    "rand([-1, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "coinA (generic function with 1 method)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We can define a function to give us the score from a new flip of coin A,\n",
    "# i.e. rand([-1, 1])\n",
    "# When you run `coinA()` that is the same as `rand([-1, 1])`.\n",
    "coinA() = rand([-1, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Run our function\n",
    "coinA()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Exercise: Define a function `coinB()` that returns -2 or 2 with equal probability."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Solution:\n",
    "coinB() = rand([-2, 2])\n",
    "coinB()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Null Strategy\n",
    "\n",
    "Here is our control strategy: just always flip coin A. (This strategy will end up with the same win rate as just always flipping coin B, but with half the score.) This calculates the score of the control strategy (i.e. not whether it wins, but the raw score itself)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-1"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# One way to write this\n",
    "control() = coinA() + coinA() + coinA() + coinA() + coinA()\n",
    "control()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# More extensible...\n",
    "control() = sum(coinA() for _ = 1:5)\n",
    "control()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "What does this do? Let’s take it apart..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1:5"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this is a range\n",
    "1:5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# it basically acts like a vector (without storing all its entries)\n",
    "length(1:5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5-element Array{Int64,1}:\n",
       " 1\n",
       " 2\n",
       " 3\n",
       " 4\n",
       " 5"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# you can convert it to a vector using `collect`\n",
    "collect(1:5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5-element Array{Int64,1}:\n",
       " 1\n",
       " 2\n",
       " 3\n",
       " 4\n",
       " 5"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# same as this\n",
    "[1, 2, 3, 4, 5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5-element Array{Int64,1}:\n",
       " -1\n",
       "  1\n",
       "  1\n",
       "  1\n",
       " -1"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This is list comprehension syntax (like Python, if you are familiar)\n",
    "# We use _ for the variable because we don’t need the number itself.\n",
    "[coinA() for _ = 1:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The sum function adds everything in a vector.\n",
    "sum([1, 2, 3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# combine everything\n",
    "# we can drop the [ ... ] because we don’t actually need this to be\n",
    "# explicitly stored in memory; by omitting them, this is a\n",
    "# generator expression and we save some memory allocations\n",
    "sum(coinA() for _ = 1:5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Very Smart Strategy\n",
    "\n",
    "Next a hopefully smarter strategy: flip coin B, unless our score is positive, then flip coin A. Goal: big losses, small wins, as discussed.\n",
    "\n",
    "To implement this strategy it is convenient to use **variables**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Score will reflect the current score. We start at 0...\n",
    "score = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Access the value of score.\n",
    "score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Now flip coin A. Note: a += b is shorthand for a = a + b.\n",
    "# (You can do this with most other operators too.)\n",
    "# The \"result\"/\"value\" (i.e. what gets printed in notebook/REPL) of an assignment is the assigned value.\n",
    "score += coinB()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The value of score has now changed.\n",
    "score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Now we need to make a decision on which coin to flip. We will need to branch on a condition, `score < 0`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "false"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The value of the condition is either `true` or `false`\n",
    "score < 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Now *if* our score is positive, flip A, otherwise flip B.\n",
    "# Introducing the `if`/`else` expression.\n",
    "# The \"result\"/\"value\" of an `if`/`else` expression is the \"result\"/\"value\"\n",
    "# of the last/tail expression in the block of the branch that was taken.\n",
    "if score > 0\n",
    "    score += coinA()\n",
    "else\n",
    "    score += coinB()\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# A shorthand for if/else: ternary operator `?`/`:`\n",
    "# May be familiar to those who have used C/C++/Java/etc.\n",
    "score += score > 0 ? coinA() : coinB()\n",
    "\n",
    "# two more times...\n",
    "score += score > 0 ? coinA() : coinB()\n",
    "score += score > 0 ? coinA() : coinB()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "So far we have only run our very smart strategy once. We can make it into a function, like `control`. There is a more imperative style syntax for function that is convenient if we need multiple statements. The return value of a function is the last statement.\n",
    "\n",
    "- You can also use the `return` keyword to return earlier, or if you prefer to see an explicit `return`.\n",
    "- (Probably if you are used to C++/Java/Python you find this tail expression return odd, but it is a feature taken from Scheme and Ruby, two languages Julia derives a lot of inspiration from.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function smart()\n",
    "    score = 0\n",
    "    score += score > 0 ? coinA() : coinB()\n",
    "    score += score > 0 ? coinA() : coinB()\n",
    "    score += score < 0 ? coinA() : coinB()\n",
    "    score += score < 0 ? coinA() : coinB()\n",
    "    score += score < 0 ? coinA() : coinB()\n",
    "    score\n",
    "end\n",
    "\n",
    "smart()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "- Actually in this case the last `score` line is totally unnecessary [why?], but most coding style generally suggests to at least include it [why?]..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "This is bad since we have just copy-pasted a line of code five times... we can fix this using a loop. This is kind of like the list comprehension we saw earlier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function smart()\n",
    "    score = 0\n",
    "    for _ = 1:5\n",
    "        score += score > 0 ? coinA() : coinB()\n",
    "    end\n",
    "    score\n",
    "end\n",
    "\n",
    "smart()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Now we are ready to compare the strategies. A strategy wins if its score is positive, i.e. `> 0`. We can run two simulations using loops.\n",
    "\n",
    "- We can also compute the probabilities exactly (exercise), but it is generally useful anyway to check work using dumbest possible code, if it is computationally feasible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "winrate (generic function with 1 method)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# This is our first function that takes arguments.\n",
    "# The `strategy` argument will be a zero-argument function that implements\n",
    "# a strategy (like `control` or `smart`).\n",
    "# The `iterations` argument will be the number of times to simulate.\n",
    "winrate(strategy, iterations) =\n",
    "    count(strategy() > 0 for i = 1:iterations) / iterations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.500005"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# should be around 0.5 (we can’t tie 0 with 5 flips, an odd number)\n",
    "winrate(control, 1000000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.531065"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# should be bigger than 0.5, if our hypothesis was right\n",
    "winrate(smart, 1000000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Cool, it works! But is it the best we can do?\n",
    "\n",
    "- We can do an exhaustive search through all possible strategies to find out.\n",
    "- First nail down what we are actually searching through. A few observations/questions:\n",
    "\n",
    "\n",
    "1. Can a mixed strategy (a strategy that involves randomness in making the decisions, not just flipping the coins) ever be preferred to both pure strategies (i.e. given the same information the strategy will always decide to flip A or always decide to flip B consistently)?\n",
    "2. Do we actually care about the results of all previous flips, or is the current score and number of coinflips remaining enough for the strategy to make a decision?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Answers: No, No (exercise)\n",
    "\n",
    "Suffices to consider: given $n$ current score and $m$ flips remaining, the next flip should be A / the next flip should be B."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "That is still a lot of strategies, even for $5$ flips. (For $100$ flips, there is no way this will be practical.)\n",
    "\n",
    "But consider only: given $n$ current score, $m$ flips remaining, choose between A and B _and then use the best possible strategy_ for new $n' := n + (\\text{result of flip})$, $m' := m - 1$.\n",
    "\n",
    "(Essentially: all optimal strategies are strategies that make a choice now and optimal choices in the future. There are no optimal strategies that make suboptimal choices in the future.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bestwinrate (generic function with 1 method)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Recursive implementation\n",
    "function bestwinrate(score, flips)\n",
    "    # Base case\n",
    "    if flips == 0\n",
    "        # No flips left. Our win rate is 1.0 if we won, 0.0 otherwise.\n",
    "        score > 0 ? 1.0 : 0.0\n",
    "    else\n",
    "        # win rate if we choose to flip A\n",
    "        winrateA = (\n",
    "            bestwinrate(score + 1, flips - 1) / 2\n",
    "          + bestwinrate(score - 1, flips - 1) / 2\n",
    "        )\n",
    "        # win rate if we choose to flip B\n",
    "        winrateB = (\n",
    "            bestwinrate(score + 2, flips - 1) / 2\n",
    "          + bestwinrate(score - 2, flips - 1) / 2\n",
    "        )\n",
    "        # pick the best of those two\n",
    "        max(winrateA, winrateB)\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.53125"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bestwinrate(0, 5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "(Actually this is pretty close to the smart strategy we chose, so that strategy is probably optimal in the five-flip case.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "This implementation makes no assumptions about the number of flips; we can ask `bestwinrate(0, 100)` if we want to.\n",
    "\n",
    "- Any issues with that?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "- (If you tried this, press Ctrl-C. In my experience, in the Jupyter notebook you may need to restart the whole kernel. This is really annoying and I am not sure why. One of the advantages of the REPL is it does not have this issue. If you know how to avoid this issue in the Juypter notebook, please let me know.)\n",
    "- Too slow; algorithm is exponential.\n",
    "- Problem: We are making four recursive calls in `bestwinrate`, each of which makes four more, ...\n",
    "- How to fix?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- One method: Memoize the result of the function (for every pair `m, n` that we are passed, remember the result; if we are asked that again, just fetch it and return)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Probably the easiest way to do this is to use the `Memoize` package. Install it in REPL with `]add Memoize<RETURN>`, then backspace to return to the usual REPL mode. In the notebook you can run:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m registry at `~/.julia/registries/General`\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[?25l\u001b[2K"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m git-repo `https://github.com/JuliaRegistries/General.git`\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[?25h"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m\u001b[1m  Resolving\u001b[22m\u001b[39m package versions...\n",
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m `~/.julia/environments/v1.5/Project.toml`\n",
      "\u001b[90m [no changes]\u001b[39m\n",
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m `~/.julia/environments/v1.5/Manifest.toml`\n",
      "\u001b[90m [no changes]\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "using Pkg  # standard library\n",
    "Pkg.add(\"Memoize\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "bestwinrate (generic function with 1 method)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Using Memoize is really easy. It provides a macro that does everything for you.\n",
    "using Memoize\n",
    "\n",
    "@memoize function bestwinrate(score, flips)\n",
    "    # Base case\n",
    "    if flips == 0\n",
    "        # No flips left. Our win rate is 1.0 if we won, 0.0 otherwise.\n",
    "        score > 0 ? 1.0 : 0.0\n",
    "    else\n",
    "        # win rate if we choose to flip A\n",
    "        winrateA = (\n",
    "            bestwinrate(score + 1, flips - 1) / 2\n",
    "          + bestwinrate(score - 1, flips - 1) / 2\n",
    "        )\n",
    "        # win rate if we choose to flip B\n",
    "        winrateB = (\n",
    "            bestwinrate(score + 2, flips - 1) / 2\n",
    "          + bestwinrate(score - 2, flips - 1) / 2\n",
    "        )\n",
    "        # pick the best of those two\n",
    "        max(winrateA, winrateB)\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6403174472759772"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# cool, it works\n",
    "bestwinrate(0, 100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Exercises for Part 2\n",
    "\n",
    "Use the documentation at https://docs.julialang.org/en/v1/ to help you with these exercises.\n",
    "\n",
    "1. Compute the win rate exactly using rational numbers.\n",
    "  - Hint: Replace `1.0` with `1//1`, `0.0` with `0//1`, and all four instances of `/ 2` with `// 2`. `//` is the rational division operator in Julia.\n",
    "  - You will probably see this error: `OverflowError: 4611686018427387904 * 2 overflowed for type Int64`. Here is the first gotcha. The default integer type in Julia is `Int`, which is (probably) a 64-bit signed integer on your machine. These can overflow. In many cases, this is silent. Rational numbers are kind enough to warn you with an error when this happens.\n",
    "  - To fix this issue, replace `1//1` with `BigInt(1)//1` and `0//1` with `BigInt(0)//1`. `BigInt(1)` is the same as `1` except that it is a `BigInt`, which means operations on it will not overflow.\n",
    "2. Show that the smart strategy is actually optimal by computing its probability exactly.\n",
    "  - Hint: You can do this with a function that looks a lot like `bestwinrate`, except instead of taking the best option, just take the option prescribed by the smart strategy (A iff positive)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "This slide is skipped in the slideshow. Since you are diligent enough to follow through the notebook, perhaps you could try some additional exercises 😄.\n",
    "\n",
    "## More exercises\n",
    "\n",
    "1. Write your own memoization. Hint: Look up Dict in the documentation.\n",
    "2. Write the function without explicit memoization. You can a two-dimensional array (https://docs.julialang.org/en/v1/manual/arrays/#man-multi-dim-arrays-1) to hold the win rates. Often this technique is called [dynamic programming](https://en.wikipedia.org/wiki/Dynamic_programming) (Wikipedia article linked).\n",
    "3. Modify the function to record which coin, A or B, is actually chosen (i.e. is the max win rate) for each pair of score/flips. Use a dict or two-dimensional array to record this information. (You can resolve ties arbitrarily. Alternatively, record when a tie happens, i.e. when choosing to flip A or B are equally optimal.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Solution to Exercise 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "811698796376000066208208781649//1267650600228229401496703205376"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using Memoize\n",
    "\n",
    "@memoize function bestwinrate(score, flips)\n",
    "    # Base case\n",
    "    if flips == 0\n",
    "        # No flips left. Our win rate is 1.0 if we won, 0.0 otherwise.\n",
    "        score > 0 ? BigInt(1)//1 : BigInt(0)//1\n",
    "    else\n",
    "        # win rate if we choose to flip A\n",
    "        winrateA = (\n",
    "            bestwinrate(score + 1, flips - 1) // 2\n",
    "          + bestwinrate(score - 1, flips - 1) // 2\n",
    "        )\n",
    "        # win rate if we choose to flip B\n",
    "        winrateB = (\n",
    "            bestwinrate(score + 2, flips - 1) // 2\n",
    "          + bestwinrate(score - 2, flips - 1) // 2\n",
    "        )\n",
    "        # pick the best of those two\n",
    "        max(winrateA, winrateB)\n",
    "    end\n",
    "end\n",
    "\n",
    "bestwinrate(0, 100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Solution to Exercise 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "811698796376000066208208781649//1267650600228229401496703205376"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using Memoize\n",
    "\n",
    "@memoize function smartwinrate(score, flips)\n",
    "    # Base case\n",
    "    if flips == 0\n",
    "        # No flips left. Our win rate is 1.0 if we won, 0.0 otherwise.\n",
    "        score > 0 ? BigInt(1)//1 : BigInt(0)//1\n",
    "    else\n",
    "        # win rate if we choose to flip A\n",
    "        winrateA = (\n",
    "            smartwinrate(score + 1, flips - 1) // 2\n",
    "          + smartwinrate(score - 1, flips - 1) // 2\n",
    "        )\n",
    "        # win rate if we choose to flip B\n",
    "        winrateB = (\n",
    "            smartwinrate(score + 2, flips - 1) // 2\n",
    "          + smartwinrate(score - 2, flips - 1) // 2\n",
    "        )\n",
    "        # pick the option prescribed by the smart strategy\n",
    "        score > 0 ? winrateA : winrateB\n",
    "    end\n",
    "end\n",
    "\n",
    "smartwinrate(0, 100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Part 3: A Quick More Theoretical Discussion...\n",
    "\n",
    "You might have heard that Julia walks like Python but runs like C..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "**How?**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Types\n",
    "\n",
    "In a lot of dynamic languages it is convenient to think about values and forget about types. (This is a large reason why Python / R / etc. are popular for rapid iteration in data science, compared to C++ / Java.)\n",
    "\n",
    "But in real programming, we often have to think about types.\n",
    "\n",
    "In Julia, types play a few important roles:\n",
    "\n",
    "- Functions in Julia are all multimethods, i.e. you can define multiple methods of a function for a variety of types. This allows for polymorphism which replaces the object-oriented paradigm in many other languages.\n",
    "- User-defined types (`struct`s, `mutable struct`s, etc.) allow the user to abstract and encapsulate data. This replaces the `class` from object-oriented languages.\n",
    "- Types are used by the compiler to make code efficient. The compiler is pretty good at type inference, but only if the types are actually inferrable (more on that later)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- **If you are not comfortable with types you can still learn and use Julia.**\n",
    "- But if you want to be very productive with Julia you will need to learn the type system. (Takes practice + some reading. See end of slides.)\n",
    "- Types are much more important in Julia than in Python / R / MATLAB / etc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Efficiency\n",
    "\n",
    "- Three important things to remember when writing efficient Julia code:\n",
    "  - Use efficient algorithms\n",
    "  - Use functions\n",
    "  - Check for type stability\n",
    "\n",
    "- What is type stability? Julia is a dynamic language and variables are allowed to change types. But this is not efficient, because the compiler will need to insert type checks every time you use the variable.\n",
    "\n",
    "- In a type stable function, the return type is a function of the input types, not of the input values.\n",
    "\n",
    "- *Important*: Contrary to popular belief, the presense of type annotations is not necessarily correlated to efficiency. (The compiler is usually able to figure out types itself. More on this soon.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# This function is type-stable (for all sane numeric / matrix types)\n",
    "# It will be efficient, even if called on a wide variety of input types.\n",
    "# For each input type, there will be native code specialized for that type.\n",
    "square1(x) = x * x\n",
    "\n",
    "# This function is not always type stable, even for integer types\n",
    "# The type annotation `::Integer` includes e.g. `Int`, `BigInt`, others, but\n",
    "# it does _not_ improve the compiler’s ability to infer types because it is on an input argument.\n",
    "# (the compiler will already specialize on the argument types)\n",
    "function square2(x::Integer)\n",
    "    if x < 0\n",
    "        return square2(-x)\n",
    "    end\n",
    "    result = 0\n",
    "    for i = 1:x\n",
    "        result += x\n",
    "    end\n",
    "    result\n",
    "end\n",
    "\n",
    "# test the code\n",
    "@assert square1(15) == square2(15) == 225"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Use `@code_warntype` to check for type instability. For `Int`s both functions are fine, but for `BigInt(1)`..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variables\n",
      "  #self#\u001b[36m::Core.Compiler.Const(square1, false)\u001b[39m\n",
      "  x\u001b[36m::BigInt\u001b[39m\n",
      "\n",
      "Body\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m1 ─\u001b[39m %1 = (x * x)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m└──\u001b[39m      return %1\n"
     ]
    }
   ],
   "source": [
    "@code_warntype square1(big\"1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variables\n",
      "  #self#\u001b[36m::Core.Compiler.Const(square2, false)\u001b[39m\n",
      "  x\u001b[36m::BigInt\u001b[39m\n",
      "  result\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "  @_4\u001b[33m\u001b[1m::Union{Nothing, Tuple{BigInt,BigInt}}\u001b[22m\u001b[39m\n",
      "  i\u001b[36m::BigInt\u001b[39m\n",
      "\n",
      "Body\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "\u001b[90m1 ─\u001b[39m       Core.NewvarNode(:(result))\n",
      "\u001b[90m│  \u001b[39m       Core.NewvarNode(:(@_4))\n",
      "\u001b[90m│  \u001b[39m %3  = (x < 0)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #3 if not %3\n",
      "\u001b[90m2 ─\u001b[39m %5  = -x\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %6  = Main.square2(%5)\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       return %6\n",
      "\u001b[90m3 ─\u001b[39m       (result = 0)\n",
      "\u001b[90m│  \u001b[39m %9  = (1:x)\u001b[36m::UnitRange{BigInt}\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (@_4 = Base.iterate(%9))\n",
      "\u001b[90m│  \u001b[39m %11 = (@_4 === nothing)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %12 = Base.not_int(%11)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #6 if not %12\n",
      "\u001b[90m4 ┄\u001b[39m %14 = @_4::Tuple{BigInt,BigInt}\u001b[36m::Tuple{BigInt,BigInt}\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (i = Core.getfield(%14, 1))\n",
      "\u001b[90m│  \u001b[39m %16 = Core.getfield(%14, 2)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (result = result + x)\n",
      "\u001b[90m│  \u001b[39m       (@_4 = Base.iterate(%9, %16))\n",
      "\u001b[90m│  \u001b[39m %19 = (@_4 === nothing)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %20 = Base.not_int(%19)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #6 if not %20\n",
      "\u001b[90m5 ─\u001b[39m       goto #4\n",
      "\u001b[90m6 ┄\u001b[39m       return result\n"
     ]
    }
   ],
   "source": [
    "@code_warntype square2(big\"1\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- Problem: `square2` will start with `result` having value `0::Int`, i.e. the `0` that is an `Int`.\n",
    "- But when a normal `Int` is added to a `BigInt`, we get a `BigInt` result.\n",
    "- So given `BigInt` input, we get `BigInt` output if the input is not a nonzero `BigInt`, but `Int` output if the input is `big\"0\"`.\n",
    "- Not type-stable because return type is a function of input _values_ not just input _types_.\n",
    "- How to fix it?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# square3 is the same algorithm as square2, but it’s type-stable because we have annotated\n",
    "# the initial value of `result`. The compiler will convert the `0::Int` to the appropriate `0`\n",
    "# of the same type as `x`.\n",
    "# If `result` is annotated like this, all assignments to `result` will be converted to the\n",
    "# annotated type.\n",
    "function square3(x::Integer)\n",
    "    if x < 0\n",
    "        return square2(-x)\n",
    "    end\n",
    "    result::typeof(x) = 0\n",
    "    for i = 1:x\n",
    "        result += x\n",
    "    end\n",
    "    result\n",
    "end\n",
    "\n",
    "@assert square3(15) == square2(15) == 225"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variables\n",
      "  #self#\u001b[36m::Core.Compiler.Const(square3, false)\u001b[39m\n",
      "  x\u001b[36m::BigInt\u001b[39m\n",
      "  result\u001b[36m::BigInt\u001b[39m\n",
      "  @_4\u001b[33m\u001b[1m::Union{Nothing, Tuple{BigInt,BigInt}}\u001b[22m\u001b[39m\n",
      "  i\u001b[36m::BigInt\u001b[39m\n",
      "\n",
      "Body\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "\u001b[90m1 ─\u001b[39m       Core.NewvarNode(:(result))\n",
      "\u001b[90m│  \u001b[39m       Core.NewvarNode(:(@_4))\n",
      "\u001b[90m│  \u001b[39m %3  = (x < 0)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #3 if not %3\n",
      "\u001b[90m2 ─\u001b[39m %5  = -x\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %6  = Main.square2(%5)\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       return %6\n",
      "\u001b[90m3 ─\u001b[39m %8  = Main.typeof(x)\u001b[36m::Core.Compiler.Const(BigInt, false)\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %9  = Base.convert(%8, 0)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (result = Core.typeassert(%9, %8))\n",
      "\u001b[90m│  \u001b[39m %11 = (1:x)\u001b[36m::UnitRange{BigInt}\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (@_4 = Base.iterate(%11))\n",
      "\u001b[90m│  \u001b[39m %13 = (@_4 === nothing)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %14 = Base.not_int(%13)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #6 if not %14\n",
      "\u001b[90m4 ┄\u001b[39m %16 = @_4::Tuple{BigInt,BigInt}\u001b[36m::Tuple{BigInt,BigInt}\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (i = Core.getfield(%16, 1))\n",
      "\u001b[90m│  \u001b[39m %18 = Core.getfield(%16, 2)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %19 = (result + x)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %20 = Main.typeof(x)\u001b[36m::Core.Compiler.Const(BigInt, false)\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %21 = Base.convert(%20, %19)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (result = Core.typeassert(%21, %20))\n",
      "\u001b[90m│  \u001b[39m       (@_4 = Base.iterate(%11, %18))\n",
      "\u001b[90m│  \u001b[39m %24 = (@_4 === nothing)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %25 = Base.not_int(%24)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #6 if not %25\n",
      "\u001b[90m5 ─\u001b[39m       goto #4\n",
      "\u001b[90m6 ┄\u001b[39m       return result\n"
     ]
    }
   ],
   "source": [
    "@code_warntype square3(big\"1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# square4 is the most natural way to write this code\n",
    "# zero(x) returns a zero of the same type as x\n",
    "# this is a little nicer than square3 because it does not require a type annotation\n",
    "function square4(x::Integer)\n",
    "    if x < 0\n",
    "        return square2(-x)\n",
    "    end\n",
    "    result = zero(x)\n",
    "    for i = 1:x\n",
    "        result += x\n",
    "    end\n",
    "    result\n",
    "end\n",
    "\n",
    "@assert square4(15) == square2(15) == 225"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Variables\n",
      "  #self#\u001b[36m::Core.Compiler.Const(square4, false)\u001b[39m\n",
      "  x\u001b[36m::BigInt\u001b[39m\n",
      "  result\u001b[36m::BigInt\u001b[39m\n",
      "  @_4\u001b[33m\u001b[1m::Union{Nothing, Tuple{BigInt,BigInt}}\u001b[22m\u001b[39m\n",
      "  i\u001b[36m::BigInt\u001b[39m\n",
      "\n",
      "Body\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "\u001b[90m1 ─\u001b[39m       Core.NewvarNode(:(result))\n",
      "\u001b[90m│  \u001b[39m       Core.NewvarNode(:(@_4))\n",
      "\u001b[90m│  \u001b[39m %3  = (x < 0)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #3 if not %3\n",
      "\u001b[90m2 ─\u001b[39m %5  = -x\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %6  = Main.square2(%5)\u001b[91m\u001b[1m::Union{Int64, BigInt}\u001b[22m\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       return %6\n",
      "\u001b[90m3 ─\u001b[39m       (result = Main.zero(x))\n",
      "\u001b[90m│  \u001b[39m %9  = (1:x)\u001b[36m::UnitRange{BigInt}\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (@_4 = Base.iterate(%9))\n",
      "\u001b[90m│  \u001b[39m %11 = (@_4 === nothing)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %12 = Base.not_int(%11)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #6 if not %12\n",
      "\u001b[90m4 ┄\u001b[39m %14 = @_4::Tuple{BigInt,BigInt}\u001b[36m::Tuple{BigInt,BigInt}\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (i = Core.getfield(%14, 1))\n",
      "\u001b[90m│  \u001b[39m %16 = Core.getfield(%14, 2)\u001b[36m::BigInt\u001b[39m\n",
      "\u001b[90m│  \u001b[39m       (result = result + x)\n",
      "\u001b[90m│  \u001b[39m       (@_4 = Base.iterate(%9, %16))\n",
      "\u001b[90m│  \u001b[39m %19 = (@_4 === nothing)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m│  \u001b[39m %20 = Base.not_int(%19)\u001b[36m::Bool\u001b[39m\n",
      "\u001b[90m└──\u001b[39m       goto #6 if not %20\n",
      "\u001b[90m5 ─\u001b[39m       goto #4\n",
      "\u001b[90m6 ┄\u001b[39m       return result\n"
     ]
    }
   ],
   "source": [
    "@code_warntype square4(big\"1\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Confusing? Probably...\n",
    "\n",
    "- If you have never used a static language before, type stability is not a natural concept.\n",
    "- Even if you _have_ used a static language before, having to think about type stability is not a natural concept.\n",
    "  - (it is usually enforced by the language)\n",
    "- If efficiency is not important no need to worry — but IMO is useful to try to grok this concept."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Personal Notes\n",
    "\n",
    "- It is not always straightforward to achieve efficiency. In practice, if you write code like you would in Python you often get Python-like performance. (That is fine sometimes / often.)\n",
    "- I needed to learn to recognize common sources of type instability and patch them up using idioms like `zero`. Came with practice and liberal use of `@code_warntype`.\n",
    "- Even in the worst case, it does solve the two-language problem. You can always write Julia like you would C and get C performance. (Personally I do not find C code fun to write, so I avoid this if possible. A lot of the code I write does not need to perform like C.)\n",
    "- So my advice: **Julia is not a magic wand that makes your dynamic code run faster.** What Julia does do is make it easy and natural to write what is essentially static code without the type annotation burden (thanks to multimethods and good type inference)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Part 4: Something different\n",
    "\n",
    "Goal: make the poster for this talk."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Well... actual goal:\n",
    "\n",
    "- Using libraries\n",
    "- IO\n",
    "- Side effects\n",
    "- Multimedia in the notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Using libraries\n",
    "\n",
    "Most of the time spent programming is actually just using code someone else wrote.\n",
    "\n",
    "Resources to find Julia packages:\n",
    "\n",
    "- https://pkg.julialang.org/docs/ (centralized package documentation)\n",
    "- Google* (seriously)\n",
    "\n",
    "\\* (personally I usually use DuckDuckGo)"
   ]
  },
  {
   "attachments": {
    "image3.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "In our case:\n",
    "\n",
    "![image3.png](attachment:image3.png)\n",
    "\n",
    "- 2D plotting does not seem appropriate\n",
    "- Luxor looks promising though!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "(Notebook recommended, since we are using multimedia, but the REPL will work... you just will have to open the files that will clutter your working directory)\n",
    "\n",
    "At REPL type `]add Luxor<RETURN>`. In notebook, run:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m\u001b[1m  Resolving\u001b[22m\u001b[39m package versions...\n",
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m `~/.julia/environments/v1.5/Project.toml`\n",
      "\u001b[90m [no changes]\u001b[39m\n",
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m `~/.julia/environments/v1.5/Manifest.toml`\n",
      "\u001b[90m [no changes]\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "using Pkg\n",
    "Pkg.add(\"Luxor\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "The first time you run `using Luxor`, it will precompile the package. This will take a while because `Luxor` has quite a few dependencies (and is a large package in itself). Luckily, this only needs to be done once every time a new version of the package (or a dependency) is installed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "using Luxor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "First, we will try to recreate the UW Data Science Club logo (so that we have a vector version).\n",
    "\n",
    "- I do not have the specific font used in the real logo installed (maybe you do not either).\n",
    "  - On my computer the default font seems to be Helvetica.\n",
    "- Hopefully this font is fairly ubiquitous; otherwise swap it for something you know you have.\n",
    "- We will worry about the placement later.\n",
    "  - For now we just use some sizes that are big enough to hold the stuff we are drawing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# We didn’t need to make this a function, but it will be more convenient for later.\n",
    "function dsclogo()\n",
    "    # `@layer` saves and restore the current color/font/scale/translation/etc.\n",
    "    # which let us change things like the font locally within the function,\n",
    "    # and restore the original settings afterward.\n",
    "    @layer begin\n",
    "        fontsize(40)\n",
    "        text(\"UW\", Point(0, 0))\n",
    "        text(\"DSC.\", Point(0, 35))\n",
    "        text(\"_\", Point(0, 70))\n",
    "    end\n",
    "end\n",
    "\n",
    "@png begin\n",
    "    dsclogo()\n",
    "end 200 200"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "This is not quite the right colour scheme. We need to use `sethue` and `box` to fix that.\n",
    "\n",
    "- The coordinates here are just trial and error.\n",
    "- You might need to adjust them if you have a different font.\n",
    "- `:fill` is a symbol: essentially a kind of string that is cheaper to compare, used frequently in APIs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "@png begin\n",
    "    box(Point(0, 0), Point(95, 131), :fill)\n",
    "    \n",
    "    @layer begin\n",
    "        translate(8, 43)\n",
    "        sethue(\"white\")\n",
    "        dsclogo()\n",
    "    end\n",
    "end 300 300"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Next we will draw the Julia logo, which Luxor provides a function to do.\n",
    "\n",
    "- These are **keyword arguments** which we have not seen so far.\n",
    "- They are basically like a named version of regular arguments. Keyword arguments are preferred when there are a lot of arguments, especially optional ones.\n",
    "- Actually, both these arguments are the default 😉. So the only point here is to show keyword arguments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "@png begin\n",
    "    scale(0.5)\n",
    "    julialogo(action=:fill, color=true)\n",
    "end 350 250"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "We need to move the logo by using a local coordinate system, through `@layer`.\n",
    "\n",
    "- Coordinates obtained through trial and error. If you changed the font you may need to change them.\n",
    "- Useful syntax: `...` (splat) to pass around for any number of arguments or keyword arguments.\n",
    "- Here we are using it to forward keyword arguments to another function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "positionedjulialogo (generic function with 1 method)"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Once again it is convenient to make this a function\n",
    "function positionedjulialogo(; kwargs...)\n",
    "    @layer begin\n",
    "        translate(65, 59)\n",
    "        scale(0.25)\n",
    "        julialogo(; kwargs...)\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "@png begin\n",
    "    box(Point(0, 0), Point(95, 131), :fill)\n",
    "    \n",
    "    @layer begin\n",
    "        translate(8, 43)\n",
    "        sethue(\"white\")\n",
    "        dsclogo()\n",
    "    end\n",
    "\n",
    "    positionedjulialogo()\n",
    "end 300 300"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "The black text cannot be read on a black background.\n",
    "\n",
    "- Workaround: draw the Julia logo once in white (clipped to the black box), and once in black.\n",
    "- Kind of a neat effect."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(Point(0.0, 0.0), Point(95.0, 131.0))"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Write the box coordinates and a function to draw a scaled julia logo once\n",
    "boxcoords = (Point(0, 0), Point(95, 131))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "@png begin\n",
    "    box(boxcoords..., :fill)\n",
    "\n",
    "    @layer begin\n",
    "        sethue(\"white\")\n",
    "        translate(8, 43)\n",
    "        dsclogo()\n",
    "    end\n",
    "    \n",
    "    positionedjulialogo(bodycolor=\"black\")\n",
    "\n",
    "    box(boxcoords..., :clip)\n",
    "    positionedjulialogo(bodycolor=\"white\")\n",
    "end 300 300"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Finally, we can add in the remaining text elements for the poster. Also, we will size the drawing appropriately for letter-sized paper. (This will also fix the weird centering arising from our choice of origin in the coordinate system.)\n",
    "\n",
    "- This size is provided through additional parameters to the `@svg` macro."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "@png begin\n",
    "    # set coordinates so (-70, -70) is top-left\n",
    "    # and so boxcoords[2].y + 70 is bottom of page (to leave some margin)\n",
    "    translate(-480, -355)\n",
    "    scale(710 / boxcoords[2].y)\n",
    "    box(boxcoords..., :fill)\n",
    "\n",
    "    @layer begin\n",
    "        sethue(\"white\")\n",
    "        translate(8, 43)\n",
    "        dsclogo()\n",
    "    end\n",
    "    \n",
    "    positionedjulialogo(bodycolor=\"black\")\n",
    "\n",
    "    sethue(\"red\")\n",
    "    box(boxcoords..., :clip)\n",
    "    positionedjulialogo(bodycolor=\"white\")\n",
    "    clipreset()\n",
    "\n",
    "    sethue(\"black\")\n",
    "    fontsize(18)\n",
    "    text(\"workshop\", Point(103, 123))\n",
    "    \n",
    "    fontsize(7)\n",
    "    text(\"MC 2065\", Point(103, 20))\n",
    "    text(\"Tue Feb 25 • 6–8 PM\", Point(103, 28))\n",
    "end 1100 850"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# since it is cut off on the previous slide...\n",
    "preview()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Some syntax\n",
    "\n",
    "- Macros: e.g. `@png`, `@layer`.\n",
    "  - Work like Scheme macros.\n",
    "  - Essentially allows package authors to create syntax extensions.\n",
    "- Symbols: `:fill`\n",
    "  - Work like Scheme symbols.\n",
    "  - They are essentially short strings that are *interned* (hashed so they only need a pointer-size of memory to pass around).\n",
    "- Splat operator: `...`\n",
    "  - May have seen in other languages: vararg, `*`/`**` in Python, etc.\n",
    "  - Allows handling arbitrary number of arguments / keyword arguments as a single unit.\n",
    "\n",
    "There are lots of small syntax things beyond the scope of this workshop. But the above are the most common across the ecosystem."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Part 5: Julia for Programming and Data Science\n",
    "\n",
    "Goal: run a tic tac toe AI tournament\n",
    "\n",
    "Will need:\n",
    "\n",
    "- types\n",
    "- multimethods\n",
    "- AI*\n",
    "\n",
    "\\* not machine learning, 😢"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Some notes before we begin\n",
    "\n",
    "Many languages are object-oriented. Julia is not.\n",
    "\n",
    "Philosophy: a type represents some semantic meaning to data. Instances of types are called objects. Objects do not generally contain functions / methods. Think of a type as a noun, e.g. `String`, `Int`, `Symbol`. User-defined types might be `Person` or `HTTPRequest` or `JavaScriptSyntaxTree`.\n",
    "\n",
    "Functions are _verbs_ that act on objects. A function should generally be a semantic operation, like `+`, and thus might have multiple implementations depending on the representation of the objects. For example `1 * 1` is not the same mechanical operation for the computer as `A::Matrix{Int} * B::Vector{Int}`, but they are both methods of function `*`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "**Contrast** with object-oriented view: objects contain both data and methods that operate on this data.\n",
    "\n",
    "In Julia objects contain just data (and some assigned meaning to it). Functions are not contained within objects; they are first-class concepts."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## A Plan\n",
    "\n",
    "1. Make a representation for the tic-tac-toe board.\n",
    "2. Make a interface for AIs to play tic-tac-toe.\n",
    "3. Make some AIs.\n",
    "4. Run a tournament.\n",
    "\n",
    "At any step feel free to extend with your own ideas!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Bitboard Representation\n",
    "\n",
    "- Computers store numbers internally as binary (e.g. 1000110, which is 70 in decimal)\n",
    "  - means that they are particularly efficient at operations like counting bits, bitwise and, bitwise or, etc.\n",
    "- A **bitboard** is, essentially, a number whose binary representation encodes some feature of squares of a board\n",
    "- Useful to save space and efficient for many operations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualization of what is going on in a bitboard\n",
    "using Luxor\n",
    "\n",
    "drawing = @png begin\n",
    "    translate(0, 50)\n",
    "\n",
    "    function tictactoeboard()\n",
    "        for i = 1:2\n",
    "            line(Point(50 * i, 0), Point(50 * i, 150), :stroke)\n",
    "            line(Point(0, 50 * i), Point(150, 50 * i), :stroke)\n",
    "        end\n",
    "    end\n",
    "    function glyph(i, c)\n",
    "        @layer begin\n",
    "            x = i % 3\n",
    "            y = i ÷ 3\n",
    "            fontface(\"Helvetica-Bold\")\n",
    "            fontsize(60)\n",
    "            text(c, Point(x * 50 + 11, y * 50 + 47))\n",
    "        end\n",
    "    end\n",
    "\n",
    "    xs = [0, 5, 7]\n",
    "    @layer begin\n",
    "        translate(-200, -200)\n",
    "        tictactoeboard()\n",
    "        for i = xs\n",
    "            glyph(i, \"×\")\n",
    "        end\n",
    "    end\n",
    "    @layer begin\n",
    "        translate(50, -200)\n",
    "        tictactoeboard()\n",
    "        for i = 0:8\n",
    "            @layer begin\n",
    "                if i in xs\n",
    "                    sethue(\"blue\")\n",
    "                end\n",
    "                glyph(i, string(i))\n",
    "            end\n",
    "        end\n",
    "    end\n",
    "    line(Point(-125, -25), Point(-125, 0), :stroke)\n",
    "    line(Point(-125, 0), Point(0, 0), :stroke)\n",
    "    line(Point(125, -25), Point(125, 0), :stroke)\n",
    "    line(Point(125, 0), Point(0, 0), :stroke)\n",
    "    arrow(Point(0, 0), Point(0, 25))\n",
    "    \n",
    "    @layer begin\n",
    "        translate(-length(xs) * 50 + 9, 50)\n",
    "        fontface(\"Noto Sans\")\n",
    "        fontsize(25)\n",
    "        for (ind, i) in enumerate(xs)\n",
    "            @layer begin\n",
    "                text(\"2\", Point(ind * 50, 0))\n",
    "                text(ind == length(xs) ? \"=\" : \"+\", Point(ind * 50 + 30, 0))\n",
    "                fontsize(17)\n",
    "                sethue(\"blue\")\n",
    "                text(string(i), Point(ind * 50 + 15, -11))\n",
    "            end\n",
    "        end\n",
    "        text(string(sum(2 .^ xs)), Point(length(xs) * 50 + 50, 0))\n",
    "        sethue(\"red\")\n",
    "        text(\"0b\", Point(75, 30))\n",
    "        sethue(\"black\")\n",
    "        text(string(sum(2 .^ xs), base=2), Point(105, 30))\n",
    "    end\n",
    "end 500 400"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "preview() # see notebook for code"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "We will use two bitboards, one for the X’s and one for the O’s.\n",
    "\n",
    "- `UInt16` (unsigned integer 16) includes 16 bits, which is bigger than the 9 required\n",
    "- We only have `UInt8`, `UInt16`, `UInt32`, `UInt64`, `UInt128` (powers of 2) so this is the best we can do"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TicTacToe"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "A tic tac toe board.\n",
    "\"\"\"\n",
    "struct TicTacToe\n",
    "    xs::UInt16\n",
    "    os::UInt16\n",
    "\n",
    "    TicTacToe() = new(0, 0)  # blank object\n",
    "\n",
    "    # this constructor will error if the provided state is \"invalid\",\n",
    "    # i.e. involves bits >= 10 or have overlap between xs and os.\n",
    "    # our definition of invalid is a little lax:\n",
    "    # we are fine with states that cannot occur in a real game,\n",
    "    # as long as they can be displayed like a tic tac toe board.\n",
    "    function TicTacToe(xs, os)\n",
    "        xs = UInt16(xs)\n",
    "        os = UInt16(os)\n",
    "        if xs & 0b111111111 != xs\n",
    "            throw(ArgumentError(\"xs contains invalid bits\"))\n",
    "        end\n",
    "        if os & 0b111111111 != os\n",
    "            throw(ArgumentError(\"os contains invalid bits\"))\n",
    "        end\n",
    "        if xs & os != 0b000000000\n",
    "            throw(ArgumentError(\"xs and os overlap\"))\n",
    "        end\n",
    "        new(xs, os)\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "That is it for the type. All other functionality will be provided through functions.\n",
    "\n",
    "- Provide a way to access the board without accessing fields xs and os (which for external code should be considered *private*)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Base.getindex"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import Base: getindex\n",
    "\n",
    "\"\"\"\n",
    "Return :X or :O if such a symbol is present in row i, column j (counting from 1), or :empty otherwise.\n",
    "\"\"\"\n",
    "function getindex(board::TicTacToe, i::Integer, j::Integer)\n",
    "    if i < 1 || i > 3 || j < 1 || j > 3\n",
    "        throw(BoundsError(board, (i, j)))\n",
    "    end\n",
    "    bit = 3(i - 1) + (j - 1)\n",
    "    if board.xs & (1 << bit) != 0\n",
    "        :X\n",
    "    elseif board.os & (1 << bit) != 0\n",
    "        :O\n",
    "    else\n",
    "        :empty\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "A simple test..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[37m\u001b[1mTest Summary:                           | \u001b[22m\u001b[39m\u001b[32m\u001b[1mPass  \u001b[22m\u001b[39m\u001b[36m\u001b[1mTotal\u001b[22m\u001b[39m\n",
      "board with Xs at (1,1), (2,3) and (3,2) | \u001b[32m  10  \u001b[39m\u001b[36m   10\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "using Test  # standard library\n",
    "board = TicTacToe(0b0010100001, 0)\n",
    "@testset \"board with Xs at (1,1), (2,3) and (3,2)\" begin\n",
    "    @test board[1, 1] == :X\n",
    "    @test board[1, 2] == :empty\n",
    "    @test board[1, 3] == :empty\n",
    "    @test board[2, 1] == :empty\n",
    "    @test board[2, 2] == :empty\n",
    "    @test board[2, 3] == :X\n",
    "    @test board[3, 1] == :empty\n",
    "    @test board[3, 2] == :X\n",
    "    @test board[3, 3] == :empty\n",
    "    @test_throws BoundsError board[4, 2]\n",
    "end;"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "- Provide a readable way to display the board"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "X┃ ┃ \n",
       "━╋━╋━\n",
       " ┃ ┃X\n",
       "━╋━╋━\n",
       " ┃X┃ \n"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import Base: show\n",
    "\n",
    "function show(io::IO, ::MIME\"text/plain\", board::TicTacToe)\n",
    "    for row in 1:3\n",
    "        for column in 1:3\n",
    "            entry = board[row, column]\n",
    "            if entry == :empty\n",
    "                print(io, \" \")\n",
    "            else\n",
    "                print(io, entry)\n",
    "            end\n",
    "            if column < 3\n",
    "                print(io, \"┃\")\n",
    "            end\n",
    "        end\n",
    "        println(io)\n",
    "        if row < 3\n",
    "            println(io, \"━╋━╋━\")\n",
    "        end\n",
    "    end\n",
    "end\n",
    "\n",
    "TicTacToe(0b0010100001, 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Now some basic rules...\n",
    "\n",
    "- `count_xs()`: number of X\n",
    "- `count_os()`: number of O\n",
    "- helper function `haswon(::UInt16)`: check if there is three in a row on individual bitboard\n",
    "\n",
    "Note: generating constants at the top level is pretty common in Julia (if you make your code a package, the top-level is precompiled, so the generation code only happens once \\[per package update\\])\n",
    "\n",
    "Lots of packages rely on this."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "haswon (generic function with 1 method)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count_xs(board::TicTacToe) = count_ones(board.xs)\n",
    "count_os(board::TicTacToe) = count_ones(board.os)\n",
    "\n",
    "const all_winning_patterns = UInt16[]\n",
    "for i = 0:2\n",
    "    push!(all_winning_patterns, sum(2 .^ (3i:3i+2)))  # row i+1\n",
    "    push!(all_winning_patterns, sum(2 .^ (i:3:6+i)))  # column i+1\n",
    "end\n",
    "push!(all_winning_patterns, sum(2 .^ (0:4:8)))  # diagonal \\\n",
    "push!(all_winning_patterns, sum(2 .^ (2:2:6)))  # diagonal /\n",
    "\n",
    "haswon(bb::UInt16) =\n",
    "    any(bb & pattern == pattern\n",
    "        for pattern in all_winning_patterns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "X┃X┃X\n",
       "━╋━╋━\n",
       " ┃ ┃ \n",
       "━╋━╋━\n",
       " ┃ ┃ \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "X┃ ┃ \n",
       "━╋━╋━\n",
       "X┃ ┃ \n",
       "━╋━╋━\n",
       "X┃ ┃ \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       " ┃ ┃ \n",
       "━╋━╋━\n",
       "X┃X┃X\n",
       "━╋━╋━\n",
       " ┃ ┃ \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       " ┃X┃ \n",
       "━╋━╋━\n",
       " ┃X┃ \n",
       "━╋━╋━\n",
       " ┃X┃ \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       " ┃ ┃ \n",
       "━╋━╋━\n",
       " ┃ ┃ \n",
       "━╋━╋━\n",
       "X┃X┃X\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       " ┃ ┃X\n",
       "━╋━╋━\n",
       " ┃ ┃X\n",
       "━╋━╋━\n",
       " ┃ ┃X\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "X┃ ┃ \n",
       "━╋━╋━\n",
       " ┃X┃ \n",
       "━╋━╋━\n",
       " ┃ ┃X\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       " ┃ ┃X\n",
       "━╋━╋━\n",
       " ┃X┃ \n",
       "━╋━╋━\n",
       "X┃ ┃ \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check: winning patterns are what we expect\n",
    "for pattern in all_winning_patterns\n",
    "    display(TicTacToe(pattern, 0))\n",
    "end"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "continued...\n",
    "\n",
    "- `status()`: `:X_wins`, `:O_wins`, `:Draw`, `:X_to_move`, `:O_to_move` or `:Invalid`\n",
    "- `isresolved()`: true or false (game is valid and done)\n",
    "- `xscore()`: 1, 0, 0.5, or error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "xscore (generic function with 1 method)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function status(board::TicTacToe)\n",
    "    xs = count_xs(board)\n",
    "    os = count_os(board)\n",
    "    xwon = haswon(board.xs)\n",
    "    owon = haswon(board.os)\n",
    "    \n",
    "    if xwon && owon || xs < os || xs > os + 1\n",
    "        :Invalid\n",
    "    elseif xwon\n",
    "        :X_wins\n",
    "    elseif owon\n",
    "        :O_wins\n",
    "    elseif xs == 5 && os == 4\n",
    "        :Draw\n",
    "    elseif xs > os\n",
    "        :O_to_move\n",
    "    else\n",
    "        @assert xs == os\n",
    "        :X_to_move\n",
    "    end\n",
    "end\n",
    "\n",
    "isresolved(board::TicTacToe) = status(board) in (:X_wins, :O_wins)\n",
    "\n",
    "const XSCOREMAP = Dict(:X_wins => 1.0, :O_wins => 0.0, :Draw => 0.5)\n",
    "xscore(board::TicTacToe) = XSCOREMAP[status(board)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "continued...\n",
    "\n",
    "- `moves()`: return vector of (row, column) pairs representing legal moves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[32m\u001b[1mTest Passed\u001b[22m\u001b[39m"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function moves(board::TicTacToe)\n",
    "    st = status(board)\n",
    "    if !(st in (:X_to_move, :O_to_move))\n",
    "        throw(ArgumentError(\"provided board is invalid or complete\"))\n",
    "    else\n",
    "        [(i, j) for i = 1:3 for j = 1:3 if board[i, j] == :empty]\n",
    "    end\n",
    "end\n",
    "                \n",
    "@test Set(moves(TicTacToe(0, 0))) == Set(Iterators.product(1:3, 1:3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "continued...\n",
    "\n",
    "- `play(board, row, column)`: return new board playing at (row, column) if allowed, error otherwise"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "play (generic function with 1 method)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "function play(board::TicTacToe, row::Integer, column::Integer)\n",
    "    st = status(board)\n",
    "    bit = 1 << (3(row - 1) + column - 1)\n",
    "    if board[row, column] == :empty\n",
    "        if st == :X_to_move\n",
    "            TicTacToe(bit | board.xs, board.os)\n",
    "        elseif st == :O_to_move\n",
    "            TicTacToe(board.xs, bit | board.os)\n",
    "        else\n",
    "            throw(ArgumentError(\"provided board is invalid or complete\"))\n",
    "        end\n",
    "    else\n",
    "        throw(ArgumentError(\"not a legal move\"))\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       " ┃ ┃ \n",
       "━╋━╋━\n",
       " ┃X┃ \n",
       "━╋━╋━\n",
       " ┃ ┃ \n"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "board = TicTacToe()\n",
    "board = play(board, 2, 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "O┃ ┃ \n",
       "━╋━╋━\n",
       " ┃X┃ \n",
       "━╋━╋━\n",
       " ┃ ┃ \n"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "board = play(board, 1, 1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Now we are ready to make some AIs. First we will define an interface and write a controller to pit two AIs against each other."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "vs (generic function with 1 method)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "An AI player for tic tac toe.\n",
    "\n",
    "AI players should implement:\n",
    "\n",
    "    selectmove(p::Player, board::Board) :: Tuple{Int, Int}\n",
    "\n",
    "which outputs the row and column of the selected move.\n",
    "\"\"\"\n",
    "abstract type AIPlayer end\n",
    "\n",
    "function vs(x::AIPlayer, o::AIPlayer)\n",
    "    board = TicTacToe()\n",
    "    st = status(board)\n",
    "    while st in (:X_to_move, :O_to_move)\n",
    "        player = st == :X_to_move ? x : o\n",
    "        board = play(board, selectmove(player, board)...)\n",
    "        st = status(board)\n",
    "    end\n",
    "    board, st\n",
    "end"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Next test it with a fully random AI:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "X┃X┃O\n",
       "━╋━╋━\n",
       "O┃X┃O\n",
       "━╋━╋━\n",
       "X┃O┃X\n"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "struct RandomPlayer <: AIPlayer end\n",
    "selectmove(::RandomPlayer, board::TicTacToe) = rand(moves(board))\n",
    "\n",
    "finalboard, st = vs(RandomPlayer(), RandomPlayer())\n",
    "finalboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       ":X_wins"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "st"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Exercise: Write an AI that picks the first move (in the order generated by `moves`). Hint: use the `first` function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "X┃O┃X\n",
       "━╋━╋━\n",
       "O┃X┃O\n",
       "━╋━╋━\n",
       "X┃ ┃ \n"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "struct LeftmostTopmostPlayer <: AIPlayer end\n",
    "selectmove(::LeftmostTopmostPlayer, board::TicTacToe) = first(moves(board))\n",
    "\n",
    "finalboard, st = vs(LeftmostTopmostPlayer(), LeftmostTopmostPlayer())\n",
    "finalboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       ":X_wins"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "st"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Next: Write a game-theoretically optimal AI.\n",
    "\n",
    "Hint: Remember what we did with the math problem earlier... we can do something very similar here.\n",
    "\n",
    "First write a function to calculate the game-theoretic score, which is 1 if X can force a win, 0 if O can force a win, and 0.5 otherwise."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[32m\u001b[1mTest Passed\u001b[22m\u001b[39m"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using Memoize\n",
    "\n",
    "@memoize function gtscore(board::TicTacToe)\n",
    "    st = status(board)\n",
    "    if st == :X_wins\n",
    "        1.0\n",
    "    elseif st == :O_wins\n",
    "        0.0\n",
    "    elseif st == :Draw\n",
    "        0.5\n",
    "    elseif st == :X_to_move\n",
    "        maximum(gtscore(play(board, move...)) for move in moves(board))\n",
    "    elseif st == :O_to_move\n",
    "        minimum(gtscore(play(board, move...)) for move in moves(board))\n",
    "    else\n",
    "        throw(ArgumentError(\"invalid position\"))\n",
    "    end\n",
    "end\n",
    "\n",
    "@test gtscore(TicTacToe()) == 0.5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Now relatively easy to write a GTO player.\n",
    "\n",
    "- We can also write a bad player that tries its best to lose.\n",
    "- (Not the same as an optimally bad player, because `gtscore` assumes perfect good play.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "X┃O┃O\n",
       "━╋━╋━\n",
       "X┃X┃O\n",
       "━╋━╋━\n",
       "X┃ ┃ \n"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "struct GTOPlayer <: AIPlayer\n",
    "    selectworst::Bool\n",
    "end\n",
    "\n",
    "function selectmove(g::GTOPlayer, board::TicTacToe)\n",
    "    st = status(board)\n",
    "    @assert st in (:X_to_move, :O_to_move)\n",
    "    # selectworst (exclusive) OR playing O, then minimize score\n",
    "    # if both, or neither, then maximize score\n",
    "    argextremum = xor(g.selectworst, st == :O_to_move) ? argmin : argmax\n",
    "    candidates = moves(board)\n",
    "    ind = argextremum([gtscore(play(board, move...)) for move in candidates])\n",
    "    candidates[ind]\n",
    "end\n",
    "\n",
    "finalboard, st = vs(GTOPlayer(false), GTOPlayer(true))\n",
    "finalboard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       ":X_wins"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "st"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "How good relatively are the AIs?\n",
    "\n",
    "- We can look at average score over some number of simulations of each pairing.\n",
    "- X has an advantage, so we want to consider average score of the ordered pairs.\n",
    "- Prefer to display this information as a table.\n",
    "\n",
    "Idea: Use `DataFrames` package for nicely formatted tables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\u001b[32m\u001b[1m  Resolving\u001b[22m\u001b[39m package versions...\n",
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m `~/.julia/environments/v1.5/Project.toml`\n",
      "\u001b[90m [no changes]\u001b[39m\n",
      "\u001b[32m\u001b[1m   Updating\u001b[22m\u001b[39m `~/.julia/environments/v1.5/Manifest.toml`\n",
      "\u001b[90m [no changes]\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "using Pkg\n",
    "Pkg.add(\"DataFrames\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "using DataFrames\n",
    "\n",
    "players = [\n",
    "    \"random\" => RandomPlayer(),\n",
    "    \"leftmost-topmost\" => LeftmostTopmostPlayer(),\n",
    "    \"gto\" => GTOPlayer(false),\n",
    "    \"bad-gto\" => GTOPlayer(true),\n",
    "]\n",
    "\n",
    "games = DataFrame(xplayer = String[], oplayer = String[], result = Float64[])\n",
    "for (xname, xplayer) in players\n",
    "    for (oname, oplayer) in players\n",
    "        for i = 1:10000\n",
    "            board, _ = vs(xplayer, oplayer)\n",
    "            push!(games, (xname, oname, xscore(board)))\n",
    "        end\n",
    "    end\n",
    "end"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"data-frame\"><thead><tr><th></th><th>xplayer</th><th>oplayer</th><th>result</th></tr><tr><th></th><th>String</th><th>String</th><th>Float64</th></tr></thead><tbody><p>160,000 rows × 3 columns</p><tr><th>1</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>2</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>3</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>4</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>5</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>6</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>7</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>8</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>9</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>10</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>11</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>12</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>13</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>14</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>15</th><td>random</td><td>random</td><td>0.5</td></tr><tr><th>16</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>17</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>18</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>19</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>20</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>21</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>22</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>23</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>24</th><td>random</td><td>random</td><td>0.5</td></tr><tr><th>25</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>26</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>27</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>28</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>29</th><td>random</td><td>random</td><td>0.0</td></tr><tr><th>30</th><td>random</td><td>random</td><td>1.0</td></tr><tr><th>&vellip;</th><td>&vellip;</td><td>&vellip;</td><td>&vellip;</td></tr></tbody></table>"
      ],
      "text/latex": [
       "\\begin{tabular}{r|ccc}\n",
       "\t& xplayer & oplayer & result\\\\\n",
       "\t\\hline\n",
       "\t& String & String & Float64\\\\\n",
       "\t\\hline\n",
       "\t1 & random & random & 1.0 \\\\\n",
       "\t2 & random & random & 1.0 \\\\\n",
       "\t3 & random & random & 1.0 \\\\\n",
       "\t4 & random & random & 1.0 \\\\\n",
       "\t5 & random & random & 0.0 \\\\\n",
       "\t6 & random & random & 1.0 \\\\\n",
       "\t7 & random & random & 1.0 \\\\\n",
       "\t8 & random & random & 1.0 \\\\\n",
       "\t9 & random & random & 1.0 \\\\\n",
       "\t10 & random & random & 0.0 \\\\\n",
       "\t11 & random & random & 1.0 \\\\\n",
       "\t12 & random & random & 0.0 \\\\\n",
       "\t13 & random & random & 0.0 \\\\\n",
       "\t14 & random & random & 0.0 \\\\\n",
       "\t15 & random & random & 0.5 \\\\\n",
       "\t16 & random & random & 0.0 \\\\\n",
       "\t17 & random & random & 1.0 \\\\\n",
       "\t18 & random & random & 1.0 \\\\\n",
       "\t19 & random & random & 1.0 \\\\\n",
       "\t20 & random & random & 0.0 \\\\\n",
       "\t21 & random & random & 1.0 \\\\\n",
       "\t22 & random & random & 1.0 \\\\\n",
       "\t23 & random & random & 1.0 \\\\\n",
       "\t24 & random & random & 0.5 \\\\\n",
       "\t25 & random & random & 1.0 \\\\\n",
       "\t26 & random & random & 1.0 \\\\\n",
       "\t27 & random & random & 1.0 \\\\\n",
       "\t28 & random & random & 1.0 \\\\\n",
       "\t29 & random & random & 0.0 \\\\\n",
       "\t30 & random & random & 1.0 \\\\\n",
       "\t$\\dots$ & $\\dots$ & $\\dots$ & $\\dots$ \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/plain": [
       "160000×3 DataFrame\n",
       "│ Row    │ xplayer │ oplayer │ result  │\n",
       "│        │ \u001b[90mString\u001b[39m  │ \u001b[90mString\u001b[39m  │ \u001b[90mFloat64\u001b[39m │\n",
       "├────────┼─────────┼─────────┼─────────┤\n",
       "│ 1      │ random  │ random  │ 1.0     │\n",
       "│ 2      │ random  │ random  │ 1.0     │\n",
       "│ 3      │ random  │ random  │ 1.0     │\n",
       "│ 4      │ random  │ random  │ 1.0     │\n",
       "│ 5      │ random  │ random  │ 0.0     │\n",
       "│ 6      │ random  │ random  │ 1.0     │\n",
       "│ 7      │ random  │ random  │ 1.0     │\n",
       "│ 8      │ random  │ random  │ 1.0     │\n",
       "│ 9      │ random  │ random  │ 1.0     │\n",
       "│ 10     │ random  │ random  │ 0.0     │\n",
       "⋮\n",
       "│ 159990 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159991 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159992 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159993 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159994 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159995 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159996 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159997 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159998 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 159999 │ bad-gto │ bad-gto │ 1.0     │\n",
       "│ 160000 │ bad-gto │ bad-gto │ 1.0     │"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "games"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"data-frame\"><thead><tr><th></th><th>xplayer</th><th>oplayer</th><th>result_mean</th></tr><tr><th></th><th>String</th><th>String</th><th>Float64</th></tr></thead><tbody><p>16 rows × 3 columns</p><tr><th>1</th><td>random</td><td>random</td><td>0.6559</td></tr><tr><th>2</th><td>random</td><td>leftmost-topmost</td><td>0.5342</td></tr><tr><th>3</th><td>random</td><td>gto</td><td>0.09715</td></tr><tr><th>4</th><td>random</td><td>bad-gto</td><td>0.9146</td></tr><tr><th>5</th><td>leftmost-topmost</td><td>random</td><td>0.80455</td></tr><tr><th>6</th><td>leftmost-topmost</td><td>leftmost-topmost</td><td>1.0</td></tr><tr><th>7</th><td>leftmost-topmost</td><td>gto</td><td>0.0</td></tr><tr><th>8</th><td>leftmost-topmost</td><td>bad-gto</td><td>1.0</td></tr><tr><th>9</th><td>gto</td><td>random</td><td>0.99825</td></tr><tr><th>10</th><td>gto</td><td>leftmost-topmost</td><td>1.0</td></tr><tr><th>11</th><td>gto</td><td>gto</td><td>0.5</td></tr><tr><th>12</th><td>gto</td><td>bad-gto</td><td>1.0</td></tr><tr><th>13</th><td>bad-gto</td><td>random</td><td>0.33055</td></tr><tr><th>14</th><td>bad-gto</td><td>leftmost-topmost</td><td>0.0</td></tr><tr><th>15</th><td>bad-gto</td><td>gto</td><td>0.0</td></tr><tr><th>16</th><td>bad-gto</td><td>bad-gto</td><td>1.0</td></tr></tbody></table>"
      ],
      "text/latex": [
       "\\begin{tabular}{r|ccc}\n",
       "\t& xplayer & oplayer & result\\_mean\\\\\n",
       "\t\\hline\n",
       "\t& String & String & Float64\\\\\n",
       "\t\\hline\n",
       "\t1 & random & random & 0.6559 \\\\\n",
       "\t2 & random & leftmost-topmost & 0.5342 \\\\\n",
       "\t3 & random & gto & 0.09715 \\\\\n",
       "\t4 & random & bad-gto & 0.9146 \\\\\n",
       "\t5 & leftmost-topmost & random & 0.80455 \\\\\n",
       "\t6 & leftmost-topmost & leftmost-topmost & 1.0 \\\\\n",
       "\t7 & leftmost-topmost & gto & 0.0 \\\\\n",
       "\t8 & leftmost-topmost & bad-gto & 1.0 \\\\\n",
       "\t9 & gto & random & 0.99825 \\\\\n",
       "\t10 & gto & leftmost-topmost & 1.0 \\\\\n",
       "\t11 & gto & gto & 0.5 \\\\\n",
       "\t12 & gto & bad-gto & 1.0 \\\\\n",
       "\t13 & bad-gto & random & 0.33055 \\\\\n",
       "\t14 & bad-gto & leftmost-topmost & 0.0 \\\\\n",
       "\t15 & bad-gto & gto & 0.0 \\\\\n",
       "\t16 & bad-gto & bad-gto & 1.0 \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/plain": [
       "16×3 DataFrame\n",
       "│ Row │ xplayer          │ oplayer          │ result_mean │\n",
       "│     │ \u001b[90mString\u001b[39m           │ \u001b[90mString\u001b[39m           │ \u001b[90mFloat64\u001b[39m     │\n",
       "├─────┼──────────────────┼──────────────────┼─────────────┤\n",
       "│ 1   │ random           │ random           │ 0.6559      │\n",
       "│ 2   │ random           │ leftmost-topmost │ 0.5342      │\n",
       "│ 3   │ random           │ gto              │ 0.09715     │\n",
       "│ 4   │ random           │ bad-gto          │ 0.9146      │\n",
       "│ 5   │ leftmost-topmost │ random           │ 0.80455     │\n",
       "│ 6   │ leftmost-topmost │ leftmost-topmost │ 1.0         │\n",
       "│ 7   │ leftmost-topmost │ gto              │ 0.0         │\n",
       "│ 8   │ leftmost-topmost │ bad-gto          │ 1.0         │\n",
       "│ 9   │ gto              │ random           │ 0.99825     │\n",
       "│ 10  │ gto              │ leftmost-topmost │ 1.0         │\n",
       "│ 11  │ gto              │ gto              │ 0.5         │\n",
       "│ 12  │ gto              │ bad-gto          │ 1.0         │\n",
       "│ 13  │ bad-gto          │ random           │ 0.33055     │\n",
       "│ 14  │ bad-gto          │ leftmost-topmost │ 0.0         │\n",
       "│ 15  │ bad-gto          │ gto              │ 0.0         │\n",
       "│ 16  │ bad-gto          │ bad-gto          │ 1.0         │"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "using Statistics  # standard library\n",
    "\n",
    "aggregate(games, [:xplayer, :oplayer], mean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"data-frame\"><thead><tr><th></th><th>xplayer</th><th>bad-gto</th><th>gto</th><th>leftmost-topmost</th><th>random</th></tr><tr><th></th><th>String</th><th>Float64⍰</th><th>Float64⍰</th><th>Float64⍰</th><th>Float64⍰</th></tr></thead><tbody><p>4 rows × 5 columns</p><tr><th>1</th><td>bad-gto</td><td>1.0</td><td>0.0</td><td>0.0</td><td>0.33055</td></tr><tr><th>2</th><td>gto</td><td>1.0</td><td>0.5</td><td>1.0</td><td>0.99825</td></tr><tr><th>3</th><td>leftmost-topmost</td><td>1.0</td><td>0.0</td><td>1.0</td><td>0.80455</td></tr><tr><th>4</th><td>random</td><td>0.9146</td><td>0.09715</td><td>0.5342</td><td>0.6559</td></tr></tbody></table>"
      ],
      "text/latex": [
       "\\begin{tabular}{r|ccccc}\n",
       "\t& xplayer & bad-gto & gto & leftmost-topmost & random\\\\\n",
       "\t\\hline\n",
       "\t& String & Float64⍰ & Float64⍰ & Float64⍰ & Float64⍰\\\\\n",
       "\t\\hline\n",
       "\t1 & bad-gto & 1.0 & 0.0 & 0.0 & 0.33055 \\\\\n",
       "\t2 & gto & 1.0 & 0.5 & 1.0 & 0.99825 \\\\\n",
       "\t3 & leftmost-topmost & 1.0 & 0.0 & 1.0 & 0.80455 \\\\\n",
       "\t4 & random & 0.9146 & 0.09715 & 0.5342 & 0.6559 \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/plain": [
       "4×5 DataFrame\n",
       "│ Row │ xplayer          │ bad-gto  │ gto      │ leftmost-topmost │ random   │\n",
       "│     │ \u001b[90mString\u001b[39m           │ \u001b[90mFloat64⍰\u001b[39m │ \u001b[90mFloat64⍰\u001b[39m │ \u001b[90mFloat64⍰\u001b[39m         │ \u001b[90mFloat64⍰\u001b[39m │\n",
       "├─────┼──────────────────┼──────────┼──────────┼──────────────────┼──────────┤\n",
       "│ 1   │ bad-gto          │ 1.0      │ 0.0      │ 0.0              │ 0.33055  │\n",
       "│ 2   │ gto              │ 1.0      │ 0.5      │ 1.0              │ 0.99825  │\n",
       "│ 3   │ leftmost-topmost │ 1.0      │ 0.0      │ 1.0              │ 0.80455  │\n",
       "│ 4   │ random           │ 0.9146   │ 0.09715  │ 0.5342           │ 0.6559   │"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unstack(ans, :xplayer, :oplayer, :result_mean)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Exercise\n",
    "\n",
    "Write a bot that plays in the first of the following list of squares that is available: `[(2, 2), (1, 1), (3, 3), (1, 3), (3, 1), (1, 2), (3, 2), (2, 1), (3, 2)]`, i.e. this bot does not actually look at the board at all but just tries to occupy the most important squares early.\n",
    "\n",
    "- Hint: look at help for the `intersect` function, which makes this a breeze\n",
    "- In my experience this bot does surprisingly well against random, even though it is essentially blind.\n",
    "- Is there a better order (for exploiting random)? I have not tried, but $9! = 362880$ can be brute forced."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "## More exercises (not in slideshow)\n",
    "\n",
    "1. Write a misère GTO player, i.e. a player that tries to minimize the score against best (also trying to lose) opposition optimally.\n",
    "  - This is not the same as `bad-gto`, which assumes opposition will try to win.\n",
    "2. Write an exploitative AI that has the best possible average score against the random player.\n",
    "  - Hint: modify `gtscore` logic so that the opponent is assumed to pick a random move, instead of picking the best move\n",
    "  - i.e. calculate expected value instead of minimum (if we are X) or instead of maximum (if we are O)\n",
    "3. Write a general exploitative AI that optimizes for win rate against a specific player through simulating some number of responses from that player (you can take the AI player being exploited as a parameter)\n",
    "  - Hint: you can `@memoize` the `selectmove` function so that it will be less slow to add this to your tournament"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "selectmove (generic function with 4 methods)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Here is a solution to 4.\n",
    "\n",
    "struct PrioritySquarePlayer <: AIPlayer\n",
    "    priorities::Vector{Tuple{Int, Int}}\n",
    "end\n",
    "# default\n",
    "PrioritySquarePlayer() = PrioritySquarePlayer(\n",
    "    [(2, 2), (1, 1), (3, 3), (1, 3), (3, 1), (1, 2), (3, 2), (2, 1), (2, 3)]\n",
    ")\n",
    "selectmove(p::PrioritySquarePlayer, board::TicTacToe) = first(intersect(p.priorities, moves(board)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Further Reading\n",
    "\n",
    "- Jeff (one of Julia co-authors) wrote his PhD thesis on the Julia type system. It is not 100% up to date, but the foundations and exposition are still relevant. (Warning: kind of technical) https://github.com/JeffBezanson/phdthesis/blob/master/main.pdf\n",
    "- Multimethods vs. function overloading: what is the difference? Excellent and brief Discourse post by Joshua Ballanco: https://discourse.julialang.org/t/is-multiple-dispatch-the-same-as-function-overloading/4145/5\n",
    "- \\[Little out of date as I wrote it for a pre-1.0 version of Julia, but I am happy to update it if anyone wants to run it on Julia 1.0+\\] my notebook on taking derivatives (good application of multimethods): https://wafy.me/documents/multdisp.html"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "Julia 1.5.0-DEV",
   "language": "julia",
   "name": "julia-1.5"
  },
  "language_info": {
   "file_extension": ".jl",
   "mimetype": "application/julia",
   "name": "julia",
   "version": "1.5.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}