16_vizualization.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Foundations of Computational Economics #16\n",
    "\n",
    "by Fedor Iskhakov, ANU\n",
    "\n",
    "<img src=\"_static/img/dag3logo.png\" style=\"width:256px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "## Visualization of data and solutions\n",
    "\n",
    "<img src=\"_static/img/lecture.png\" style=\"width:64px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=\"_static/img/youtube.png\" style=\"width:65px;\">\n",
    "\n",
    "[https://youtu.be/dJdWVkSNNpc](https://youtu.be/dJdWVkSNNpc)\n",
    "\n",
    "Description: Principles and functions of graphics. Examples of visualization of economic models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Why visualize?\n",
    "\n",
    "1. **Convey ideas to others**\n",
    "  Ability to efficiently explain your idea/work to other busy people is crucial element of success in many fields  \n",
    "1. **Check your own work**\n",
    "  Creating of new knowledge using computational tools requires absolute certainty in the code  \n",
    "1. **Aggregate large amounts of information**\n",
    "  Makes it possible to get the big picture and the message behind it  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/strava-heatmap-north-and-south-korea.png\" style=\"width:900px;\">\n",
    "\n",
    "[https://www.strava.com/heatmap#5.13/120.47297/37.61174/hot/all](https://www.strava.com/heatmap#5.13/120.47297/37.61174/hot/all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Matplotlib and other libraries\n",
    "\n",
    "- `matplotlib` - Python library that abstracts from the graphical backbone\n",
    "  of each system and ensures *code mobility*  \n",
    "\n",
    "\n",
    "**Matplotlib thumbnail gallery** [https://matplotlib.org/gallery.html](https://matplotlib.org/gallery.html)\n",
    "\n",
    "- `seaborn` - pretty plots geared towards statistical applications\n",
    "  [https://seaborn.pydata.org/examples/index.html](https://seaborn.pydata.org/examples/index.html)  \n",
    "- `bokeh` is a library for creating interactive plots\n",
    "  [http://bokeh.pydata.org/en/latest/docs/gallery.html](http://bokeh.pydata.org/en/latest/docs/gallery.html)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Plan\n",
    "\n",
    "1. Visualization examples from my own research projects  \n",
    "1. Links to compulsory online learning resources  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Graphical objects\n",
    "\n",
    "Extensive collection of objects to modify all aspects of the graphics\n",
    "\n",
    "- figure - axes - subplots  \n",
    "- lines - polygons (patches)  \n",
    "- fill color and edge color  \n",
    "- annotations and other text  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Types of plots\n",
    "\n",
    "- **bar** - categorical data, histograms  \n",
    "- **scatter** - individual data points  \n",
    "- **line** - continuous measure  \n",
    "- **area** - dynamics of composition  \n",
    "- **pie** - static composition  \n",
    "- **Sankey** - flow diagram  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### How to choose the plot type\n",
    "\n",
    "1. Number of variables to be represented  \n",
    "1. Type of variables  \n",
    "  - continuous  \n",
    "  - categorical  \n",
    "  - ordered  \n",
    "1. What is the message of the graphics?  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### General tips\n",
    "\n",
    "1. **Less visual clutter!**  \n",
    "  Every dot, line, shape and label has to convey useful information  \n",
    "1. **Read the the manual and change the options**  \n",
    "  Defaults are good for quick and dirty preliminary runs only  \n",
    "1. **Careful with 3D**  \n",
    "  Much harder to make clear  \n",
    "1. **“Animations”**  \n",
    "  May be useful in cases when there are one too many dimensions in the data to visualize  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Less visual clutter\n",
    "\n",
    "<img src=\"_static/img/viz/dcegm_value.png\" style=\"width:800px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Choice of appropriate plot type"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/cpu_vs_N.png\" style=\"width:800px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/plot1-super.png\" style=\"width:900px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/distribution_earnings.png\" style=\"width:800px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/octane95_price_composition.png\" style=\"width:900px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/waves_full_data.png\" style=\"width:900px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/map.jpg\" style=\"height:600px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/trpr2_diversification.png\" style=\"width:900px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/CDF_2.png\" style=\"width:700px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Using many dimensions in one plot\n",
    "\n",
    "- location (x,y,z)  \n",
    "- color  \n",
    "- line or marker style  \n",
    "- size  \n",
    "- animation  \n",
    "- multiple plots in a figure  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/eqb_n5_randtech-eps-converted-to.png\" style=\"width:900px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Visual debugging and dashboards\n",
    "\n",
    "Using visual representation to verify the code\n",
    "- *Seeing* a bug in a plot is easier than in the code!\n",
    "\n",
    "Dashboards are ideal for aggregation of large amounts of information\n",
    "\n",
    "1. Calibration/estimation an moment matching  \n",
    "1. Monitoring computing resources  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/phelps4.png\" style=\"width:900px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/ConsT3NumQuad_sigma0IncVar0.01.png\" style=\"width:800px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/fig_moments_lastrun.png\" style=\"width:1000px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/fig_moments2.png\" style=\"width:1000px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Visualizing economic model for new insights"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/5stage_eqb.png\" style=\"width:1200px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/eqb_map.png\" style=\"width:1000px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/eqb2bw_pureA.png\" style=\"width:1000px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=\"_static/img/viz/eqb2bw_symm.png\" style=\"width:1000px;\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Tutorials for compulsory self-study\n",
    "\n",
    "- Excellent tutorial on Matplotlib on QuantEcon DataScience\n",
    "  [https://datascience.quantecon.org/applications/visualization_rules.html](https://datascience.quantecon.org/applications/visualization_rules.html)  \n",
    "- Presentation by Hans Rosling (1948-2017, Swedish physician, academic, statistician, and public speaker)\n",
    "  [https://youtu.be/hVimVzgtD6w?t=159](https://youtu.be/hVimVzgtD6w?t=159)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Further learning resources\n",
    "\n",
    "- Excellent beginner tutorial for Matplotlib by the authors (3h)\n",
    "  [https://www.youtube.com/watch?v=6gdNUDs6QPc&t=2843s](https://www.youtube.com/watch?v=6gdNUDs6QPc&t=2843s)  \n",
    "- Playlist of lectures and tutorials\n",
    "  [https://www.youtube.com/user/EnthoughtMedia/search?query=matplotlib](https://www.youtube.com/user/EnthoughtMedia/search?query=matplotlib)  \n",
    "- Visualization of sorting algorithms\n",
    "  [https://www.youtube.com/watch?v=kPRA0W1kECg](https://www.youtube.com/watch?v=kPRA0W1kECg)  "
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "date": 1612589584.968545,
  "download_nb": false,
  "filename": "16_vizualization.rst",
  "filename_with_path": "16_vizualization",
  "kernelspec": {
   "display_name": "Python",
   "language": "python3",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "title": "Foundations of Computational Economics #16"
 },
 "nbformat": 4,
 "nbformat_minor": 4
}