{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Unit 12, Review\n",
    "====\n",
    "\n",
    "*Numerical Methods and Statistics*\n",
    "\n",
    "----\n",
    "\n",
    "#### Prof. Andrew White, April 25 2017"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "We've seen the following types of regression:\n",
    "\n",
    "1. Ordinary least squares with 1 x dimension and 1 y dimension (OLS-1D). Solve with equation. Lecture 1\n",
    "2. Ordinary least squares with N x dimensions and 1  y dimension (OLS-ND). Solve with matrix equation. Lecture 1\n",
    "3. Non-linear least squares with N x dimensions and 1 y dimension (NLS-ND). Solve with SSR minimization. Lecture 2\n",
    "4. Non-linear least squares with N x dimensions and N y dimension (NLS-ND-NY). Solve with SSR minimization. Lecture 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "We've seen the following changes we can make to a problem to make regression possible:\n",
    "\n",
    "1. Linearizing a problem, often polynomials, to make it an N dimensional regression (OSL-ND). Lecture 1\n",
    "2. Representing categories or yes/no data as binary variables. Lecture 3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Finally, we've seen the following modifications due to measurement error:\n",
    "\n",
    "1. OLS-1D with measurement error in x, y or both. Lecture 4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "We saw a sketch of a derivation for the error equations in lecture 2."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Now we'll look at some examples and choose which approach to take from the above types of regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Exampe 1\n",
    "---\n",
    "\n",
    "We have data on the US population below for the years 1998-2016. We would like to be able to predict it for the year 2017\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAc0AAAEtCAYAAABqATufAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nf/Hl/QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd80d9d94e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(pop_x, pop_y, 'o')\n",
    "plt.xlabel('Year')\n",
    "plt.ylabel('Population [Millions]')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Example 1 Answer\n",
    "----\n",
    "\n",
    "This is a simple OLS-1D problem. Although population growth is exponential, over this range it is linear enough. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Exmaple 2\n",
    "---\n",
    "\n",
    "We have data on selectivity in a membrane as a function of temperature and if we added graphene nanofibers to the membrane. Fit a model to describe the data. Selectivity is the number of moles of the desired product in the efflux over the number of moles of the undesired product. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Exmaple 2 Answer\n",
    "---\n",
    "\n",
    "This is a 2 dimensional problem: temperature and yes/no graphene fibers. This can be answered using OLS-ND. One more issue though is that we need an intercept. We would not expect selectivity to be 0 at absolute zero. Therefore our x-matrix will have the following columns: $[1, T, \\delta_{g}]$ where $\\delta_g$ indicates if graphene fibers are used. \n",
    "\n",
    "Here's an example of building this with real data. Let's say we did 5 experiments with and 5 experiments without graphene at a range of temperatures. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[   1.  280.    1.]\n",
      " [   1.  285.    1.]\n",
      " [   1.  290.    1.]\n",
      " [   1.  295.    1.]\n",
      " [   1.  300.    1.]\n",
      " [   1.  280.    0.]\n",
      " [   1.  285.    0.]\n",
      " [   1.  290.    0.]\n",
      " [   1.  295.    0.]\n",
      " [   1.  300.    0.]]\n"
     ]
    }
   ],
   "source": [
    "graphene_used = np.concatenate( (np.ones(5), np.zeros(5)) )\n",
    "temperature = np.concatenate( (T, T) )\n",
    "intercept = np.ones(10)\n",
    "\n",
    "x_mat = np.column_stack( (intercept, temperature, graphene_used) )\n",
    "print(x_mat)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Example 3\n",
    "---\n",
    "\n",
    "We have data on traffic fatalities per at each day of the year and whether or not it's raining. Create a model equation for the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Example 3 Answer\n",
    "---\n",
    "\n",
    "This is actually a very complex regression. Rather than try to regress against time, we could use the following category variables:\n",
    "\n",
    "1. Is it a holiday\n",
    "2. Is it raining\n",
    "3. Is it a weekend or weekday"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Example 4\n",
    "---\n",
    "\n",
    "Someone has proposed that conversion efficiency in a chemical reaction depends on the product of the temperature and concentration of one of the reagents. What equation could you use to represent this?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Example 4 Answer\n",
    "---\n",
    "\n",
    "We could take two approaches here. We could use non-linear regression over $T$, $c$ and an intercept. This would give an x-matrix of $[T, c, 1]$. OR, we could linearize it to the $T\\times c$ dimension. Then our x-matrix would be $[Tc, 1]$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Example 5\n",
    "---\n",
    "\n",
    "You are making a regression of reaction rate vs concentration of reactant 1. Through error propogation, you estimate that your reaction rate measurement has an uncertainty of 2.5 inverse seconds. What regression technique should you use?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Example 5 Answer\n",
    "----\n",
    "\n",
    "We do not know if it is linear or non-linear, but we will for sure have to use a measurement error technique. The error is in the dependent variable. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "#Ignore these\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('fivethirtyeight')\n",
    "import scipy.stats as ss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "pop_x = np.arange(1998, 2016)\n",
    "pop_y = 275.9 * np.exp((pop_x - 1998) * 0.005) + ss.norm.rvs(size=len(pop_x)) * 0.2\n",
    "T = np.arange(280, 280 + 5 * 5, 5)"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}