aicam/diversity_MSE.ipynb

## diversity_MSE.ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "\n",
    "\n",
    "# Number of mus as population number\n",
    "mu_count = 100\n",
    "# theta initialization\n",
    "theta = 10\n",
    "# variances that will be tested in each iteration\n",
    "var_range = np.arange(10, 300)\n",
    "# indicates how many times each population should generate new numbers\n",
    "confidence = 80\n",
    "# boundary of our distribution\n",
    "bins = np.arange(-100,100)\n",
    "\n",
    "diversity = []\n",
    "MSE = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "# generates a normal distribution histogram\n",
    "def generate_ideal_normal(sigma):\n",
    "    return np.array(1/(sigma * np.sqrt(2 * np.pi)) * np.exp( - (bins - theta)**2 / (2 * sigma**2)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [],
   "source": [
    "MSE = []\n",
    "for var in var_range:\n",
    "    SE = []\n",
    "    for _ in range(confidence):\n",
    "        # generate initial population by choosing random numbers with normal distribution \n",
    "        # (mean = theta, sigma = each variance in var_range)\n",
    "        mus = np.random.normal(theta, np.sqrt(var), mu_count)\n",
    "        \n",
    "        # each individual get a scalar value from a normal distribution \n",
    "        # (mean = theta, sigma = e^-|theta - mu for each mu of individual|)\n",
    "        mus_signals = np.array([np.random.normal(theta, np.exp(-1 * np.abs(theta - m))) for m in mus])#generate_ideal_normal(np.exp(-1 * np.abs(theta - m))) for m in mus])    \n",
    "        \n",
    "        # each squared error is the difference between the average of all individual signals and theta\n",
    "        SE.append((np.mean(mus_signals) - theta)**2)\n",
    "        \n",
    "    # generate new theta\n",
    "    theta = np.random.uniform(-100,100)\n",
    "    MSE.append(np.mean(SE))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f2f4bb8f278>]"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.xlabel('Diversity (Variance)')\n",
    "plt.ylabel('MSE')\n",
    "plt.plot(var_range, MSE, 'bo')"
   ]
  }
 ],
 "metadata": {
  "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}
	{
	"cells": [
	{
	"cell_type": "code",
	"execution_count": 129,
	"metadata": {},
	"outputs": [],
	"source": [
	"import numpy as np\n",
	"import matplotlib.pyplot as plt\n",
	"\n",
	"\n",
	"\n",
	"# Number of mus as population number\n",
	"mu_count = 100\n",
	"# theta initialization\n",
	"theta = 10\n",
	"# variances that will be tested in each iteration\n",
	"var_range = np.arange(10, 300)\n",
	"# indicates how many times each population should generate new numbers\n",
	"confidence = 80\n",
	"# boundary of our distribution\n",
	"bins = np.arange(-100,100)\n",
	"\n",
	"diversity = []\n",
	"MSE = []"
	]
	},
	{
	"cell_type": "code",
	"execution_count": 130,
	"metadata": {},
	"outputs": [],
	"source": [
	"# generates a normal distribution histogram\n",
	"def generate_ideal_normal(sigma):\n",
	" return np.array(1/(sigma * np.sqrt(2 * np.pi)) * np.exp( - (bins - theta)*2 / (2 sigma**2)))"
	]
	},
	{
	"cell_type": "code",
	"execution_count": 133,
	"metadata": {},
	"outputs": [],
	"source": [
	"MSE = []\n",
	"for var in var_range:\n",
	" SE = []\n",
	" for _ in range(confidence):\n",
	" # generate initial population by choosing random numbers with normal distribution \n",
	" # (mean = theta, sigma = each variance in var_range)\n",
	" mus = np.random.normal(theta, np.sqrt(var), mu_count)\n",
	" \n",
	" # each individual get a scalar value from a normal distribution \n",
	" # (mean = theta, sigma = e^-\|theta - mu for each mu of individual\|)\n",
	" mus_signals = np.array([np.random.normal(theta, np.exp(-1 * np.abs(theta - m))) for m in mus])#generate_ideal_normal(np.exp(-1 * np.abs(theta - m))) for m in mus]) \n",
	" \n",
	" # each squared error is the difference between the average of all individual signals and theta\n",
	" SE.append((np.mean(mus_signals) - theta)**2)\n",
	" \n",
	" # generate new theta\n",
	" theta = np.random.uniform(-100,100)\n",
	" MSE.append(np.mean(SE))"
	]
	},
	{
	"cell_type": "code",
	"execution_count": 134,
	"metadata": {},
	"outputs": [
	{
	"data": {
	"text/plain": [
	"[<matplotlib.lines.Line2D at 0x7f2f4bb8f278>]"
	]
	},
	"execution_count": 134,
	"metadata": {},
	"output_type": "execute_result"
	},
	{
	"data": {
	"image/png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAEGCAYAAABCa2PoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8GearUAAAgAElEQVR4nO2de7QlVX3nP7++/ZAriOF2xxCwH0gnpnEZonf5iEkmDjE8MiPJGpa2uZiOkOkJYCQ4k1mQnkmMazrRmTXJYBCZjkAIfSON+Egvh4j4SExG03AxgA2kscXuFsZA0/gA29jQ/Zs/apddt7rep+qce875ftaqdevsqtq1d51z97d++/fbe5u7I4QQQvTKokEXQAghxGggQRFCCNEKEhQhhBCtIEERQgjRChIUIYQQrbB40AUYJMuXL/fVq1cPuhhCCDFU3HPPPU+6+4p0+lgLyurVq5mbmxt0MYQQYqgws71Z6eryEkII0QoSFCGEEK0gQRFCCNEKEhQhhBCt0KmgmNk5ZrbLzHab2ZUZx5eZ2bZwfIeZrU4cuyqk7zKzsxPpN5jZE2a2M5XX/zCzfzKz+83sY2b2wi7rJoQQYj6dCYqZTQDvB84F1gFvMbN1qdMuBr7p7qcDfwK8N1y7DlgPnAGcA1wb8gP485CW5k7gZe7+cuBh4KpWK9Qjs7OwejUsWhT9nZ0ddImEEKJdurRQXgXsdvdH3P0QcAtwfuqc84Gbwv5twFlmZiH9Fnf/vrt/Ddgd8sPdPw88lb6Zu3/K3Z8LH/8BOLXtCjVldhY2boS9e8E9+rtxo0RFCDFadCkopwBfT3x+NKRlnhPE4NvAVMVri7gI+OusA2a20czmzGxu//79NbJszqZNcPDg/LSDB6N0IYQYFUbOKW9mm4DngMz3f3ff4u7T7j69YsUxAz07Yd++eulCCDGMdCkojwEvTnw+NaRlnmNmi4ETgQMVrz0GM/t14N8AM96HlcOSfpHly6Mty0eycmX29XnpQggxjHQpKHcDa81sjZktJXKyb0+dsx3YEPYvAD4bhGA7sD5Ega0B1gJ3Fd3MzM4B/jPwRnc/WHRuG6T9IgcORFuWj2TzZpicnH/95GSULoQQo0JnghJ8Im8H7gAeAm519wfM7N1m9sZw2vXAlJntBt4JXBmufQC4FXgQ+CRwmbsfBjCzDwFfBH7czB41s4tDXtcAJwB3mtm9ZnZdV3WDbL9IkqSPZGYGtmyBVavALPq7ZUuULoQQo4KN85ry09PT3nRyyEWLImukCDM4cqRR9kIIsWAxs3vcfTqdPnJO+X5Rxf8hH4kQYpyQoNQkdsTv3RtZIHnIRyKEGDfGej2UusSO+Nh34h6JijtMTUVpTz0VWSabN8tHIoQYLyQoNchyxLtHTvY9ewZSJCGEWDCoy6sGGqAohBD5SFBqoAGKQgiRjwSlBhqgKIQQ+UhQaqABikIIkY+c8jWZmZGACCFEFrJQhBBCtIIERQghRCtIUIQQQrSCBEUIIUQrSFCEEEK0ggRFCCFEK0hQhBBCtIIERQghRCtIUIQQQrSCBEUIIUQrSFCEEEK0ggRFCCFEK0hQhBBCtIIERQghRCtIUDpidhZWr4ZFi6K/s7ODLpEQQnSL1kPpgNlZ2LgRDh6MPu/dG30GraUihBhdZKF0wKZNR8Uk5uDBKF0IIUYVCUoH7NtXL10IIUYBCUoDyvwjK1dmX5eXLoQQo4AEpSaxf2TvXnA/6h9JisrmzTA5Of+6yckoXQghRpVOBcXMzjGzXWa228yuzDi+zMy2heM7zGx14thVIX2XmZ2dSL/BzJ4ws52pvE4yszvN7Cvh7w91Uacq/pGZGdiyBVatArPo75YtcsgLIUabzgTFzCaA9wPnAuuAt5jZutRpFwPfdPfTgT8B3huuXQesB84AzgGuDfkB/HlIS3Ml8Bl3Xwt8JnxunTw/yN6987u/ZmZgzx44ciT6KzERQow6XVoorwJ2u/sj7n4IuAU4P3XO+cBNYf824Cwzs5B+i7t/392/BuwO+eHunweeyrhfMq+bgF9uszIxRX6QrO4vIYQYF7oUlFOAryc+PxrSMs9x9+eAbwNTFa9N8yJ3/0bY/2fgRVknmdlGM5szs7n9+/dXqcc8svwjSRQeLIQYV0bSKe/uDnjOsS3uPu3u0ytWrKidd9I/ksfevbWzFUKIoadLQXkMeHHi86khLfMcM1sMnAgcqHhtmsfN7OSQ18nAE41LXkLsH8kTFTN1ewkhxo8uBeVuYK2ZrTGzpURO9u2pc7YDG8L+BcBng3WxHVgfosDWAGuBu0rul8xrA/BXLdShkM2bI/FI465uLyHE+NGZoASfyNuBO4CHgFvd/QEze7eZvTGcdj0wZWa7gXcSIrPc/QHgVuBB4JPAZe5+GMDMPgR8EfhxM3vUzC4Oeb0HeIOZfQX4hfC5U2ZmIvHIQqPihRDjhnleizgGTE9P+9zcXE95rF6d7TNZtSrqFhNCiFHDzO5x9+l0+kg65fuJRsULIUSEBKVHNCpeCCEiJCgtkBwVv3lz5JDXwlpCiHFDC2y1iBbWEkKMM7JQWkQLawkhxhkJSotoYS0hxDgjQWkRLawlhBhnJCgtohBiIcQ4I0FpSNYywAohFkKMMxKUBhQtA5y1sFbZGvRCCDEKSFAaUBbNlRSQ5cvhoouK16AXQohRQILSgKJorrT1cuAAHDo0/zyFEgshRhEJSgOKormyrJcs6oYSq9tMCLHQkaA0oCiaq6pQ1AklLvLZCCHEQkGC0oCiaK4qQlE3lFgj8IUQw4AEpSFZ0VyQbb0sWQJTU81DiTUCXwgxDEhQWibLernxRnjyyWPFpyoagS+EGAYkKB2QZ700RSPwhRDDgARlCNAIfCHEMKD1UIaEmRkJiBBiYSMLZQGgMSZCiFFAFsqA0SqPQohRQRbKgNEYEyHEqCBBGTAaYyKEGBUkKANGY0yEEKOCBGXAaIyJEGJUkKAMGI0xEUKMChKUATM7Gzng9+2Lurk2b5aYCCGGE4UNDxCFDAshRglZKANEIcNCiFGiU0Exs3PMbJeZ7TazKzOOLzOzbeH4DjNbnTh2VUjfZWZnl+VpZmeZ2ZfM7F4z+3szO73LulWhbAS8QoaFEKNEZ4JiZhPA+4FzgXXAW8xsXeq0i4FvuvvpwJ8A7w3XrgPWA2cA5wDXmtlESZ4fAGbc/UzgL4H/0lXdqlBllUWFDAshRokuLZRXAbvd/RF3PwTcApyfOud84KawfxtwlplZSL/F3b/v7l8Ddof8ivJ04AVh/0Tg/3VUr0pU6c5SyLAQYpTo0il/CvD1xOdHgVfnnePuz5nZt4GpkP4PqWtPCft5ef4GcLuZfQ/4DvCarEKZ2UZgI8DKDk2BKt1ZseN906bIgpmYmC86cswLIYaJUXLKXwGc5+6nAjcCf5x1krtvcfdpd59esWJFZ4Wp2p01M3PUUjl8OErL6h4TQoiFTpeC8hjw4sTnU0Na5jlmtpioq+pAwbWZ6Wa2AvhJd98R0rcBP91ONZpRpztL0V5CiFGgS0G5G1hrZmvMbCmRk3176pztwIawfwHwWXf3kL4+RIGtAdYCdxXk+U3gRDP7sZDXG4CHOqxbKXVGwCvaSwgxCnTmQwk+kbcDdwATwA3u/oCZvRuYc/ftwPXAzWa2G3iKSCAI590KPAg8B1zm7ocBsvIM6f8e+IiZHSESmIu6qltVkj6SffvyfSMrV0bdXGkU7SWEGCYsMgjGk+npaZ+bm+ss//RIeIi6vdKWStXzhBBiIWBm97j7dDp9lJzyC46qvhFNECmEGAVkoXRooSxaFA1qTGMGR450dlshhOgUWSgDQCPhhRDjhASlQzQSXggxTkhQOkS+ESHEOKH1UDpmZkYCIoQYD2ShCCGEaAUJygKnbE0VIYRYKKjLawGjJYKFEMNEoYViZhcm9l+XOvb2rgo1rqStkcsv16SRQojhoazL652J/T9NHRv4XFmjRNYKjwcOZJ+rSSOFEAuRMkGxnP2sz6IHsqZpyUMDI4UQC5EyQfGc/azPogeqWh0aGCmEWKiUOeVfamb3E1kjLwn7hM+ndVqyMSNvCvupKTj++EhwVq6MxEQOeSHEQqRMUH6iL6UYQ2Znj66TsnIlnHce3HTTsVPYX321BEQIMRwUdnm5+97kBjwDvAJYHj6LBmQ54G+6CTZs0DQtQojhpSxs+BNm9rKwfzKwkyi662Yz++0+lG8kyVsn5fbbYc+eaGr7PXuqiUk61PjSSzUQUggxGMqc8mvcfWfYfxtwp7v/W+DVKGy4NnHjn+UrgfrhwFmWzgc+MP/zxo0SFSFEfygTlGcT+2cBtwO4+9OAloiqQbLxz6NuOHCVUGMNhBRC9IsyQfm6mf2Wmf0Kke/kkwBmdhywpOvCjRJljb9Z/XDgqhbN3r3qAhNCdE+ZoFwMnAH8OvBmd/9WSH8NcGOH5Ro5yhp/93KfSdpfctJJ1e+vLjAhRNdoTfkO15RPUuQ7gSiqa8+e/OPpiSIBliyJLJtDh+qVpexeQghRRN6a8oXjUMxse9Fxd39jrwUbFzZvhre+NbIU0lTp7srqMnv22WMHPp53XhQttm9f9r1Ac4EJIbqh0EIxs/3A14EPATtIzd/l7n/baek6pp8WCkTCkUeZobhoUb4YHUmFR8SDJvMsIlkoQoheyLNQynwoPwL8LvAy4GrgDcCT7v63wy4mg2DVqmrpWYtq5UWAuc93tpdFk2kuMCFEV5SNlD/s7p909w1EjvjdwN9oLZRmbN4cNehJ0g181tiSjRujrqz0tTFJZ3tRNJlG3wshuqTUKW9my4BfAt4CrAa2Aze4+2Odl65j+t3lBcfO4ZWe7DHPeT8xAYcPH/2bxapV+b6TrK4xIYRoQl6XV5kP5S+IurtuB25JjJofCQYhKGXk+UqqYJY/a7H8JkKItmjqQ7kQWAtcDnzBzL4TtqfN7DtdFHTcqTO2JE1s8ZR1qwkhRBeU+VAWufsJYXtBYjvB3V9QlrmZnWNmu8xst5ldmXF8mZltC8d3mNnqxLGrQvouMzu7LE+L2GxmD5vZQ2b2jqoPYaEwOwvfaSjTsWjMzER+Es1aLIToN2XroTTGzCaA9xNFhj0K3G1m2939wcRpFwPfdPfTzWw98F7gzWa2DlhPNEr/R4FPm9mPhWvy8vx14MXAS939iJn9cFd1a4u0P+WZZ6KxJVWYmor+xuvOH3fc0WMzMxIQIUT/6UxQgFcBu939EQAzuwU4H0gKyvnAu8L+bcA1ZmYh/RZ3/z7wNTPbHfKjIM9LgF919yMA7v5Eh3XrmfTI96JR9GnMIiFJjms5cCDKDyQmQojBUOZD6YVTiAZFxjwa0jLPcffngG8DUwXXFuX5EiLrZs7M/trM1rZUj06oMlNwzNTU0bEqZked9mnn/cGD0SJdmghSCDEIuhSUfrMM+JcQefBnwA1ZJ5nZxiA6c/v37+9rAZNUnf4kXgZ4z55IVMoiwA4fLp8IMmvgpBBC9EqXgvIYkU8j5tSQlnmOmS0GTgQOFFxblOejwEfD/seAl2cVyt23uPu0u0+vWLGiZpXaI2/ke9IamZg4up7J7Gz9Obiy1kLJGzgpURFC9EqXgnI3sNbM1pjZUiIne3qyye3AhrB/AfBZjwbGbAfWhyiwNUShy3eV5Plx4PVh/18BD3dUr1bIC++9+uqjx+IBjHGj3ySkOC1CecsPaxEuIUSvdCYowSfyduAO4CHgVnd/wMzebWbxLMXXA1PB6f5O4Mpw7QPArUTO9k8Cl4VpYDLzDHm9B/h3ZvZl4I+A3+iqbm1QFN6b1+hD/vQreaTn+sqzctLp6hYTQtRF66EssJHyUDyz8M03zw81rhodNjkZCVbeLMTJkfRZa6/E1yuCTAjRdKS8GAB5/pWVK6MGfc+eaF6u2FFfhYMH4fLLo7EuadIj6XvpFpNlI8T4IkFZgNSZPiXr3DwOHDg6EDJmaupYy6Nqt1gaOfyFGG8kKAuQsulTklbApk3R2JPkufEo+iocf/yx3VhFFlIRcvgLMd5IUBYo6a6tpJikrYCbbooslfjcq6+ubrVkWR1NJ5hsatkIIUYDCcqQUcUKSFo4ULz0cJbV0XSCyaaWjRBiNJCgDBlVrYDYwikaXV9kdeRZSEVo6nwhxhsJypBR1woo6m5qOwxYU+cLMd5IUIaMulZAntCsWtVNQ9/EshFCjAYSlCGjrhWgbighRL+QoAwhdayAugI0OwvLl0fnmkX7/RpHokGRQgw3EpQxoKoAzc7C2942f/DjgQNw0UXNG/cskchL06BIIYYbzeW1AOfyGhTLlx87kj5mYiIa71LHJ5I1J9iSJZHlc+jQ0bTJyWgJ46x7J+cYE0IsDDSXlyhkdjZfTCCaSv/CC+t1gWWNmXn22fliAtE5efduY1CkutKE6A9drikvhoiq06PUWbu+DTHodVBk2kqKu9JAEWhCtI0sFAHUa/wPHoyslbK3/V7FoCgaLWl1LF8ebVkWiOYXE6J/SFAE0Kzxz3Ocx4393r3HTvuyZAksXVqe98REfjRa2oEfz6Kc5czX/GJC9A8JyghTx3dQZxr8JOm3/WRjD1EjH4vKqlVw441www1Hw5jzOHIkX0w2bDjW6sgq0+xsVPcsNL+YEO0jH8qI0sR3cNxxR8+fmoI3vQluvbXYWQ/R2/7sbP5qkO7HRmvFZYgtmTRZDX5cp8OHi8sDR+ubda4GdgrRDbJQRpQ6voO4oU4Kx/e+B697HTz5ZPn6KiedNN8qySKvi6nOSP6sOuUxMZF/btKCEUK0hwRlRKnjOygTn6eeKr9fWUOf18VUZyR/Vb+HWbkVo4GTQrSPBGVEqTMrcZn4FPkbFi0q7xKDyNLJi8RKjuTfvDkSsqzzqvg9zOA3f/PoWjBFKNpLiHaRoIwodbqSysRn8+YoOiuLI0eKnesx3/3usZFYl146P2jg0kuLp1/JqtOSJVGXXGzd3HwzXHtt9SADRXsJ0SLuPrbbK1/5Sh9ltm51X7XK3Sz6u3Vr/nmTk+5RMx5tk5Pzz9+61X1qav45yc2s+HMv16xalV+OqaniesX1n5goz1sIUQ1gzjPa1IE36oPcRl1QysgTiaJGukgokuJVJiZ1NrPi8qbFL6+uZaIphKhGnqCoy2tMyZpZOObpp/OvK1qwKzmjcRUfRlVWrsyORIup4gvRapJCdI8EZUzZtCmaqDGLQ4fyG+iqvpmmAyXT/pg477KQ4Sq+kF5Xk9Qkk8Xo+YiBdzsNchvnLq8yH0fczZRFHd9MHR/K5KT7JZfMzzv+XNYt1rUvRF1mxej5jBfIhyJBSVLkYK/aQMfCAked3mmBKRODWFSyhCmrkcraqvpQqohgHkX1aJLfqJH3fBT0MJpIUCQoP2DrVvclS/IbyKVLmzm5sxr4qqKQbJiTQlW2FQUQlJW1yrUxZRZd12/jvQpi1+Q9nyJLVwwvEhQJyg8oaqyrNrJlDX461LeqQCxZEglaVfHptb55QpBuwMssui7fxoehO0kWyngxEEEBzgF2AbuBKzOOLwO2heM7gNWJY1eF9F3A2TXyfB/wTJXyjaugtPE22cQH01Y4cd1Gqqys6fyyGvCqQteFBTEMjfUwiJ5oj74LCjABfBU4DVgK3AesS51zKXBd2F8PbAv768L5y4A1IZ+JsjyBaeBmCUoxdRqovK6WOhZKMq+q3V91LYom9c0Tv7zzp6aqi2KbjemwdCct9G450R6DEJTXAnckPl8FXJU65w7gtWF/MfAkYOlz4/OK8gxi8zngZAlKMVXfJovOq+pDybp33qj1sq3qqPh0Y1YmZGnxK2vAqwpjWxbEMFgoYrwYhKBcAHww8fmtwDWpc3YCpyY+fxVYDlwDXJhIvz7kl5sncDlwRdjPFRRgIzAHzK1cubKLZz0UVHmbLGvIqkR55d07r0Eu6loq8ncUCV9cxiyhSJ9XdZqW5PlVLZ+m1O1OkqUgumakBQX4UeDvgcVeIijJbVwtlKpU6Wpp2ngViVGRFZP1Vl7URZXlC5maml/eKhZHUQPeDwui13nZ0uN7JDKiF0a6ywv4JeCfgT1hOwLsLiujBKWYKhZKV47Yojf/dP5VJqIsaujz6jkxUa0B7uI5NBXqvLpkDSKVqIimDEJQFgOPEDnVYwf6GalzLmO+U/7WsH8G853yjwQfSWme4XpZKDnUaajKGsou38zrhPrWjR5Ld0W14fRus5vpkkvKBSDrfmUzE2RZcEI0oe+CEt2T84CHQ1fWppD2buCNYf95wIeJQoDvAk5LXLspXLcLOLcoz4z7SlAyaPImXdRQdhl9VMeRnlevvLEjcbRW2RiTJsLYq7Bs3Vo+jX8vYc1l1p4QVRiIoCz0bdwEpW2LouvpSIreuNOilffGXqXhzUpr0iXURtdX0TON69zm8gB5370c+/UZp2cmQZGgtG5RlFkRbfTT9yqCVUe8J8eY1IlYa7Os7uXrzdTt1irbsr57DVKsz7g9MwmKBKUTn0eVaVV6eVur849a5Q2xSFR7bRTqCHbTAaN1AxCqWChVRVfjXvIZt7FCEpSMbdwEZVBRWb3ep4pQVK1b0T9+lUahqCxVG5WmA0aLtiY+lDicuOr9FtrI/CL63f00LLMZtIUEJWMbN0Fx7+4frUq/fpdva0VjUZIUNeZNRsgnx7VMTeX7YqpYcvEzSo4ZqTqrQJV7pIMRqpapH99fmwyi+0kWigRlLAWlK6q8WXf5tlZn3Erd7qa4UajS+DYdOJnX+FXp4qpjAdV5bv1skNtkEI171Wc+Ko57CUrGJkFpl7I33kFYKGX33bq1eGp6s8hicK/e+FYdOFklj7Jre/EnVS3bxES33ZVNyct7UN1PZXUdJce9BCVjk6B0wyD+ceqEGCevKVpoLF32OsKQpIkjvaibrWiVyybPrSvLssvfQVHevc580EbZ0hGDU1PuixbllyvZZTkM1osEJWOToHTHIP456kYn1fUfVO26iiPGmtwnq8xdP8uuLMtBzKRQ9XuqI2x1nn+bQRUL2XqRoGRsEpTRou4bcR3LIWkxxI1LUVdZ2Uj+OoLUL/KeX9OJJbvseqoSQFEW2FBF2KpMg5OkzUGnbYlvF0hQMjYJysIl/VZYtVGr8zbZRoRTmQClywXVIre6Ju85ZT33piIzKAslTdkSA0Wh6EXT4GQ9w7bHCdUV3371DEhQMjYJysKkapdFXqNW9Z/qkkuq/VOnI7eqhPVmheim6Uc0UtazqGPJ5ZWxylt7lfs0bQDzfiNZi7BVeXGoM2YpeU3RM2ljqzOBZz99lxKUjE2C0l/ajDzKa9Ty3qirNHbJPJPO1LoDBqv2h5c1AL2+beblX8fXVKeRzApfTt4r3dj32gDmRehliVaTFTaL6t501dGi30xWnkuXVn8e/QyXlqBkbBKU/tHF2Ig6/+hNQnmLGt+s+5ZNXZIVflvU9dTlRJNFjWrTgY/J7pkqvoeiaKy2G9Hkcy6rQ1nd6/w+ky8meVZufL+8383zn1/txaKf4dISlIxNgtI/6rw9te3YzPqn6rKvu6zPPq53E/9OnbfNunVMn79kSdSYVb0+LltRV2Ky/EXlqyqeTRrRqi8TWdZuPC6pSh51G/JeB5nKQhnwJkHpH3UnTqwyNiIrvU0Lpc6WzL9q3lmj6ps8rzzyyvH85x+bfxsCmzfOIutZVbF+qjSEeXlk+R6S96w6A0EdC7Ls+rSV0sTXU/Rs5EMZ8CZB6R91357KoryKoo968aE02Zr22WeJRfxM2pjxN6scWf6dLpzJVZ5ZlYkpy4It8ganxr6HIhGp0v1V9nzz8q5Sv6xgjzq/lSr/N4ry6uMmQekfXbw9VQ19rRJinOV8z2twyvq0i/rDq25tLvqVLGuv5Wp7KxpBXtY4x8+jyPfQy0tDlUi9ot9iXSt4ctL9+OOrndtFN1YdJCgZmwSlv/Tr7akp6fLVfTuM82jL8qnToFVlENZI2dZ0+eK4Ye2qTr0I+tatzb/zOqP8B/U/JUHJ2CQoooiybqc23kqLti6ic3opX5diFFt9TcrUS52yGnCzfEuhil+kl5eKeJaEou7eIj9OHDTQNRKUjE2CIvKo0jef1QVTpwEteysvCn1t+jZat7Gr48TOEp86voEmghw/h14EJavBLotAKzrWS7di0ctKmqKw5q4tFQlKxiZBGV/K/mHLooeKxlDkpef5evIa7GRjl3VOWRdMkY+pSqMXN251G/mst+Qq95uaqhemnGw48/KvIoRNRsk33Yr8RfFzq+pvLKpb1z4WCUrGJkEZT6r8w5aF7dZ5g63S9751a73GNKvBTI5EL5uapKzBrLvIV1mD1qZvKbkV5W/mvm5d9RDhrssbi3rTkPg6oe9ddJcmkaBkbBKU8aRKCHPZOUXHm3RNlXWzVN3iLrmixqZOqG5RXcsGcGbVsc03/3QDmzU6v+ozbbO8WX6ZLEd60zonfStVxadtJCgZmwRlPKkyaLDMimk7DLrthrapONUZMHfJJc2mhs/zT9Xd2uymKipvnWcZ/waqvFT0OpPB5KT7WWcVd4V2FQUmQcnYJCjjSdVBlmX/jL0eT9J2BFUTx3CRINYZMV4lnzbq19YzLHsRqFrerJmOi6jzHIoskbwBn1UmzmyKBCVjk6CMJ/2YoqLuPdp2Atcd29HL22udgaRt+CXyli6o+gzjEepw1MIqK3dZ917RlCpFz60o32QgR5nYlIl8+rvuFQlKxiZBGV+6HhBW1QpKlifLqXzWWc1HusdjJKq8/bZF0XOt0uCXWRkTE8XT7hTN8VY2VqRobrW6glf195QXUp2etr7Ks4vvW+XcXpGgZGwSFNEVTSZ3rCJydd7y43uVNTJtOXDLrLIysShrEJOD/vLqUVU064hbXSGv+0zT3VNx11l6AGUVi7Oq/6xu91waCUrGJkERXVHXQqlDWniqrr/SdVdf08i4+JyqI8CLxLrqc6/rb8lbAKuKgDWN+ms6wWid7tOmwiJBydgkKKJNyt4ou5pKvKjxaWvJ3SqUWWV1BK1J11lRd1gbEWHpLsSiLsXkiPcmIt7UpxaLXpPItDoMRFCAc4BdwG7gyozjy4Bt4fgOYHXi2FUhfRdwdlmewGxI3wncACwpK58ERSENa1QAAAx9SURBVLRF3T75Lu7fJIS3TapYB3kRSXVErkxA8+a+St6/bqObtjqKypJsoJtaqm1E/dXJo+5vpO+CAkwAXwVOA5YC9wHrUudcClwX9tcD28L+unD+MmBNyGeiKE/gPMDC9iHgkrIySlBEW3TZxVWVJn6bOlQJk677Nt70Db6JgOZ1pcXWRt251ao8l6bfSVtRf1VmL27yGxmEoLwWuCPx+SrgqtQ5dwCvDfuLgSeDIMw7Nz6vSp4h/Qpgc1kZJSiiLbpuzKvQtd+m6sJldayNJtFwZY7nvGdedq8ih3/T7sqm30lbIdZxIENZUMEwWCgXAB9MfH4rcE3qnJ3AqYnPXwWWA9cAFybSrw/5VclzCfAl4GdzyrURmAPmVq5cWe8pCpHDQrBQunC6dx0hVkeIqzayaYEoc1Qn75U3fUvTKeF7+U7qdM9VsdbaHOyYJyiLGD2uBT7v7n+XddDdt7j7tLtPr1ixos9FE6PK5s0wOTk/bXIySu8XMzOwZQusWgVm0d8tW6L0JszOwsaNsHdv/jn79jXLO2blyurpmzbBwYPF+cXPPFl29+ivWfG9Zmfhuuui85O4w+23F983j16+k5kZ2LMnuv/NN8PERPZ5ZlFd07+/JUvgmWdg0SJYvTpKe/JJ2Lq1vd/IMWSpTBsbA+jyAn4f+DiwqEoZ1eUl2qTrwZL9pko/fq8WSp03+LJuruQzLxrLknevOuNYBkFROHXa+shab6fNKEMG0OW1GHiEyKkeO9DPSJ1zGfOd8reG/TOY75R/hMghn5sn8BvAF4DjqpZRgiJEPlUHIvZKVSGu061YtlZIHQd6v7sui8iLkqu7aFqv31vfBSW6J+cBDxP5RjaFtHcDbwz7zwM+TBQCfBdwWuLaTeG6XcC5RXmG9OdC2r1h+72y8klQhMin6kDEflHHmmni0yobob9QaRIR1uvLwEAEZaFvEhQh8unHJJpNylTFmmkrhLkXh3y/aHOpgqpIUCQoQtRmmP1CTac8Gbb6Nh2z0otfKE9QLDo2nkxPT/vc3NygiyGEEI2JI9rKIuDSrFoVRZE1wczucffpdPoohg0LIcTYkAxNhmPDo5csgaVL56d1FdIuQRFCiCEnPWYlOc7kxhvhhhs6HHuSQF1e6vISQohaqMtLCCFEp0hQhBBCtIIERQghRCtIUIQQQrSCBEUIIUQrjHWUl5ntB9KTcy8nmvV4VBi1+oDqNAyMWn1AdUqyyt2PWf9jrAUlCzObywqHG1ZGrT6gOg0Do1YfUJ2qoC4vIYQQrSBBEUII0QoSlGPZMugCtMyo1QdUp2Fg1OoDqlMp8qEIIYRoBVkoQgghWkGCIoQQohUkKAEzO8fMdpnZbjO7ctDlaYqZ7TGzL5vZvWY2F9JOMrM7zewr4e8PDbqcRZjZDWb2hJntTKRl1sEi3he+t/vN7BWDK3k2OfV5l5k9Fr6ne83svMSxq0J9dpnZ2YMpdTFm9mIz+5yZPWhmD5jZ5SF9KL+ngvoM7fdkZs8zs7vM7L5Qpz8I6WvMbEco+zYzWxrSl4XPu8Px1bVvmrWM47htwATwVeA0YClwH7Bu0OVqWJc9wPJU2n8Hrgz7VwLvHXQ5S+rwc8ArgJ1ldQDOA/4aMOA1wI5Bl79ifd4F/KeMc9eF398yYE34XU4Mug4Z5TwZeEXYPwF4OJR9KL+ngvoM7fcUnvXxYX8JsCM8+1uB9SH9OuCSsH8pcF3YXw9sq3tPWSgRrwJ2u/sj7n4IuAU4f8BlapPzgZvC/k3ALw+wLKW4++eBp1LJeXU4H/gLj/gH4IVmdnJ/SlqNnPrkcT5wi7t/392/Buwm+n0uKNz9G+7+pbD/NPAQcApD+j0V1CePBf89hWf9TPi4JGwO/GvgtpCe/o7i7+424Cyz9PqPxUhQIk4Bvp74/CjFP6aFjAOfMrN7zGxjSHuRu38j7P8z8KLBFK0n8uowzN/d20P3zw2Jbsihq0/oGvkpojfgof+eUvWBIf6ezGzCzO4FngDuJLKkvuXuz4VTkuX+QZ3C8W8DU3XuJ0EZPX7G3V8BnAtcZmY/lzzokT071LHio1AH4APAS4AzgW8A/3OwxWmGmR0PfAT4bXf/TvLYMH5PGfUZ6u/J3Q+7+5nAqUQW1Eu7vJ8EJeIx4MWJz6eGtKHD3R8Lf58APkb0I3o87l4If58YXAkbk1eHofzu3P3x8M9+BPgzjnaXDE19zGwJUeM76+4fDclD+z1l1WcUvicAd/8W8DngtUTdjYvDoWS5f1CncPxE4ECd+0hQIu4G1oboh6VEDqntAy5Tbczs+WZ2QrwP/CKwk6guG8JpG4C/GkwJeyKvDtuBXwtRRK8Bvp3oclmwpPwHv0L0PUFUn/Uh4mYNsBa4q9/lKyP0rV8PPOTuf5w4NJTfU159hvl7MrMVZvbCsH8c8AYi39DngAvCaenvKP7uLgA+G6zM6gw6EmGhbERRKA8T9TFuGnR5GtbhNKLIk/uAB+J6EPWDfgb4CvBp4KRBl7WkHh8i6l54lqiP9+K8OhBFsrw/fG9fBqYHXf6K9bk5lPf+8I98cuL8TaE+u4BzB13+nDr9DFF31v3AvWE7b1i/p4L6DO33BLwc+MdQ9p3A74X004jEbzfwYWBZSH9e+Lw7HD+t7j019YoQQohWUJeXEEKIVpCgCCGEaAUJihBCiFaQoAghhGgFCYoQQohWkKCIkcTMDofZYR8Is63+RzNbFI5Nm9n7Or7/F8Lf1Wb2qw2u/ykzuz5c/2hc9sTxe83s1XXL0zVmttTMPp8YOCfGCAmKGFW+5+5nuvsZRAO6zgV+H8Dd59z9Hb3eoKjRdPefDrurgdqCAvwu8D533wPsA342cd+XAie4+46ca48pY6I8neLR5KqfAd7cj/uJhYUERYw8Hk1Ds5Fokj8zs583s0+Y2SKL1o95YXyuRet4vCiMMv6Imd0dtteF4+8ys5vN7P8CN5vZGWHNiXvDBIJrw3nxLK/vAX42HL8ivL2fmbjf35vZTybLG2Y7eLm73xeSPkQ0e0PMeuCWYL38nZl9KWw/Ha7/+ZC+HXgwWR4zO97MPhPO/7KZnR/SV5vZQ2b2Z8Gq+1QYXY2ZnW5mnw6W3pfM7CUh/XfCs7nfwlobgY8DMw2/LjHMDHo0pzZtXWzAMxlp3yKa/fbngU+EtKuBt4X9VwOfDvt/STTRJsBKoik5IFof4x7guPD5T4GZsL80kf5M+PuDe4XPG4D/FfZ/DJjLKOfrgY8kPr+IaKT94vD5IeBlwCTwvJC2Ns4r3PO7wJr08wAWAy8I+8uJRkUbkSX1HHBmOHYrcGHY3wH8Sth/XrjvLwJbwrWLgE8APxfOmQD2D/o3oK3/m/o5xbizDfg94EbCokIh/ReAdXZ0OYgXhJloAba7+/fC/heBTWZ2KvBRd/9Kyf0+DPxXM/sd4CLgzzPOORnYH39w98ctWu3xLDN7HHjO3Xea2YnANcHiOUwkUDF3ebRORxoD/tCiWaiPEE1ZHk8x/zV3vzfs3wOsDtbSKe7+sVCWfwEws18kEpV/DOcfTyRqn3f3w2Z2yMxO8GhtETEmSFDEWGBmpxE1uk8AP5E49EXgdDNbQbTQ0H8L6YuA18QNaCIfiN7+AXD3vzSzHcAvAbeb2X9w98/mlcPdD5rZnUSLGb0JeGXGad8jsgSSxN1ej4d9gCvC558M5U2W9btkMwOsAF7p7s+a2Z7Evb6fOO8wcFxePYiE6Y/c/X/nHF+WKo8YA+RDESNPEIvrgGvcfd7kdeHzx4A/JurWiqfr/hTwW4k8ziSDIFSPuPv7iGZtfXnqlKeJlpRN8kHgfcDd7v7NjGwfAk5PpX2UaLLCNxOtKArR9OLf8Ghq9bcSdTWVcSLwRBCT1wOrik4OFsajZvbL8IN1xyeBO4CLYqvNzE4xsx8O+1PAk+7+bIXyiBFCgiJGlePisGGiWW8/BfxBzrnbgAs52t0F8A5gOjicHwR+M+faNwE7LVoV72XAX6SO3w8cDg7tKwDc/R7gO0TdbMfg7v8EnBi6m+K0bxFZU4+7+yMh+Vpgg5ndR7RwUp5VkmQ21OvLwK8B/1ThmrcC7zCz+4EvAD/i7p8i8jN9MeR1G0eF8/XA/6mQrxgxNNuwEH3GzH4U+BvgpcG6yDrnCuBpd/9gP8vWBmb2UeBKd3940GUR/UUWihB9xMx+jShqalOemAQ+wHyfxlBg0QJ1H5eYjCeyUIQQQrSCLBQhhBCtIEERQgjRChIUIYQQrSBBEUII0QoSFCGEEK3w/wFyTJfoYuftEwAAAABJRU5ErkJggg==\n",
	"text/plain": [
	"<Figure size 432x288 with 1 Axes>"
	]
	},
	"metadata": {
	"needs_background": "light"
	},
	"output_type": "display_data"
	}
	],
	"source": [
	"plt.xlabel('Diversity (Variance)')\n",
	"plt.ylabel('MSE')\n",
	"plt.plot(var_range, MSE, 'bo')"
	]
	}
	],
	"metadata": {
	"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": "3.6.9"
	}
	},
	"nbformat": 4,
	"nbformat_minor": 4
	}