statistics/statistic-exercise.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "# Write yor imports here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Statistics Exercise\n",
    "## Statistical Distributions. Properties of distributions. Applications of Probability and Statistics in Computer Science"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 1. Plotting a Single Distribution. Digits in $\\pi$ and $e$\n",
    "We expect that the decimal digits in $\\pi$ and $e$ will be randomly distributed and there's no reason for any digit to dominate over others. Let's verify this.\n",
    "\n",
    "Using an algorithm, the first 10 004 digits of $\\pi$ and $e$ were generated:\n",
    "$$\n",
    "\\pi = 3.(141592 \\dots 5678)5667\n",
    "$$\n",
    "$$\n",
    "e = 2.(718281 \\dots 6788)5674\n",
    "$$\n",
    "\n",
    "The 10 000 digits in brackets were counted. You can see the results in `digits.dat`. Each column corresponds to one digit from 0 to 9. The first row is for $\\pi$ and the second row is for $e$.\n",
    "\n",
    "How are these digits distributed? Are the two distributions different?\n",
    "\n",
    "**Note:** The dataset is **not properly formatted** to work easily. You can transpose it. Now, digit counts will be in rows and variables - in columns. \n",
    "```python\n",
    "digits = pd.read_table(\"digits.dat\", header = None).T\n",
    "```\n",
    "\n",
    "You can also specify column names like this:\n",
    "```python\n",
    "digits.columns = [\"pi\", \"e\"]\n",
    "```\n",
    "\n",
    "Also note that **we are not creating the histogram of the distribution**. We already have the counts, we need to plot them. In a sense, the histogram has already been calculated.\n",
    "\n",
    "To do this, we can create a \"bar chart\" (using `plt.bar()`). We have to provide values for the x-axis and y-axis. For the x-axis, we have the numbers 0 through 9 (we can use the *index* of the dataset like this: `digits.index`). For the y-axis, we need to plot the digit counts directly.\n",
    "\n",
    "We can see that even the simplest datasets sometimes need a bit of preprocessing. This is always the case when we're working with data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pi</th>\n",
       "      <th>e</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>968</td>\n",
       "      <td>974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1026</td>\n",
       "      <td>989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1021</td>\n",
       "      <td>1004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>974</td>\n",
       "      <td>1008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1012</td>\n",
       "      <td>982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1046</td>\n",
       "      <td>992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1021</td>\n",
       "      <td>1079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>970</td>\n",
       "      <td>1008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>948</td>\n",
       "      <td>996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1014</td>\n",
       "      <td>968</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     pi     e\n",
       "0   968   974\n",
       "1  1026   989\n",
       "2  1021  1004\n",
       "3   974  1008\n",
       "4  1012   982\n",
       "5  1046   992\n",
       "6  1021  1079\n",
       "7   970  1008\n",
       "8   948   996\n",
       "9  1014   968"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "digits_data = pd.read_table(\"digits.dat\", header = None).T\n",
    "digits_data.columns = [\"pi\", \"e\"]\n",
    "digits_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd566db70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEICAYAAABWJCMKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAEFlJREFUeJzt3X+s3XV9x/Hnyxbkhz/KjwuBtlsxNkxCMmkqoiTEWOcAnWWLJJhNO4Lp/kAFWSLVf8hcsmBixJktJB3F1QxRBhg6R1TGj7n9AdoCClgdFZBeqfQafjmJQfS9P86n80JvaXtP+72Fz/ORnJzv9/P9fM/7c3pvz+t8P+d8vzdVhSSpP6+a6wFIkuaGASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEg7aUkxye5IclUkoeTfGyuxyTNhgEg7YUkrwL+DfgesBBYAVyc5I/ndGDSLBgA0t55CzBRVZ+uqueq6iHgn4Dz5nhc0l6bP9cDkF5mfh84PslT09rmAf81R+ORZs0AkPbOVuDhqlo61wORxuUUkLR3vgM8k+TSJIcmmZfk5CRvmeuBSXvLAJD2QlX9BvgT4M3Aw8DPgauA18/luKTZiH8QRpL65BGAJHXKAJCkThkAktQpA0CSOnVAnwdw9NFH15IlS+Z6GJL0srJp06afV9XE7vod0AGwZMkSNm7cONfDkKSXlSQ/2ZN+TgFJUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDugzgSXt3pI1/77fazxy+Xv2ew0NzyMASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpU7sNgCRXJ9me5P5pbUcmuSXJg+3+iNaeJF9IsiXJ95Msm7bPqtb/wSSr9s/TkSTtqT25GNw/A/8AfGla2xrg1qq6PMmatn4pcBawtN3eClwJvDXJkcBlwHKggE1JNlTVk/vqiejA4IXJ+uLP++VttwFQVd9OsuRFzSuBd7Tl9cAdjAJgJfClqirgziQLkhzX+t5SVU8AJLkFOBO4duxnoJ34n1LSnpjtZwDHVtU2gHZ/TGtfCGyd1m+yte2qfSdJVifZmGTj1NTULIcnSdqdff33ADJDW71E+86NVWuBtQDLly+fsY90oPGoa3j+m49vtgHweJLjqmpbm+LZ3tongcXT+i0CHmvt73hR+x2zrP2ysL9/OV/pv5iS9r/ZBsAGYBVwebu/aVr7R5J8hdGHwE+3kPgm8Hc7vi0EvBv45OyHLe3Md4TS3tltACS5ltG796OTTDL6Ns/lwHVJLgAeBc5t3W8Gzga2AM8C5wNU1RNJ/hb4buv36R0fCEuS5saefAvoA7vYtGKGvgVcuIvHuRq4eq9GNyanYSRp1/yj8JK0l14p041eCkKSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdGisAknw8yQNJ7k9ybZJDkpyQ5K4kDyb5apKDW99Xt/UtbfuSffEEJEmzM+sASLIQ+BiwvKpOBuYB5wGfAa6oqqXAk8AFbZcLgCer6o3AFa2fJGmOjDsFNB84NMl84DBgG/BO4Pq2fT1wTlte2dZp21ckyZj1JUmzNOsAqKqfAp8FHmX0wv80sAl4qqqeb90mgYVteSGwte37fOt/1IsfN8nqJBuTbJyamprt8CRJuzHOFNARjN7VnwAcDxwOnDVD19qxy0ts+11D1dqqWl5VyycmJmY7PEnSbowzBfQu4OGqmqqqXwM3Am8HFrQpIYBFwGNteRJYDNC2vx54Yoz6kqQxjBMAjwKnJTmszeWvAH4A3A68v/VZBdzUlje0ddr226pqpyMASdIwxvkM4C5GH+beDdzXHmstcClwSZItjOb417Vd1gFHtfZLgDVjjFuSNKb5u++ya1V1GXDZi5ofAk6doe+vgHPHqSdJ2nc8E1iSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVNjBUCSBUmuT/LDJJuTvC3JkUluSfJguz+i9U2SLyTZkuT7SZbtm6cgSZqNcY8A/h74RlX9AfCHwGZgDXBrVS0Fbm3rAGcBS9ttNXDlmLUlSWOYdQAkeR1wBrAOoKqeq6qngJXA+tZtPXBOW14JfKlG7gQWJDlu1iOXJI1lnCOANwBTwBeT3JPkqiSHA8dW1TaAdn9M678Q2Dpt/8nWJkmaA+MEwHxgGXBlVZ0C/JLfTffMJDO01U6dktVJNibZODU1NcbwJEkvZZwAmAQmq+qutn49o0B4fMfUTrvfPq3/4mn7LwIee/GDVtXaqlpeVcsnJibGGJ4k6aXMOgCq6mfA1iQntqYVwA+ADcCq1rYKuKktbwA+1L4NdBrw9I6pIknS8OaPuf9HgWuSHAw8BJzPKFSuS3IB8Chwbut7M3A2sAV4tvWVJM2RsQKgqu4Fls+wacUMfQu4cJx6kqR9xzOBJalTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqfGDoAk85Lck+Trbf2EJHcleTDJV5Mc3Npf3da3tO1Lxq0tSZq9fXEEcBGwedr6Z4Arqmop8CRwQWu/AHiyqt4IXNH6SZLmyFgBkGQR8B7gqrYe4J3A9a3LeuCctryyrdO2r2j9JUlzYNwjgM8DnwB+29aPAp6qqufb+iSwsC0vBLYCtO1Pt/4vkGR1ko1JNk5NTY05PEnSrsw6AJK8F9heVZumN8/QtfZg2+8aqtZW1fKqWj4xMTHb4UmSdmP+GPueDrwvydnAIcDrGB0RLEgyv73LXwQ81vpPAouBySTzgdcDT4xRX5I0hlkfAVTVJ6tqUVUtAc4DbquqPwduB97fuq0CbmrLG9o6bfttVbXTEYAkaRj74zyAS4FLkmxhNMe/rrWvA45q7ZcAa/ZDbUnSHhpnCuj/VdUdwB1t+SHg1Bn6/Ao4d1/UkySNzzOBJalTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6NesASLI4ye1JNid5IMlFrf3IJLckebDdH9Hak+QLSbYk+X6SZfvqSUiS9t44RwDPA39dVW8CTgMuTHISsAa4taqWAre2dYCzgKXtthq4cozakqQxzToAqmpbVd3dln8BbAYWAiuB9a3beuCctrwS+FKN3AksSHLcrEcuSRrLPvkMIMkS4BTgLuDYqtoGo5AAjmndFgJbp+022dpe/Firk2xMsnFqampfDE+SNIOxAyDJa4AbgIur6pmX6jpDW+3UULW2qpZX1fKJiYlxhydJ2oWxAiDJQYxe/K+pqhtb8+M7pnba/fbWPgksnrb7IuCxcepLkmZvnG8BBVgHbK6qz03btAFY1ZZXATdNa/9Q+zbQacDTO6aKJEnDmz/GvqcDHwTuS3Jva/sUcDlwXZILgEeBc9u2m4GzgS3As8D5Y9SWJI1p1gFQVf/NzPP6ACtm6F/AhbOtJ0natzwTWJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktQpA0CSOmUASFKnDABJ6pQBIEmdMgAkqVMGgCR1ygCQpE4ZAJLUKQNAkjplAEhSpwwASeqUASBJnTIAJKlTBoAkdcoAkKROGQCS1CkDQJI6ZQBIUqcMAEnqlAEgSZ0yACSpUwaAJHVq8ABIcmaSHyXZkmTN0PUlSSODBkCSecA/AmcBJwEfSHLSkGOQJI0MfQRwKrClqh6qqueArwArBx6DJAlIVQ1XLHk/cGZVfbitfxB4a1V9ZFqf1cDqtnoi8KPBBghHAz8fsJ61rW1ta+8Pv19VE7vrNH+IkUyTGdpekEBVtRZYO8xwXijJxqpabm1rW9var5TaL2XoKaBJYPG09UXAYwOPQZLE8AHwXWBpkhOSHAycB2wYeAySJAaeAqqq55N8BPgmMA+4uqoeGHIMuzEnU0/Wtra1rT0XBv0QWJJ04PBMYEnqlAEgSZ0yAJjby1MkuTrJ9iT3D1m31V6c5PYkm5M8kOSiAWsfkuQ7Sb7Xav/NULWnjWFeknuSfH3guo8kuS/JvUk2Dlx7QZLrk/yw/dzfNlDdE9vz3XF7JsnFQ9Ru9T/efs/uT3JtkkMGrH1Rq/vAkM95j1RV1zdGH0b/GHgDcDDwPeCkAeufASwD7p+D534csKwtvxb4n6GeO6NzQl7Tlg8C7gJOG/j5XwJ8Gfj6wHUfAY4e+ufdaq8HPtyWDwYWzMEY5gE/Y3Sy0hD1FgIPA4e29euAvxyo9snA/cBhjL508x/A0rn42c908whgji9PUVXfBp4Yqt6Lam+rqrvb8i+AzYz+swxRu6rqf9vqQe022DcSkiwC3gNcNVTNuZbkdYzecKwDqKrnquqpORjKCuDHVfWTAWvOBw5NMp/Ri/FQ5x+9Cbizqp6tqueB/wT+dKDau2UAjF7wtk5bn2SgF8EDSZIlwCmM3okPVXNeknuB7cAtVTVYbeDzwCeA3w5Yc4cCvpVkU7v0yVDeAEwBX2xTX1clOXzA+jucB1w7VLGq+inwWeBRYBvwdFV9a6Dy9wNnJDkqyWHA2bzwZNg5ZQDsweUpXumSvAa4Abi4qp4Zqm5V/aaq3szojPBTk5w8RN0k7wW2V9WmIerN4PSqWsboqrgXJjljoLrzGU03XllVpwC/BIb+zOtg4H3Avw5Y8whGR/UnAMcDhyf5iyFqV9Vm4DPALcA3GE0xPz9E7T1hAHR+eYokBzF68b+mqm6cizG0aYg7gDMHKnk68L4kjzCa8ntnkn8ZqDZV9Vi73w58jdE05BAmgclpR1rXMwqEIZ0F3F1Vjw9Y813Aw1U1VVW/Bm4E3j5U8apaV1XLquoMRtO9Dw5Ve3cMgI4vT5EkjOaDN1fV5wauPZFkQVs+lNF/0h8OUbuqPllVi6pqCaOf921VNcg7wiSHJ3ntjmXg3YymCfa7qvoZsDXJia1pBfCDIWpP8wEGnP5pHgVOS3JY+51fwejzrkEkOabd/x7wZwz//Hdp6KuBHnBqji9PkeRa4B3A0Ukmgcuqat1A5U8HPgjc1+biAT5VVTcPUPs4YH37I0GvAq6rqkG/jjlHjgW+NnodYj7w5ar6xoD1Pwpc097sPAScP1ThNgf+R8BfDVUToKruSnI9cDej6Zd7GPbSDDckOQr4NXBhVT05YO2X5KUgJKlTTgFJUqcMAEnqlAEgSZ0yACSpUwaAJHXKAJCkThkAktSp/wMBWDhR3aBrhgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd5685780>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd5a21a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.bar(digits_data.index, digits_data.pi)\n",
    "plt.xticks(range(10))\n",
    "plt.title(\"Pi\")\n",
    "plt.show()\n",
    "plt.bar(digits_data.index, digits_data.e)\n",
    "plt.xticks(range(10))\n",
    "plt.title(\"e\")\n",
    "plt.show()\n",
    "plt.bar(digits_data.index, digits_data.pi,label = \"pi\", color = \"red\")\n",
    "plt.bar(digits_data.index, digits_data.e,label = \"e\",color = \"green\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's try something else. Scientists have measured the percentage of silica ($\\text{SiO}_2$, sand / glass) for 22 meteors. You can find it in `silica.dat`. How are these distributed? What is a \"typical\" percentage? Is there such percentage at all?\n",
    "\n",
    "Print the mean, standard deviation (you can use the biased or unbiased formula), skewness and kurtosis of the distribution. What do these numbers tell you? How do they relateto the shape of the distribution? Can you characterize the distribution better? (An idea would be to characterize different parts of it on their own, as if they're different distributions.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>20.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>22.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>22.71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>22.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>27.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>27.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>27.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>27.57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>27.81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>28.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>29.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>30.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>31.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>32.88</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>33.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>33.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>33.40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>33.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>33.83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>33.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>34.82</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        0\n",
       "0   20.77\n",
       "1   22.56\n",
       "2   22.71\n",
       "3   22.99\n",
       "4   26.39\n",
       "5   27.08\n",
       "6   27.32\n",
       "7   27.33\n",
       "8   27.57\n",
       "9   27.81\n",
       "10  28.69\n",
       "11  29.36\n",
       "12  30.25\n",
       "13  31.89\n",
       "14  32.88\n",
       "15  33.23\n",
       "16  33.28\n",
       "17  33.40\n",
       "18  33.52\n",
       "19  33.83\n",
       "20  33.95\n",
       "21  34.82"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "silica_data = pd.read_table(\"silica.dat\", header = None)\n",
    "silica_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd5a86a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean value =  29.165000000000006\n",
      "Standart derivation =  4.29153513436903\n",
      "Skewness =  -0.44644211385029625\n",
      "Kurtosis =  -0.9242970726953743\n"
     ]
    }
   ],
   "source": [
    "plt.hist(silica_data[0], bins = 22)\n",
    "plt.vlines(silica_data[0].mean(),0,4,color = \"red\", linewidth = 2, linestyle = \"dashed\", label = \"mean\")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "print(\"Mean value = \", silica_data[0].mean())\n",
    "print(\"Standart derivation = \", silica_data[0].std())\n",
    "print(\"Skewness = \", silica_data[0].skew())\n",
    "print(\"Kurtosis = \", silica_data[0].kurt())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 2. Categorical Variables. Comparing Categories\n",
    "In addition to numeric variables (like age and salary), in statistics we also use **categorical variables**. These are descriptions of quality (as opposed to quantity). Such variables can be gender, smoker / non-smoker, results of a medical study (healthy / not healthy), colors (red, green, blue), etc. To plot values of categories, we use *bar charts*. Since category names can be long, it's sometimes useful to plot the lines horizontally.\n",
    "\n",
    "<p style=\"color: #d9534f\"><strong>There is a very significant difference between histograms and bar charts. Histograms are used to plot the frequency distribution of one numeric variable. Bar charts are used to plot categorical variables - how each value compares to other values.</strong></p>\n",
    "\n",
    "The dataset `us_budget.dat` contains the figures for the eight main items in the US budget for 1978 and 1979 in billions\n",
    "of dollars.\n",
    "\n",
    "Display the two budgets separately. Use `xlabel()` (or `ylabel()` if your plot is horizontal) to write the names of each category. You can use [this](https://matplotlib.org/examples/pylab_examples/barchart_demo.html) and [this](https://matplotlib.org/examples/pylab_examples/barchart_demo2.html) examples as a guide.\n",
    "\n",
    "Create another variable which shows the difference in budget $\\Delta b = b_{1979} - b_{1978}$. Add this variable to the dataset (find out how). Plot it. How does the budget differ?\n",
    "\n",
    "Since the numbers are different, a better comparison will be if we convert them to percentages of the total budget. Create two more variables for 1978 and 1979 and add them to the dataset. Plot these now. Also plot the difference in percentage, like you did before."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd5bc4080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd5c4dda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda\\lib\\site-packages\\ipykernel_launcher.py:29: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "D:\\Anaconda\\lib\\site-packages\\ipykernel_launcher.py:30: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "D:\\Anaconda\\lib\\site-packages\\ipykernel_launcher.py:31: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd55be908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "us_budget = pd.read_table(\"budget.dat\")\n",
    "us_budget.columns = [\"Category\" , \"FirstYear\", \"SecondYear\"]\n",
    "index = np.arange(8)\n",
    "bar_width = 0.35\n",
    "plt.bar(index, us_budget.FirstYear, bar_width, color = \"red\", label = \"For year 1978\")\n",
    "plt.bar(index+bar_width, us_budget.SecondYear, bar_width, color = \"green\", label = \"For year 1979\")\n",
    "plt.xticks(index + bar_width / 2, (\"MS\",\"SS\",\"HC\",\"DS\",\"W\",\"Edu\",\"Ene\",\"VB\"))\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "budget = np.sum(us_budget)\n",
    "FirstYear_budget = budget[1]\n",
    "SecondYear_budget = budget[2]\n",
    "\n",
    "diffs = []\n",
    "for i in range(0,len(us_budget.FirstYear)):\n",
    "    diffs.append(us_budget.SecondYear[i] - us_budget.FirstYear[i])\n",
    "us_budget = us_budget.assign(BudgetDifferences = diffs)\n",
    "\n",
    "plt.bar(index, us_budget.FirstYear, bar_width, color = \"red\", label = \"For year 1978\")\n",
    "plt.bar(index+bar_width, us_budget.SecondYear, bar_width, color = \"green\", label = \"For year 1979\")\n",
    "plt.bar(index+bar_width+bar_width, us_budget.BudgetDifferences, bar_width, color = \"blue\", label = \"Budget difference\")\n",
    "plt.xticks(index + bar_width / 2, (\"MS\",\"SS\",\"HC\",\"DS\",\"W\",\"Edu\",\"Ene\",\"VB\"))\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "# Converting all values from the dataset in procents\n",
    "for i in range(0, len(us_budget.FirstYear)):\n",
    "    us_budget.FirstYear[i] = us_budget.FirstYear[i] * 100 / FirstYear_budget\n",
    "    us_budget.SecondYear[i] = us_budget.SecondYear[i] * 100 / SecondYear_budget\n",
    "    us_budget.BudgetDifferences[i] = abs(us_budget.FirstYear[i] - us_budget.SecondYear[i])\n",
    "\n",
    "plt.bar(index, us_budget.FirstYear, bar_width, color = \"red\", label = \"For year 1978 in procents from the year budget\")\n",
    "plt.bar(index+bar_width, us_budget.SecondYear, bar_width, color = \"green\", label = \"For year 1979 in procents from the year budget\")\n",
    "plt.bar(index+bar_width+bar_width, us_budget.BudgetDifferences, bar_width, color = \"blue\", label = \"Budget difference in procents\")\n",
    "plt.xticks(index + bar_width / 2, (\"MS\",\"SS\",\"HC\",\"DS\",\"W\",\"Edu\",\"Ene\",\"VB\"))\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 3. Correlations between Variables. Alcohol and Tobacco Usage\n",
    "The dataset `alcohol_tobacco.dat` shows the average weekly household spending, in British pounds, on tobacco products and alcoholic beverages for each of the 11 regions of Great Britain.\n",
    "\n",
    "Create a scatter plot. Print the correlation coefficient. You can use the **correlation matrix** (find out how).\n",
    "\n",
    "There's a major outlier. Which is it?\n",
    "\n",
    "Remove the outlier from the dataset (find out how). Calculate the correlation coefficient once again. It should be much higher.\n",
    "\n",
    "This example is useful to show what an outlier is, and how an outlier can influence the results of an experiment.\n",
    "\n",
    "**Note:** Be careful with outliers. Sometimes they indicate human error (e.g. human height 1588 cm is obviously wrong) but sometimes they indicate important patterns in the data. Should you remove, replace, or leave them is a difficult question and should be answered separately for each dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd6ea8630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlation coefficient: \n",
      "           Alcohol   Tobacco\n",
      "Alcohol  1.000000  0.223572\n",
      "Tobacco  0.223572  1.000000\n",
      "\n",
      "\n",
      "Correlation coefficient after removing the outlier: \n",
      "           Alcohol   Tobacco\n",
      "Alcohol  1.000000  0.784287\n",
      "Tobacco  0.784287  1.000000\n"
     ]
    }
   ],
   "source": [
    "alcohol_tobacco = pd.read_table(\"alcohol_tobacco.dat\")\n",
    "plt.scatter(alcohol_tobacco[\"Alcohol\"],alcohol_tobacco[\"Tobacco\"])\n",
    "plt.xlabel(\"Alcohol price\")\n",
    "plt.ylabel(\"Tobacco price\")\n",
    "plt.show()\n",
    "\n",
    "print(\"Correlation coefficient: \\n\", alcohol_tobacco.corr())\n",
    "\n",
    "# On eye we can see that the outliner is the first point on the plot (Northern Ireland - 4.02, 4.56) - index 10\n",
    "alcohol_tobacco = alcohol_tobacco.drop(10)\n",
    "print(\"\\n\\nCorrelation coefficient after removing the outlier: \\n\", alcohol_tobacco.corr())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem 4. Simulation\n",
    "Another prediction technique based on statistics, is simulation. This means recreating a system's parameters and running the experiment on a computer instead of running it in real life. Simulation can give us many insights. It's useful for prediction, \"what-if\" analysis, etc. It's also very useful if we have very limited \"real experimentation\" resources and want to narrow down our possibilities.\n",
    "\n",
    "Let's see how we can simulate the profit of a grocery shop.\n",
    "\n",
    "The profit is dependent on the customers and what items they buy. Let's assume that the number of customers per months follows a normal distribution with mean 500 and standard deviation 20.\n",
    "\n",
    "$$ C \\sim N(500, 20) $$\n",
    "\n",
    "In the shop, there are several items, each having a different popularity. The popularity represents the probability of buying each item.\n",
    "\n",
    "| Item               | Price | Popularity |\n",
    "|--------------------|-------|------------|\n",
    "| Bread              | 0.99  | 0.5        |\n",
    "| Milk               | 2.89  | 0.15       |\n",
    "| Eggs, dozen        | 2.00  | 0.2        |\n",
    "| Chicken fillet, kg | 6.39  | 0.15       |\n",
    "\n",
    "Each customer buys *exactly one* article at random. Each customer will generate an expected profit equal to $\\text{price} . \\text{popularity}$. Total profit: sum of all profits."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_customer_profit():\n",
    "    n = np.random.random()\n",
    "    if n <= 0.5:\n",
    "        return 0.99\n",
    "    elif n < 0.65:\n",
    "        return 2.89\n",
    "    elif n <= 0.85:\n",
    "        return 2\n",
    "    else:\n",
    "        return 6.39"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_simulation():\n",
    "    days = 1000\n",
    "    profits = []\n",
    "    for day in range(days):\n",
    "        customers = np.floor(np.random.normal(500, 20))\n",
    "        profit = sum([get_customer_profit() for c in np.arange(customers)])\n",
    "        profits.append(profit)\n",
    "    return profits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd700eb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "profits = run_simulation()\n",
    "plt.hist(profits, bins = 50)\n",
    "plt.xlabel(\"Profit for 1000 days [$]\")\n",
    "plt.ylabel(\"Count\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can answer questions like:\n",
    "* What's the probability of profit less than \\$1100? \n",
    "* What's the probability of profit between \\$1300 and \\$1400?\n",
    "\n",
    "We can also change our model. Let's suppose now that one customer can take 1, 2 or 3 items, with probabilities 0.5, 0.3 and 0.2 respectively. The picked items are independent. How does this change the distribution?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_customer_profit_many_items(items = 1):\n",
    "    customer_sum = sum([get_customer_profit() for i in range(items)])\n",
    "    return customer_sum\n",
    "\n",
    "def get_total_customer_profit():\n",
    "    n = np.random.random()\n",
    "    if n <= 0.5:\n",
    "        return get_customer_profit_many_items(1)\n",
    "    elif n <= 0.8:\n",
    "        return get_customer_profit_many_items(2)\n",
    "    else:\n",
    "        return get_customer_profit_many_items(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "def run_simulation_many_items():\n",
    "    days = 1000\n",
    "    profits_many_items = []\n",
    "    for day in range(days):\n",
    "        customers = np.floor(np.random.normal(500, 20))\n",
    "        profit = sum([get_total_customer_profit() for c in np.arange(customers)])\n",
    "        profits_many_items.append(profit)\n",
    "    return profits_many_items"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd71c3dd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "profits_many_items = run_simulation_many_items()\n",
    "plt.hist(profits_many_items, bins = 50)\n",
    "plt.xlabel(\"Profit for 1000 days [$]\")\n",
    "plt.ylabel(\"Count\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x27fd731d8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title(\"Comparison of profits: 1 vs 3 items\")\n",
    "plt.hist(profits, bins = 20)\n",
    "plt.hist(profits_many_items, bins = 20)\n",
    "plt.xlabel(\"Profit\")\n",
    "plt.ylabel(\"Count\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ** Problem 5. Monte Carlo Simulation\n",
    "One common technique to apply simulations is called **Monte Carlo simulation**. It's similar to the simulation from the previous example. The main idea is to use random sampling to solve deterministic problems.\n",
    "\n",
    "Research what these simulations are. Give examples. Implement at least one case of a Monte Carlo simulation. You can use the following checklist to help with your research and work:\n",
    "* What is a simulation?\n",
    "    * How is simulation used in science?\n",
    "    * Why is a simulation useful?\n",
    "* How are statistics useful in simulation? How can we simulate unknown, random processes?\n",
    "* What is a Monte Carlo simulation (also known as \"Monte Carlo method\")?\n",
    "* A common use of Monte Carlo methods is numeric integration\n",
    "    * Define the problem. Propose the solution. Implement it and test with some common functions\n",
    "    * How does this method compare to other methods, e.g. the trapezoidal rule? Compare the performance (accuracy and time to execute) of both methods\n",
    "* Apply Monte Carlo simulation to a real-life system. There are many examples. You can see [Wikipedia](https://en.wikipedia.org/wiki/Monte_Carlo_method#Applications) or some other resource for inspiration."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ** Problem 6. Probabilistic Data Structures\n",
    "A very interesting application of probability in computer science is a kind of data structures which have a probabilistic behaviour. Examples of these are **Bloom filter**, **Skip list**, **Count-min sketch** and **HyperLogLog**.\n",
    "\n",
    "Research how one of these structures works. Or write about many of them, if you wish. You can use the following checklist as a guide:\n",
    "* What is a data structure? \n",
    "* What is a probabilistic data structure?\n",
    "    * Where does the probabilistic behaviour emerge?\n",
    "    * What advantages do these structures provide?\n",
    "* For your chosen structure, how is it constructed?\n",
    "    * What parts do you need? What are the details?\n",
    "* How does the structure work?\n",
    "    * What operations can you do?\n",
    "    * What are the typical probabilities associated with these operations?\n",
    "* Analyze the structure\n",
    "    * Analyze the runtimes for all operations\n",
    "    * Analyze the space usage\n",
    "    * Compare to a similar, non-probabilistic data structure\n",
    "    * What advantages does the new data structure have? What drawbacks do you need to be aware of?\n",
    "* Give at least one example where this structure is useful\n",
    "    * E.g. Bloom filter - spell checkers\n",
    "    * Analyze the use case\n",
    "    * If possible, implement the use case\n",
    "    * Display some metrics (e.g. % conserved space, % reduced time)"
   ]
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}