data_analysis/Statistical_Tests_Significance_testing_AB_Testing_Simpson_paradox.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4jLE0nxT_P19"
   },
   "source": [
    "<center><font size=\"4\"><b>Statistical Tests Significance testing; AB Testing; Simpson paradox </b></font></center>\n",
    "\n",
    "***"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "rgsImuK1xlp9"
   },
   "source": [
    "## Task 1: testing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "QuyjM_D5fw1z"
   },
   "source": [
    "A research study was conducted to examine the differences between older and younger adults on perceived life satisfaction. A pilot study was conducted to examine this hypothesis. Ten older adults (over the age of 70) and ten younger adults (between 20 and 30) were give a life satisfaction test (known to have high reliability and validity). Scores on the measure range from 0 to 60 with high scores indicative of high life satisfaction; low scores indicative of low life satisfaction. The data are presented below. Compute the appropriate t-test.\n",
    "\n",
    "1. What would be the null hypothesis in this study?\n",
    "2. What would be the alternate hypothesis?\n",
    "3. What probability level did you choose and why?\n",
    "4. What is your tcrit?\n",
    "5. Is there a significant difference between the two groups?\n",
    "6. Interpret your answer.\n",
    "7. If you have made an error, would it be a Type I or a Type II error? "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "w8Yi6y7bfpT9"
   },
   "outputs": [],
   "source": [
    "older_adults = [45, 38, 52, 48, 25, 39, 51, 46, 55, 46]\n",
    "younger_adults = [34, 22, 15, 27, 37, 41, 24, 19, 26, 36]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "MdE86uAc-Nor"
   },
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dHA8ewOC81z9"
   },
   "outputs": [],
   "source": [
    "mu_1 = np.mean(older_adults)\n",
    "mu_2 = np.mean(younger_adults)\n",
    "\n",
    "sigma_1 = np.std(older_adults, ddof=1)\n",
    "sigma_2 = np.std(younger_adults, ddof=1)\n",
    "\n",
    "n = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19.84962406015038"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mu_1 = 1.3\n",
    "mu_2 = 1.6\n",
    "sigma_1 = 0.25\n",
    "sigma_2 = 0.09\n",
    "\n",
    "n_1 = 22\n",
    "n_2 = 24\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "t = (mu_2 - mu_1) / (sigma_1/n_1 + sigma_2/n_2)\n",
    "t "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "eIQTt66-9gZL",
    "outputId": "15fc7699-af59-4c63-b285-5e7695f59c4d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-3.570416691943344"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t = (-mu_2 + mu_1) / (np.sqrt(sigma_2**2/n + sigma_1**2/n))\n",
    "t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "is6sckD2-MBO",
    "outputId": "cbe8c73a-7bc7-4aed-c62e-a55708ffe1e5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.10092204024096"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t_crit = stats.t.ppf(1-0.05/2, 18)\n",
    "t_crit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "bmquvzvc9Nxk",
    "outputId": "3915299c-bbda-4058-9bdc-db680cc56752"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t < t_crit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "Z_UKloXIAkyt",
    "outputId": "ff0008e8-904e-4cfe-9a6b-c1a06724f38d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.998979362829552"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = 1 - stats.t.cdf(t,df=19)\n",
    "p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "T6r9Bq1OAo56",
    "outputId": "836ac5de-3439-47a7-f7a3-8eb1e045d405"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alpha = 0.05\n",
    "p < alpha"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "1ZvHLLlq6quU"
   },
   "source": [
    "## Task 2: testing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "SaNVjkqB6quV"
   },
   "source": [
    "A researcher hypothesizes that electrical stimulation of the lateral habenula will result in a decrease in food intake (in this case, chocolate chips) in rats. Rats undergo stereotaxic surgery and an electrode is implanted in the right lateral habenula. Following a ten day recovery period, rats (kept at 80 percent body weight) are tested for the number of chocolate chips consumed during a 10 minute period of time both with and without electrical stimulation. The testing conditions are counter balanced. Compute the appropriate t-test for the data provided below.\n",
    "\n",
    "1. What would be the null hypothesis in this study?\n",
    "2. What would be the alternate hypothesis?\n",
    "3. What probability level did you choose and why?\n",
    "4. What were your degrees of freedom?\n",
    "5. Is there a significant difference between the two testing conditions?\n",
    "6. Interpret your answer.\n",
    "7. If you have made an error, would it be a Type I or a Type II error?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "V_tQjLmz6quW"
   },
   "outputs": [],
   "source": [
    "stimulation\t= [12, 7, 3, 11, 8, 5, 14, 7, 9, 10]\n",
    "no_stimulation = [8, 7, 4, 14, 6, 7, 12, 5, 5, 8]\n",
    "alpha = 0.05\n",
    "n = 10\n",
    "\n",
    "mu_1 = np.mean(no_stimulation)\n",
    "mu_2 = np.mean(stimulation)\n",
    "\n",
    "sigma_1 = np.std(no_stimulation, ddof=1)\n",
    "sigma_2 = np.std(stimulation, ddof=1)\n",
    "\n",
    "# t_crit = stats.t.ppf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "s3uenX14ESnk",
    "outputId": "b75ac188-13b0-4943-a0ec-48e8661e4c4e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "t =  0.69, t_critical =  2.10\n",
      "t <= t critical. Accept H_0.\n"
     ]
    }
   ],
   "source": [
    "t = (mu_2 - mu_1) / (np.sqrt(sigma_2**2/n + sigma_1**2/n))\n",
    "\n",
    "print(f't = {t: .2f}, t_critical = {t_crit: .2f}')\n",
    "if t <= t_crit:\n",
    "    print('t <= t critical. Accept H_0.')\n",
    "else:\n",
    "    print('t > t critical. Reject H_0.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "E6bc2rF4EsQn",
    "outputId": "f23a236c-f861-4668-83bd-b8987fb991be"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.24913574227679747"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = 1 - stats.t.cdf(t,df=19)\n",
    "p"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "c6hjIKh_8FGN"
   },
   "source": [
    "## Task 3: testing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "6x1sOCTK8FGO"
   },
   "source": [
    "Researchers want to examine the effect of perceived control on health complaints of geriatric patients in a long-term care facility. Thirty patients are randomly selected to participate in the study. Half are given a plant to care for and half are given a plant but the care is conducted by the staff. Number of health complaints are recorded for each patient over the following seven days. Compute the appropriate t-test for the data provided below.\n",
    "\n",
    "1. What would be the null hypothesis in this study?\n",
    "2. What would be the alternate hypothesis?\n",
    "3. What probability level did you choose and why?\n",
    "4. What were your degrees of freedom?\n",
    "5. Is there a significant difference between the two testing conditions?\n",
    "6. Interpret your answer.\n",
    "7. If you have made an error, would it be a Type I or a Type II error?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "S6-pvFOE8FGP"
   },
   "outputs": [],
   "source": [
    "control_over_plant = [23, 12, 6, 15, 18, 5, 21, 18, 34, 10, 23, 14, 19, 23, 8]\n",
    "no_control_over_plant = [35, 21, 26, 24, 17, 23, 37, 22, 16, 38, 23, 41, 27, 24, 32]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "71hCe0JpFOp6"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from scipy import stats as st"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "S_V8NbfwGcDf",
    "outputId": "fc31d65f-0904-4f29-9b97-e1011d0900d8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h0 = 'there is affect'\n",
    "h1 = 'there is no affect'\n",
    "\n",
    "n = 15\n",
    "alpha = 0.05\n",
    "ddf = 15 * 2 -1\n",
    "\n",
    "mu_1 = np.mean(control_over_plant)\n",
    "mu_2 = np.mean(no_control_over_plant)\n",
    "\n",
    "sigma_0 = np.std(control_over_plant, ddof=1)\n",
    "sigma_1 = np.std(no_control_over_plant, ddof=1)\n",
    "\n",
    "t_crit = st.t.ppf((1-alpha)/2, ddf)\n",
    "\n",
    "t = abs((mu_1 - mu_2)) / (np.sqrt(sigma_0**2/n + sigma_1**2/n))\n",
    "\n",
    "t < t_crit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "jJ8HpZg_PoPE"
   },
   "source": [
    "## Task 4: transactions dataset investigation\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "89zTGZluPw7T"
   },
   "source": [
    "In this task you should perform investigation of the transactions dataset of some e-commerce service that provides different kind of **service-packages** and performs on both **desktop** and **mobile**. Each record corresponds to the website visit session by certain user and could be finished by buying the product of not (see `is_paid_order` column)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 359
    },
    "colab_type": "code",
    "id": "oJUMz1RtGOrt",
    "outputId": "373d7e42-2141-4482-ae9f-35fe95532029"
   },
   "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>order_date</th>\n",
       "      <th>package_level</th>\n",
       "      <th>device</th>\n",
       "      <th>is_paid_order</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2012-01-01 00:10:32</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2012-01-01 01:02:35</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2012-01-01 01:05:35</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012-01-01 01:20:52</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2012-01-01 01:38:27</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2012-01-01 02:04:11</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2012-01-01 03:08:33</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2012-01-01 03:43:02</td>\n",
       "      <td>premium</td>\n",
       "      <td>desktop</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2012-01-01 05:35:16</td>\n",
       "      <td>standard</td>\n",
       "      <td>mobile</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2012-01-01 06:00:20</td>\n",
       "      <td>standard</td>\n",
       "      <td>desktop</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           order_date package_level   device  is_paid_order\n",
       "0 2012-01-01 00:10:32      standard  desktop              1\n",
       "1 2012-01-01 01:02:35      standard  desktop              0\n",
       "2 2012-01-01 01:05:35      standard  desktop              1\n",
       "3 2012-01-01 01:20:52      standard  desktop              1\n",
       "4 2012-01-01 01:38:27      standard  desktop              0\n",
       "5 2012-01-01 02:04:11      standard  desktop              0\n",
       "6 2012-01-01 03:08:33      standard  desktop              1\n",
       "7 2012-01-01 03:43:02       premium  desktop              1\n",
       "8 2012-01-01 05:35:16      standard   mobile              1\n",
       "9 2012-01-01 06:00:20      standard  desktop              0"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df = pd.read_csv(\"https://raw.githubusercontent.com/devrepublik/data-science-course/master/data/a-b-testing/2.3.7.%20conversion_dataset.csv\", index_col=0, parse_dates=[\"order_date\"])\n",
    "df.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "package_level\n",
       "platinum    10580\n",
       "premium     23742\n",
       "standard    65678\n",
       "Name: is_paid_order, dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby('package_level')['is_paid_order'].count()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "saGvy0IaJ7Ym"
   },
   "source": [
    "### A. Average conversion rate, SE and CI\n",
    "Calculate the average conversion rate (order payment rate), its standard error (SE) and confidence interval (CI) with significance 95%. \n",
    "\n",
    "**Hint:** You can treat the column `is_paid_order` as the observation variable that follows **Bernoulli distribution**. Then convertion rate is calculated as a mean value of the observations sample."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "rpYdG113Gdtp"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.69303"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# YOUR CODE GOES HERE\n",
    "\n",
    "alpha = 0.05\n",
    "convertion_rate = len(df[df['is_paid_order'] == 1]) / len(df)\n",
    "convertion_rate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0014585589432724341"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "se =  np.sqrt(convertion_rate * (1-convertion_rate)) / np.sqrt(len(df))\n",
    "se"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.9599877077718446"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t_score = st.t.ppf(1-alpha/2, len(df) - 1)\n",
    "t_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "lower = convertion_rate - t_score * se\n",
    "upper = convertion_rate + t_score * se"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.6901712424001254, 0.6958887575998747)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lower, upper"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "qxckB-bDLooa"
   },
   "source": [
    "### B. comparing conversion by segments\n",
    "\n",
    "Compare conversion rate by different **devices** and **package_level**. Is the difference statistically significant?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ttest_ind(data, package, device):\n",
    "    \"\"\"Returns the p_value of the t-test\"\"\"\n",
    "    \n",
    "    data = df[(df['package_level'] == package) & ((df['device'] == device))]\n",
    "    \n",
    "    alpha = 0.05\n",
    "    \n",
    "    convertion_rate = len(data[data['is_paid_order'] == 1]) / len(data)\n",
    "    se =  np.sqrt(convertion_rate * (1-convertion_rate)) / np.sqrt(len(data))\n",
    "    \n",
    "    t_score = st.t.ppf(1-alpha/2, len(data) - 1)\n",
    "\n",
    "    lower = convertion_rate - t_score * se\n",
    "    upper = convertion_rate + t_score * se\n",
    "    \n",
    "    t = (mu_2 - mu_1) / (np.sqrt(sigma_2**2/n + sigma_1**2/n))\n",
    "\n",
    "    \n",
    "    return lower, upper, t_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "OrFhd4G0LpM8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.7291396031343815, 0.7409674412259233, 1.9600748859601018)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest_ind(df, 'premium', 'desktop') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.6902208118003856, 0.697815672679569, 1.9600059198271116)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest_ind(df, 'standard', 'desktop')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.6062187569609947, 0.6453776670492224, 1.9609748336798793)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest_ind(df, 'premium', 'mobile') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.5474518212624, 0.5678567664819445, 1.9602245645790968)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest_ind(df, 'standard', 'mobile') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.9599877077718446 29.321913142542687\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def ttest_ind_2(data, arr1, arr2):\n",
    "    \n",
    "    mean_1 = df['is_paid_order'][df['device'] == arr1].mean()\n",
    "    mean_2 = df['is_paid_order'][df['device'] == arr2].mean()\n",
    "    \n",
    "    sigma_0 = np.std(df['is_paid_order'][df['device'] == arr1], ddof=1)\n",
    "    sigma_1 = np.std(df['is_paid_order'][df['device'] == arr2], ddof=1)\n",
    "    \n",
    "    n1 = len(df['is_paid_order'][df['device'] == arr1])\n",
    "    n2 = len(df['is_paid_order'][df['device'] == arr2])\n",
    "    \n",
    "    t_crit = st.t.ppf(1-0.05/2, n1+n2-1)\n",
    "    \n",
    "    t_calc = abs((mean_1 - mean_2)) / (np.sqrt(sigma_0**2/n1 + sigma_1**2/n2))\n",
    "    \n",
    "    print(t_crit, t_calc)\n",
    "\n",
    "    return t_calc <= t_crit\n",
    "\n",
    "ttest_ind_2(df, 'desktop', 'mobile')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Kr_6K5CDLs-J"
   },
   "source": [
    "### C. conversion timeseries plot\n",
    "\n",
    "Sample data by different time periods (1 hour, day, week, month) and plot the conversion rate timeseries (dependency on time) and its 95% significance confidence interval (CI). \n",
    "\n",
    "What time period of sampling would you use for conversion timeseries presentation? Why?\n",
    "\n",
    "**Hint:** in order to sample data you can use `pandas` `resample` function for grouping and then process the samples using `apply`.\n",
    "\n",
    "\n",
    "**Hint 2:** in order to visualize CI you can use `matplotlib` `fill_between` function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe9cebdaa58>"
      ]
     },
     "execution_count": 34,
     "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": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "df.groupby(df['order_date'].dt.hour)[['is_paid_order']].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe9cf33e160>"
      ]
     },
     "execution_count": 35,
     "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": [
    "df.groupby(df['order_date'].dt.day)[['is_paid_order']].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe9cf21b6a0>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAELCAYAAAA1AlaNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xl82+WV6P/PkWVb3vclsZ3YsZ2VbBACJQQCvWyXlqULkC5T6EJpC3SjM+2v3K4z906HOwOdlnZKOy2FW0gpBZq20BDWNCFA9tVZbMdJvC+xLe+2rOf3hySjyJIl27It2+f9euWF9dVX0mNhHz863/OcR4wxKKWUmh0sUz0ApZRSk0eDvlJKzSIa9JVSahbRoK+UUrOIBn2llJpFNOgrpdQsokFfKaVmEQ36Sik1i2jQV0qpWcQ61QPwlZmZaQoLC6d6GEopNa3s2bOn2RiTFey8iAv6hYWF7N69e6qHoZRS04qInA7lPE3vKKXULKJBXymlZhEN+kopNYtEXE5fKRU5BgYGqK6upre3d6qHotxsNhv5+flER0eP6fEa9JVSAVVXV5OUlERhYSEiMtXDmfWMMbS0tFBdXU1RUdGYnkPTO0qpgHp7e8nIyNCAHyFEhIyMjHF98tKgr5QakQb8yDLe/x8a9NWs0O9w8sK+GvacPkdnn2Oqh6PUlNGcvpoVfr/7LP/rhcNDtwsz4lk6N5klucksnZvMxUXpJNvGdmFMqelEZ/pqVvjz/lqKsxL49Z1reODahSydm8zRWjv/vvUEn/ntbu5/et9UD1EFcNlll034a/zXf/0XTzzxxLDjVVVVXHDBBWF/vTvvvJNnn3027M8bCp3pqxmvtq2Hd6vO8fVrFnL14hyuXpwzdF9nn4OvbNrP8Qb7FI5QjeStt96a8Ne45557Juy5HQ4HVuv4Qm04nsNDg76a8f5ysBaAD66cO+y+xFgrC3MSeeN4I4NOQ5RFL1oG8v0/H+FobXj/OC6dm8x3P7hsxHMSExPp7Oykrq6O22+/HbvdjsPh4Oc//znr168P+JjPf/7zvP7666SlpbFp0yaysrL45S9/yWOPPUZ/fz8lJSU8+eSTxMfH873vfY/ExEQeeOAB9uzZw6c//Wni4+O5/PLLRxxbb28vX/jCF9i9ezdWq5X/+I//4KqrruLxxx/nr3/9K729vXR1dfHqq69y33338dprr1FUVIQxZug59uzZw9e+9jU6OzvJzMzk8ccfZ86cOWzYsIHLLruMHTt2cNNNN/H1r3999G+wH5reUTPenw/UsSI/hcLMBL/356XF4XAaGjt0AVIke+qpp7juuuvYv38/Bw4cYNWqVQHP7erq4sILL2Tv3r1ceeWVfP/73wfgQx/6ELt27eLAgQMsWbKE//7v/x722Lvuuov//M//ZOfOnUHH9OijjwJw6NAhnn76aT71qU8NlVPu3LmT3/72t7z22ms8//zzHD9+nEOHDvHLX/5y6NPLwMAA9913H88+++zQH5tvf/vbQ8/f1tbGm2++GbaADzrTVzPcqeYuDtW08+CNSwKek5caB7jSQHNS4iZraNNOsBn5RLv44ov59Kc/zcDAALfccsuIQd9isXD77bcD8IlPfIIPfehDABw+fJgHH3yQtrY2Ojs7ue666857XHt7O21tbVx55ZUAfPKTn+Sll14K+Drbt2/nvvvuA2Dx4sXMnz+fEydOAHDNNdeQnp4OwLZt29i4cSNRUVHMnTuXq6++GoDjx49z+PBhrrnmGgAGBweZM2fO0PN7vodw0pm+mtH+fKAWEbhxxZyA53iCfnVrz2QNS43BFVdcwbZt28jLy+OTn/yk3wuvgXhq2++8805++tOfcujQIb773e8OW+RkjBlVHbx3msZXQsL5nyz9Pa8xhmXLlrF//37279/PoUOHePnllwM+Rzho0FczljGGzQdqubgwfcQZfF6a676aNg36kez06dNkZ2fzuc99js985jPs3bs34LlOp3OoOuapp54ays13dHQwZ84cBgYG+N3vfjfscampqaSkpLB9+3YAv+d4u+KKK4bOOXHiBGfOnGHRokV+z9u0aRODg4PU1dXx+uuvA7Bo0SKampqGUkkDAwMcOXIk2FsxLpreUTNWWV0H5Y2d/PMtI5fcxcdYSYuPpkZn+hHtjTfe4KGHHiI6OprExMQRZ/oJCQkcOXKEiy66iJSUFH7/+98D8MMf/pBLLrmE+fPns3z5cjo6OoY99je/+c3QhVzf9I+vL37xi9xzzz0sX74cq9XK448/Tmxs7LDzbr31Vl577TWWL1/OwoULh9JHMTExPPvss9x///20t7fjcDj4yle+wrJlE5dKk5E+nkyFNWvWGN05S4XDj/52jMe2VfLu//d+MhKH/yJ6u/E//052Uiy/uWvtJI1ueigrK2PJksDXQyKVp+JnpvL3/0VE9hhj1gR7bEjpHRG5XkSOi0i5iHzTz/0Pi8h+978TItLmdd88EXlZRMpE5KiIFIbymkqNhzGGPx+o5fKSzKABH1x5fU3vqNkgaHpHRKKAR4FrgGpgl4hsNsYc9ZxjjPmq1/n3Aau9nuIJ4F+MMVtFJBFwhmvwSgWy72wb1a09fOV/LAzp/Ly0OHaUN4/6Qp6aWpdccgl9fX3nHXvyySfDPsvfsmUL//RP/3TesaKiIp5//vmwvs5kCCWnvxYoN8ZUAojIJuBm4GiA8zcC33WfuxSwGmO2AhhjZu7nLRVRNu+vJcZq4dplOcFPxjXT7+ofxN7jICVee/B4i+Q/hO+8886kvM51110XNL8/Wcabkg8lvZMHnPW6Xe0+NoyIzAeKgNfchxYCbSLynIjsE5GH3J8cfB93t4jsFpHdTU1No/sOlPIx6DT89VAdVy3KCrmJ2lDZZlv3RA4trOrbe2nu7At+4jjYbDZaWlrGHWhUeHg2UbHZbGN+jlBm+v7+xAf6CbgDeNYYM+j1/OtxpXvOAL8H7gTOWwZnjHkMeAxcF3JDGJNSAb1T2UJTRx83rfQ7N/FrqGyztYdlc1Mmamhh9fkndxNrjeKZe943Ya+Rn59PdXU1OhmLHJ7tEscqlKBfDRR43c4HagOcewfwJZ/H7vNKDb0AXIpP0FcqnDYfqCUhJoqrF2eH/Ji5Xqtyp4N+h5OjdXYGBg1VzV0BW0yMV3R09Ji35VORKZT0zi6gVESKRCQGV2Df7HuSiCwC0oCdPo9NE5Es9+2rCXwtQKlx63c4eelwPdcszSEuZlgmMaCMhBhs0ZZpU8FT2dzJwKDrQ/Fz+2qmeDRqOgka9I0xDuBeYAtQBjxjjDkiIj8QkZu8Tt0IbDJeyT93mucB4FUROYQrVfTLcH4DSnn7+8km2nsGuGnV8I6aIxER5k6jss1jda5FRflpcTy/r1pz7ipkIa3INca8CLzoc+w7Pre/F+CxW4EVYxyfUqOy+UAtKXHRXF6SFfxkH65a/cnvtNnd7+BorZ3SnCRS4kK78FxWbycmysL9V5fyj388yO7TrVxcmD7BI1UzgbZhUDPKWxUtvH9xNjHW0beVykuNo6yscQJGdb7mzj52V7Wyu+ocu063cqSmHYfT8IlL5/HPtywP6TmO1XVQkp3IjSvm8L0/H+G5vdUa9FVINOirGaN3YJCmjj6KxnhRMy81jubOPnoHBrFFh349IFQ7K1r49vOHqGzuAiDGamFVfip3X7GAN080sed0W5BneM+xejvrSjJJiLVy/QW5/OVgHd/94LIJGbeaWTToqxnDU3njqcQZLc/j6tp7x/yHYyRPv3uGlq5+vnnDYi4uTOOCvBRira4gHWURfvZGBT39g0EvQJ/r6qfB3seS3GQAPnxhPs/treGVsgY+sGJ01zLU7KOtldWM4bkI66m5Hy3vWv1wM8bwVkULGxZlcc+VxVw0P30o4AOsyE9l0Gk4Utse9LmO1bu2LFw8JwmASxdkMCfFxnN7tYpHBadBX80YnmCdN8aZvudxNROwKvdkYyfNnX1cVpzh9/6V+a4FYQeqQwj67sqdxe6ZfpRFuGV1Hm+eaKKpY2JX6KrpT4O+mjFq2nqwCOSmjG2Jem6KDYtMzEz/rfJmAC4rzvR7f3ayjTkpNg6cDZ7XP1ZvJzMxhqyk97qHfmh1HoNO16YxSo1Eg76aMWpae8hNthEdNbYf6+goCznJtgkp23yrooWC9DgK0uMDnrMiP4WD1aEE/Y6hWb5HaU4SK/JTeG5v9bjHqmY2Dfpqxqhp6xlzPt/DVasf3vTOoNPwdmULly3wP8v3WJGfSlVLN+3dAyM+1/H6DhbnJg2770Or8zhSax/K+fvz+vFGth5tCH3wasbRoK9mjJq2njFX7nhMxKrco7V27L0OLivxn8/3WJmfCsDBmsCz/aqWLvocThbPSR523wdXzsVqEZ73c0HXMejk/7xYxl2/2cXXfr+fgUHd1mK20qCvZoRBp6G+vXfMF3E98tLiqG/vZdAZvrYGb1W48vnvWzBy0F/uuZg7Ql7fcxF3yZzhM/2MxFg2LMrihf01542/pbOPf/j1u/xiWyUXzU+jo8/B3tOto/4+1MygQV/NCA32XhxOE5b0zsCgCWsVzFsVLZRkJ5KdPPIF5pS4aBZkJoxYwVNWZyfKIpRkJ/q9/0MX5tNg7xv6Q3Owuo0P/mQ7u0+38m8fWcHjd12M1SK8cUJbJc9WGvTVjDBUox+Gmb7r+cKT1+93ONlVdS5gqaavYBdzj9XbKc5KOK/G39vVi7NJtll5bm8Nz+w6y0f+ayciwh/vuYzb1hSQZIvmovlpvHFcg/5spUFfzQieMsv8MMz0gbBV8BysbqO7fzDkoL+yIJUGex/17f5fv6xueOWON1t0FB9YOZcX9tfwj388yMWFafz5vsuHUkcAGxZlU1Znp8E++c3l1NTToK9mhJpxtmDwGAr6YarV31Hegohr1WwoVrgv5h7wM9u39w5Q09YztBI3kI0XzyMmysI9Vxbz27vWkp4Qc979Gxa5OpC+qbP9WUmDvpoRatp6SIuPJj5mfO2kEmKtpMZHhy2981ZFM8vmJpMaHxP8ZGDZ3GSsFvGb4jle776IO8JMH1wXhA9//zq+ecNirH7WLCzOTSI32cYbJya+o6iKPBr01YxQ0zr+Gn2PuSlx1IYhvdPTP8i+M20BV+H6Y4uOYlFuEgfODr+Ye6zu/J47IxlpgZqIcOXCLP5+shmHlm7OOhr01YxQ09Yz7ou4HnlpcWFJ7+w53Ur/oJP3hZjP91iRn8rB6rZhu2GV1XeQEhdNbpAqoFBsWJRFR6+DvWdCb+esZoaQgr6IXC8ix0WkXES+6ef+h0Vkv/vfCRFp87k/WURqROSn4Rq4Uh7GGNdMPzVwi4PRyHMv0BrvFoRvVTRjtcioNzdZmZ+CvddBVcv5KaZjdXYW5yYhIuMaF8C60kxX6eZxTfHMNkGDvohEAY8CNwBLgY0istT7HGPMV40xq4wxq4CfAM/5PM0PgTfDM2SlztfaPUDPwGDY0jt5qXF09jmw9zjG9TxvVbSwsiCVxNjRXWdYWeBemeuV13e62y8s8bMSdyySbdFcqKWbs1IoM/21QLkxptIY0w9sAm4e4fyNwNOeGyJyEZADvDyegSoVSG2YavQ93qvVH3uKx947wMHqtpBLNb2VZidii7aw32tlbnVrD139g3577ozVhkVZHNXSzVknlKCfB5z1ul3tPjaMiMwHioDX3LctwL8D3xjfMJUKrHqcffR9vVerP/agv+vUOZyGUefzAaxRFi6Ym8JBr5W5ZUMbp4Rnpg+wYWE2oKWbs00oQd9fAjFQsvMO4FljzKD79heBF40xZwOc73oBkbtFZLeI7G5q0h9ANTrj3THL19yhWv2xl23uKG8h1mrhwnlpY3r8yoJUjtS2DzVGO1bXgQgszPHffmEslsxJIic5Vks3Z5lQgn41UOB1Ox8ItFPDHXildoD3AfeKSBXwf4F/EJF/9X2QMeYxY8waY8yarKyskAaulEdNaw9x0VGkxUeH5fkyE2OItVqoDbAqNhRvVTSzpjBtzBuVr8hPoXfAyYkGV23+sXo7hRkJ416H4E1LN2enUIL+LqBURIpEJAZXYN/se5KILALSgJ2eY8aYjxtj5hljCoEHgCeMMcOqf5Qaj5q2bvLS4sJS1QKuYJiXOvayzZbOPo7Vd4yqPt/XUJtld4rnWIAe+uO1YVG2lm7OMkGDvjHGAdwLbAHKgGeMMUdE5AcicpPXqRuBTWa8dW5KjVI4a/Q95qbGUT1CTr/PMUhLp/9OnG9XngPGls/3mJ8RT0pctLt3j4Oqlq4Re+6M1bqSTKK0dHNWCemzojHmReBFn2Pf8bn9vSDP8Tjw+KhGp1QIatt6h3rWhEteahyvjRAIv7JpPy8drmdxbhJXLMziitKsoXTOWxXNJMZaWZGXEvDxwYgIK/JT2H+2nRMNnRgT2krc0UqJi+aiea7SzX+8fnHYn19FnvAlCJWaAt39Ds519Yd9pp+XFkdTRx+9A4PD8vJvV7bw0uF6rl2aQ2efg8d3VPHYtkps0RYuXZDBkVo7a4vS/fa9GY2V+an8/M0K9p1xbXgSrOfOWF25KIuHthyn0d4btOe/mv406KtpLdw1+h6eCp669l6KMhOGjjudhn/5axlzUmz858bV2KKj6Opz8HZlC9tONLHtZDNNHX1cvTh73GNYkZ/CoNPwx73VJMREjbttdCAb3EH/jRNN3LamIPgD1LSmQV9Na0M1+mEOiJ4/IrVtPecF/T8dqOFQTTv/cdvKoU8ACbFW3r8kh/cvyQFcF3LTQuyqOZJV7pW5h2vsXDgvFYslPBeqfS2dk0x2UixvHtegPxtowzU1rYVrxyxfnlm1dwVP78AgD/3tOMvzUrhlld/1iYBrr9pwBOjsZNtQc7VwLsry9V7pZpOWbs4CGvTVtFbT2oPVIuSEORedk2xDhPMqeP57+ylq23v59o1LJmzW7WuFe8erJRNQrultw6Js7L0O9o2wKbuaGTToq2mtpq2H3BQbUWEOwjFWCzlJtqGZfnNnHz9/o4JrluaEvAtWOHiar03kTB8Y6hG0X+v1ZzzN6atprbatZ9xbJAaSlxY3dKH44a0n6B0Y5Fs3TG5Z4y2r82iw9w4t1pooaQkxpMVHc6qla0JfR009nemraa2mtYf8CQr6c9199U82dPD0u2f4xKXzWZAVvt43ochLjeMHN19AjHXif1XnZyRQ1axBf6bToK+mrYFBJ/X23rBX7njkpcZR197Dv7xYRkKslfvfXzohrxMpijI16M8GGvTVtFXf3ovThL9yxyMvLY6BQcMbx5u496oS0hPGX4YZyQozEqht76V3YDD4yWra0qCvpq1wt1T25Ukb5afF8anLCifkNSJJYaZru8kz58beUlpFPg36KiINOg1f+t3eERuB1YR58xRfC3OTiLVaePDGJWNukTydFGa4FqGd0hTPjKbVOyoi7T/byl8P1VHX3sOGRf5bGngqayaseic1jsPfv47ocfbQmS4K3SuPNa8/s82On2Y17bx8tAGAvWfaOOneSMRXTVsPmYkxEzoLny0BH1wdN9MTYqhq0fTOTDZ7fqLVtGGM4eUjDazIT8FqEX6/y/9umxPRR3+2K8yIH9VM/8mdVWw/2TxxA1Jhp0FfRZyKpk5ONXfx0TUFXLM0h+f21dDvGN4Tpqa1Z8Iu4s5WhRkJVIW4QKu9Z4D/9acjfPLX7/DjV07idOr+SdOBBv0I9Mzus3x50z4a7WPfo3U623LEldq5ZkkOt11cwLmufl4pazjvHGOMzvQnQGFmAnXtvfT0By/bPF7vSrstm5vMw6+c4HNP7Ka9Z2Cih6jGSYN+BPrjnmr+tL+W63/8d1452hD8AdPAaHbR3Hq0gZX5KeSm2LiiNIs5KbZhKZ7mzn76HE4N+mHmuZgbStlmWZ0dgF/9w8V8/6ZlvHmiiZt/up1j9fYJHaMan5CCvohcLyLHRaRcRIZtbC4iD4vIfve/EyLS5j6+SkR2isgRETkoIreH+xuYiapaurisOIOcZBuffWI3D75wKKSZV6Rq6exj5fdfZsuR+qDnNth72X+2jWuX5QIQZRE+elE+2042DdXlw8RX7sxWRaMo2yyrs5MWH01OciyfuqyQTXdfSnf/ILc++hZ/2l8z0UNVYxQ06ItIFPAocAOwFNgoIku9zzHGfNUYs8oYswr4CfCc+65u4B+MMcuA64FHRGRiO0dNc119DhrsfawryeSFL13GZy8v4v+9fYabfrp9aGY13Ryts2PvdfDIKyeDzvi3uj/ZXLs0Z+jYR90bezy7u3ro2EQvzJqt5rsXaIWS1y+rs7NkTjIirg6nawrT+ct9l3NBXjJf3rSf//1i2YSOVY1NKDP9tUC5MabSGNMPbAJuHuH8jcDTAMaYE8aYk+6va4FGIGt8Q57ZPL9sRZkJxFqjePADS3ni02tp6xng5p/u4L+3n5p2F8wqm1zfU1mdne3lI1d6bD3aQFFmAiXZ7zU2K0iPZ11xJs/sPjv0vXsWZuWnxk/QqGenZFs0GQkxQSt4Bp2G4w0dLPFp+ZydbOOpz13KRy7K57FtlZzR8s+IE0rQzwO8E6rV7mPDiMh8oAh4zc99a4EYoGL0w5w9PB+rPasjAa5YmMXfvryeKxZm8sO/HOXPB2unanhjUtnUSUJMFNlJsfzizcqA53X0DvBWRTPXLM0Zmj163H5xATVtPeyocP3RqGnrITHWSnKcri8Mt8LM4BU8p5q76B1wDgv64FrbsHGt69NZeZP/NRZq6oQS9P3tThFoqnkH8Kwx5rwEtIjMAZ4E7jLGDKu9E5G7RWS3iOxuamoKYUgzl2eG5emD4pGRGMtjn1xDTJSFo9MszVPZ3EVxdiJ3rStie3kzh2va/Z73xvEmBgbNeakdj2uX5ZAaH80m9wXd6lZX5Y7vHwc1foUZCVQ1jzxD96Qal8zxv6PXgkzXJzXPpzwVOUIJ+tWA927J+UCgqeYduFM7HiKSDPwVeNAY87a/BxljHjPGrDHGrMnKmt3Zn1PN3eQm24iPGT6DtViE/LQ4qs/1+Hlk5Kps6qIoM4GPXTKPxFgrj23zP9t/+WgDmYkxrJ6XNuy+WGsUt67OY+uRBs519bvKNTWfPyEKM+Kpt49ctllWZ8dqkfPScN7SEmJIT4ihoqlzooapxiiUoL8LKBWRIhGJwRXYN/ueJCKLgDRgp9exGOB54AljzB/CM+SZ7VRz57BZvrf89Php1QWxp3+QmrYeFmQmkhIXzca1Bfz1UB1nfb6HfoeTN4418v7FOQG3Prz94gL6B508v6/GvWNWePfFVS5DPXhGSPGU1dkpyU4k1hq4BcaCzAQqdKYfcYIGfWOMA7gX2AKUAc8YY46IyA9E5CavUzcCm8z55Rm3AVcAd3qVdK4K4/hnnKqWbooyA+/ONC89jrOt0yfoe65RLMhyBZK71hUhuDYZ9/Z2ZQsdfQ6uXTY8teOxODeZlQWpPLGzivaeAfL0Iu6EKHIH/dMjBv3hF3F9FWclUqkz/YgTUp2+MeZFY8xCY0yxMeZf3Me+Y4zZ7HXO94wx3/R53P8zxkR7yjnd//aH91uYOdq7BzjX1U/RCDP9grR42roHsPdOj5WPlc2uX3pP0J+bGsdNq+by+11nae3qHzrv5aP1xMdEsa4kc8Tnu31NAafdFSGa3pkY8zNcP3+nAuT1W7v6qbf3BsznexRnJ9Dc2U979/T4WZ0tdEVuBPFsSu1dueNrXrrrF9I3PRKpPBfyPLNHgLuvWEDPwCD/7+3TADidhleONnLlwqygHTM/uHIOce5zdDXuxEiyRZOZGLhs872LuCPP9D0XcyuadbYfSTToR5Aqn1SIPwXTLuh3Mjfl/AvTi3OT2bAoi9/urKJ3YJBDNe3U23u5xk/Vjq8kWzQ3rpgDuHa0UhOjMCNhaBLiy1M9tjg3SHrHfZG3olGDfiTRoB9BKpu7sMh7gd2f94L+9KjgqWzuYkHW8GsUd1+xgObOfv64t5qXj9YTZRGuXux/sxRfD1y7iH++5QJykvVC7kQpzEwImNMvq+sgMzGWrKTYEZ+jIC2O6CihUjdliSga9CNIVXMXc1PjRqyISImLJtlmHVUFT3ljBx1TcA3AGENlU5ffTy7vW5DB8rwUfvX3U2w50sAlRemkxoe28Xhuio1PXDo/3MNVXooyE2iw99Hd7xh2n6v9wsj5fABrlIX5GQk6048wGvQjSFVL13m570AK0uNDruBxDDq55dG3+Olr5eMd3qg1dfTR2edggZ/vSUT4/JULONXcRXljZ0ipHTV5PBdzfRdpDQw6KW/sZGmQfL7HgswEnelHGA36EcIYw6mm0IL+vFHU6p85101nn2NKVvF6arT9pXcArl+WS0G6Ky+vQT+yeIoJfGv1K5o66R/0337Bn+LsRE63dOEYHL4JjpoaGvQjREtXPx19jpBn+tWtPSE1Xit3f7Qun4KP2L7lmr6sURa+f9MyvrChmPw0rbmPJIEWaIVaueOxIDOBgUHD2dbpcQ1qNtCgHyHe67kTWtDvdzhp7OgLeu5Jd7Cva++d9Lx+ZVMXtmgLc1MCV9lcvTiHf7p+8SSOSoUiMdZKVlLssLLNY3UdxERZRqww86YVPJFHg36E8OQ9i0ao0fcocJcqhpLX9/5lm+wl8ZVNnRRmJGAJ0FZBRTbXJunn/4wdrbNTmpNIdFRooaPY03hNa/Ujhgb9CFHV3IXV3VAtGM8CrVB6lZ9s7Bx6zpMNk9vmtrK5i+IA+XwV+fzV6ofSfsFbSrxroVdFo17MjRQa9CPEqeYu5qXHYw1hBpWXFodI8Jm+02kob+zk6sXZxFgtk5rX73MMcvZcd8hpABV5CjMThiqwwFWN1dzZN6qgD64L+dptM3Jo0I8Qp5q7Qsrng6vNcG6yLWgFT217Dz0DgyzKTWJBZsKkBv0zLd04zciri1Vk8228FqyHfiDFWVq2GUk06EcAp9NwuqU7pModj4K0+KB99T0XcUuzkyjJThy6PRmGyjVH6BiqIptvrb4n6Idao+9RnJXIua7+8xrsqamjQT8CNHT00jMwGPJMH1wVPMFm+hVDQT+R0uwkzrZ20zsQeGOMcApWrqkin2+tflmdnTlJfPKrAAAgAElEQVQptpBXTnt4fgb0Ym5k0KAfAU6NonLHoyA9joaOXvocgYP4yYZOMhJiSEuIoSQ7EWOYtNxqZVMXWUmxJNmiJ+X1VPglxFrJTood+vksq+tgce7oUjvA0MV8vZgbGTToRwDPx+eiUcyK56XHYwzUjLDopbypc2g7u9Ic138nK69f2dTpt/2Cml48jdf6HINUNHWO+iIuQH5aPDFRFm2xHCE06EeAU82dxFotzBlF10hPt81AKR5jDCcbOoaCfWFGAlEWmbygH6C7pppeCjPiOdXczcmGThxOM6agH2URCjPjdaYfITToR4BTzd2jXsQ0tJlKgJl+U0cf9l4HJe7AG2O1MD8jnpMNEx/0z3X109Y9QLHm86e9wswEmjv72F11Dgi9/YKv4qxEzelHiJCCvohcLyLHRaRcRL7p5/6HvfbAPSEibV73fUpETrr/fSqcg58pqlq6RtwM3Z+sxFhirJaAm6l4ZvSlOe/lYEuzEznZOPELtE7pRdwZw3Od6aXD9diiLaOqMPO2ICuBMy3dDGjjtSkXNOiLSBTwKHADsBTYKCJLvc8xxnzVswcu8BPgOfdj04HvApcAa4HvikhaeL+F6W3QaTjT0j2qyh0Ai0UoSIsLGPQ95ZmenL7n69Mt3fQ7JvYXT8s1Zw7Pz+W7VedYlJNE1BhbahRnJeJwmlHtA6EmRigz/bVAuTGm0hjTD2wCbh7h/I3A0+6vrwO2GmPOGWNaga3A9eMZ8ExT29ZD/6BzTBc9RyrbLG/sJMnmqr7wKM1OwuE0AXdECpfKpi6io0JrKaEim6dW35ixp3bgvfba2nht6oUS9POAs163q93HhhGR+UAR8NpoHztbeVYqjrQZeiAFafEjzPQ7KM1OROS9mZln1j/Ri7QqmzqZn5EQUksJFdniY6zkJLsmDuML+q6f78lu+qeGC+W30t/nuUCN3O8AnjXGeIrHQ3qsiNwtIrtFZHdTU1MIQ5o5PK1rx5IrnZcej73XQXv38JbJ5Y2d56V2wPURW2TiyzYrm7u0XHMG8UxIxhP0k23RZCXFUjnGdSKNHb3UtmlP/nAIJehXAwVet/OB2gDn3sF7qZ2QH2uMecwYs8YYsyYrKyuEIc0cp5q7SIiJCrrJtD+eXad8G6+1dvXT3NlPafb5C2niYqLIT4ub0Jm+Y9DJ6RYt15xJPBOSxaPsueOrOCthTIsDz3X1c9NPdvCxX76NMcE3DlIjCyXo7wJKRaRIRGJwBfbNvieJyCIgDdjpdXgLcK2IpLkv4F7rPqbcPI3WvNMwoQpUq1/u/sUqyRkeeEuyEid0pl/d2sPAoNGZ/gzy8Uvm860bFpM8ztXVxVmJVDR1jSpwG2P4xh8OUG/vpaqlm71nWsc1BhVC0DfGOIB7cQXrMuAZY8wREfmBiNzkdepGYJPx+j9qjDkH/BDXH45dwA/cx5Sbq1xzbAHSE/R98/qeoF7iZ7ZdmpNERVMngyFstTgW2nNn5lmen8Lnrywe9/MsyEqkvWeAc6NovPbrHVW8eqyRb1y3CFu0hef31Yx7HLOdNZSTjDEvAi/6HPuOz+3vBXjsr4Ffj3F8M1q/w0l1aw83rZw7pscn26JJjY8eNtM/2dBJXHQUeanDq2dKshPpdzg5e270ZaKhqAyyGbqavYqHGq91kZEYPJ15sLqNf32pjGuW5vDFDcUcr+/gLwfr+M4HlhFj1SKBsdJ3bgqdbe1m0GnGVLnjUZAWP2xV7snGDoqz/a/w9VzcnagUT0VTF6nx0aQnjK4To5r5ikdRttnRO8B9T+8jKzGWhz6yAhHh1tV5tHUP8OaJ2VXsEW7TJugfr+/gaK19qocRVqPZDD2QeenDyzYrGjuHXcT1mOiyTW20pgKZmxpHrNUSdEMVYwzffv4w1a09/Hjj6qFWzutLM8lIiOH5fdWTMdwZa9oE/e/86TDffuHQVA8jrDwta8cTJPPT46hp7RnK0Xf2Oaht7x1WrumRbIsmJzl2wtoxaKM1FUiURSjKTAg6039m91k2H6jlq/+jlIsL04eOW6MsfHDlXF4pa6S9Z3iZsgrNtAn6de1TX6f7D79+l5+/URG25zvV3EVKXDRp40iFzEuPp3/QSYO9F3jvo3OgoA+ulbkTsTKyo3eApo4+vYirAioOsl/uyYYOvrv5CJcVZ/CFDSXD7r91dR79Did/O1w3kcOc0aZF0DfGUG/vpamjb8KqToJp6uhj24kmfvzqCRo7esPynFUtXWNuYOVRkHZ+Bc9Jr92yAinJdpVtBiqdM8bw2d/u4v6n942qT0+l9txRQRRnJXC2tcfv5j/nuvq596l9JMRYeeT2VX77/KzIT2FBZoJW8YzDtAj67T0D9DucOA00d/ZNyRg89cG9A05+9np4ZvunmsYf9Of51OqfbOwgJsoydNyfkuxEuvoHqWv3/8fr9eONvFLWyOYDtdz39N6QOyN6yjW1pbIKZEFW4lCTQY+O3gEe3nqCK/7tdcqbOvn321aSHWBvCRHhltV5vF15jhpdoTsm0yLoN9jfC/T1AQLVRNt3po3oKFcFwVPvnBn3D1zvwCC17b3jqtwB18Uxkff66lc0dlKUOXLfm9IRLuYaY3jklZMUpMfx4I1L2HKkgS9v2ocjhMBf2dSFRWBexujaRKvZY6iCp6mL3oFBHttWwRX/9jo/fvUk60sz+duX17NhUfaIz3HLKlf7rj/t19n+WEyToN/r9+vJtPdMK0vnpvDAdYsA+MmrJ8f1fMfqXRdSR9tH31eM1cLclLjz0jv+VuJ6G6rgaRh+MffVskYOVrdz31WlfHb9Ah68cQkvHqrnq88cGDHwH65p52+H6ylIjyfWGjWO70jNZJ4tQZ98u4orH3qd//3iMZbnp7L53nX8/BMXnbf/QyDzMuJZMz+N5/fWaFuGMZgWQb9+ioO+Y9DJweo2Vhekkpcax8cumccf9lQPlVyOltNp+N9/LSPZZmVdSea4x5fv7qvfOzDI2XPdflfiestIjCU9IWbYBTVjDI+8eoJ56fHceqFrNvXZ9Qv41g2L+fOBWh74w4Fh11SqW7v5yqZ9fOAn22nu7OOb1y8e9/ejZq7EWCtzUmzsKG+hIC2eTXdfyhOfXsuK/NRRPc8tq/M42djJ0bqZVcY9GUJakTvVGt2B3iLnp3omy7H6DnoHnFw437X/yxevKmbTrjM88soJHrlj9aif76l3z/Bu1Tn+7cMryAxhZWIw89Lj2XayicqmLpzmvU3QR1KSnThs68RXyho5XGPnoY+sINorPfT5K4txOA0PbTmOxSI89JGVdPY5+Nnr5fzmrSoE+OKGYu7ZUDzu/ixq5nv49lX0O5ysL80cU88pgBuXz+H7fz7CC/tqWDY3JcwjnNmmRdBvsPeRGh+NzRp13qx/suxzX8RdXeCajWQn2bjzsiJ+sa2CL2woYVFu6N0H69p7+NeXjrGuJIOPrskPy/gK0uNpsPdxuLYdIODCLG8l2Yn89WAdxhhExJ3LP8H8jHhuXT18y4MvXVXCoNPwH1tP0NTRx6Gadtp7BvjQ6ny+fu1C5vpp+aCUP5cuyBj3c6QlxLBhUTZ/2l/LN29YMuYdvWajaZHeabD3kpNkIyfFNiXpnb1n2shKij1vJ6jPX7GAxBgrD289EfLzeFYaOpxO/s+tK8Y8y/HlqdR583gTFgntOkFptqv5VXOnq/nVy0cbOFJr576rSwNeBL7//aXc//5S/n6ymeV5Kfzlvsv599tWasBXU+LW1Xk0dvSxs6JlqocyrUyTmX4v2cmxxEVHUTXBW/35s+9MK6sLUs8L0mkJMXxmfRGPvHKSQ9XtLM8P/hFz84FaXjvWyIM3LglrhYunr/62E00UZiSEdCHV82ngZGMHGQkxPPLKSYoyE7hl1cjN3752zUI+ccm8gCV1Sk2Wqxdnk2Sz8ty+ai4vHf+1sdlimsz0+8hNtpGbYpv0nH5LZx9VLd1D+Xxvn7m8iNT4aP7vy8eDPs+5rn6+/+ejrCxI5a51RWEdo2eBVkefY8SVuN68G6+9fLSesjo7911dEtIWhxrwVSSwRUdx4/I5bDlcT3e/Y6qHM21EfNAfdBqaOvvISbaRk2yjvWeA3oHhq/kmyv6zbcB7+XxvSbZo7rmymDdPNLGrauRtAn74l6PYewb40YeXhz3/mJUUS6y71WyoQT8nOZakWCsnGjp45JWTLMhMGHOLZ6Wmyi2r8+jqH2Tr0YapHsq0EfFBv6XL1XohJzmWHPcMczLz+nvPtBJlkYAlZZ96XyFZSbE89LfjAZtAvX68kef31fDFq0pYnDv2fUYDEZGhDVVCqdzxPKY4O5Hn99ZwrL6D+98fOJevVKRaW5jO3BQbL2hbhpBF/G95Q7srnZOdbCMn2VXeOJmrcvedaWPJnCTiYvznyeNiovjy+0t5t+ocq37wMv/zx3/ne5uP8NKhOpo7++jsc/Dt5w5Rkp3Il64a/+5DgXgu5pZkhV5JVOpux1CclcAHdZavpiGLRbh5dR7bTjZPWYuW6SbiL+R6ZvW5yTbi3YG3oWNy/ucOOg0Hzrbx4YtGLq38+CXzKM1O5O3Kc7xb1cKmXWd4/K0qANITYmjt7ufZey6b0JWqnqBfnB16WwfPp4L731+qJW9q2rp1dR4/f6OCvxyo5c4wXy+biUIK+iJyPfBjIAr4lTHmX/2ccxvwPcAAB4wxH3Mf/zfgRlyfKrYCXzajWDvd4O5omZNsG5ptN0zSTP9EQwdd/YOsnjfyakER4ZIFGVyyIAMopd/h5HBtO++eOsfuqnOsKUznIj8XgsPpzssKWVmQQnxM6H/HP3RhPrHWKD6wQmf5avpamJPE0jnJPL9fg34ogkYIEYkCHgWuAaqBXSKy2Rhz1OucUuBbwDpjTKuIZLuPXwasA1a4T90OXAm8EeoAG+x9iEBmYgxRFiEuOvQFWhVNnRyuaecDK+aOaSbr6ax54bzRBewYq4UL56W5HheGDaVDUZiZMOoduDITY/nUZYUTMyClJtGtq/P4lxfLXDu36SY+Iwolp78WKDfGVBpj+oFNwM0+53wOeNQY0wpgjGl0HzeADYgBYoFoYFSX2Rvae8lMjMUaZUFEyEmODflC7iOvnOTLm/bz0f96i/Ix7BS170wb6QkxI7YpVkpNvZtWzUUEXthfO9VDiXihBP084KzX7Wr3MW8LgYUiskNE3nangzDG7AReB+rc/7YYY8p8X0BE7haR3SKyu6np/E2PGzp6yfWqC89JDn1V7qnmTgrS46hs7uJ//ng7P33tZMi94cE1079wXmrYVs4qpSZGTrKNdcWZvLBPO28GE0rQ9xfxfN9VK1AKbAA2Ar8SkVQRKQGWAPm4/lBcLSJXDHsyYx4zxqwxxqzJyso6774Ge99Q1Q54gn7wC7nGGKqau7l6UTavfO1KrlmWw/99+QQ3/3QHh2vagz6+rbufyqYuVo8ytaOUmhq3rM7jzLlu9p5pm+qhRLRQgn41UOB1Ox/w/QxVDfzJGDNgjDkFHMf1R+BW4G1jTKcxphN4Cbh0NANstPeetwI0N8VGvb036F/z5s5+OvscFGYmkJkYy6Mfu5BffPIimjr7uPnRHfzb34753bLNY59nUVaQi7hKqchw3bIcbNGWoDX7xhie2XV2yvfcniqhBP1dQKmIFIlIDHAHsNnnnBeAqwBEJBNXuqcSOANcKSJWEYnGdRF3WHonkD7HIC1d/eQkvRf0s5Ni6Xc4Ay6E8vD06PG+uHndslxe+eqVfGh1Hj97o4Jv/fFQwMfvO92KRWDlKPt8K6WmRpItmmuW5vKXg7Uj7u386x1V/OMfD/LdzUcmcXSRI2jQN8Y4gHuBLbgC9jPGmCMi8gMRucl92hagRUSO4srhf8MY0wI8C1QAh4ADuEo5/xzq4Jrc9fi5Ke+ld3JTXH8AglXwnHJvcFLksx1hSnw0D310Jfe/v5Tn9tXw/L5qv4/fd7aNRbnJJMRG/FIGpZTbravn0to9wLYTTX7v33emlf/zomsDo1fKGoZtJDQbhLQi1xjzojFmoTGm2BjzL+5j3zHGbHZ/bYwxXzPGLDXGLDfGbHIfHzTGfN4Ys8R939dGMzhP7j7b50Ku932BVDV3YbXIee2Qvd1/dQlrC9N58PnDw3bAcjoN+8+0aWpHqWlmfWkW6QkxPO9n/9y27n7ufWofuSk2nv/SOqKjLPzq76emYJRTK6LbMHh2zPJO73gqeYIt0Kpq6aIgPT5gPxlrlIWH71iFNcrC/Zv2nfdxsLypk44+x6jr85VSUys6ysIHV8zhlaMN2HvfSwEbY3jgDwdo7Ojl0Y9dSHFWIh++MJ8/7q2ede0bIjroe1I43tU72e6vg5VtnmrupjBIz/q81Dh+9OEVHKxu59+92iPvPe3eKUtn+kpNO7eszqPP4eRvh+uHjv3y75W8UtbI//c/l7DS3TH3s+uLGBh08sTO01M11CkR0UG/wd5HdJSQnhAzdCzWGkVafPSIOX1jDKdbukJaoXr9Bbl84tJ5/GJbJW+684D7zrSREhfNglGucFVKTb1VBakUZsQPVfHsOX2OH/3tODdckMudXivQi7MS+R9LcnhyZxU9/ZPXrn2qRXTQb7T3kp1kG7Y4KlitfmNHH939gxSFGLQfvHEpi3KS+Poz+2nq6GPvmVZW66IspaYlEeGW1XnsrGzhaK2de5/a5/pU/5HhW5TefcUCWrsHeHbP2QDPNvNEdNBv6Og9L7XjEWxVrqdypzAjtKBvi47iJx9bTUevg3uf2kt5U6fm85Waxm5ZlYcxcMdjO2np7OdnH7+QZFv0sPPWzE9j9bxUfrX9FIPO2bGSN6KDfn1771C1jrfcZNuI6R1PNU6oM31wder7zgeX8s6pcxij+XylprPCzARWz0vF3uvgf31gCRfk+d/DWkS4e/0CTrd0s/Vovd9zZpqIDvqN9j6/QT8nOZbmzj4cAfronGrpIibKwtxU/+WagXxs7TxuuCCXGKtl6GKPUmp6+u4Hl/HgjUv4xKXzRzzv2mW5zEuP5xfbKmdF356IDfpdfQ46+hz+g36KDWOgKUCpVVVzF/My4kfdTllEeOSOVbz05fV+PwoqpaaPVQWpfHb9gqDX5qIswmfXF7HvTBt73JV7M1nEBv1G92pcvzn9pJEXaFU1d4ecz/cVa42iWPtxKzWrfPSiAtLio3lsW+VUD2XCRWzQ9+yD6zen72nF4GeBltNpqGrpoihTe+ArpUITFxPFJy+dz9ayBipneGuGiA36jR2Bg75ngZbnHG/19l76HM5R7yKllJrdPvm+Qldrhu0zuzVDxAb9Bj+rcT0yE2KJsojfmX6gRmtKKTWSrKRYV2uGPdW0dvVP9XAmTAQH/T7iY6JI9NPl0mIRspNi/eb0h2r0daavlBqlT102nz6Hkz/u9d99dyaI2KBfb3fV6Ae68h5ogVZVcxexVst5WywqpVQoFucms2Z+Gr9758yMLd+M2KDvasEwPLXjEWiD9KqWLgozErCMslxTKaUAPn7pPE41d7GzomWqhzIhIjboN9j7hqp0/Am0KvdUcxeFWrmjlBqjGy6YQ2p8NL9758xUD2VCRGTQN8bQYPffgsEjJ8VGR6+D7n7H0LFBp+HsuR7N5yulxswWHcVHL8pny5F6vxWC011IQV9ErheR4yJSLiLfDHDObSJyVESOiMhTXsfnicjLIlLmvr8w2Ou19wzQ53COnN7xs0Crtq2H/kGnVu4opcZl49p5OJyuDdRnmqBBX0SigEeBG4ClwEYRWepzTinwLWCdMWYZ8BWvu58AHjLGLAHWAo3BXtMTyEea6ftboKWVO0qpcFiQlci6kgyefvfsjOu+GcpMfy1QboypNMb0A5uAm33O+RzwqDGmFcAY0wjg/uNgNcZsdR/vNMZ0B3tBzwXakXL6OX4WaFW1jL67plJK+fPxS+ZT09bDmyeCzlOnlVCCfh7g/Rmn2n3M20JgoYjsEJG3ReR6r+NtIvKciOwTkYfcnxxG1OBnb1xfnk8BvjP9+JioEdNCSikVimuW5pCVFMvv3p5ZF3RDCfr+ah99P+9YgVJgA7AR+JWIpLqPrwceAC4GFgB3DnsBkbtFZLeI7G5qahoK+tl+VuN6JMZaiY+JOi+nX9XcxfyMBN3xSik1btFRFu64uIDXjjdS3Ro0QTFthBL0q4ECr9v5QK2fc/5kjBkwxpwCjuP6I1AN7HOnhhzAC8CFvi9gjHnMGLPGGLMmKyuLBnsfKXHR2KIDfygQEXJ9FmhVtXRrozWlVNjcsXYeAvx+Bl3QDSXo7wJKRaRIRGKAO4DNPue8AFwFICKZuNI6le7HpolIlvu8q4GjwV6wwd4b0orabK8FWo5BJ2fPjb2lslJK+cpLjeOqRdls2nWWgQCbNk03QYO+e4Z+L7AFKAOeMcYcEZEfiMhN7tO2AC0ichR4HfiGMabFGDOIK7XzqogcwpUq+mWw12zo6BsxtePhvUCrurUHh9No5Y5SKqw+fuk8mjr6eOVow1QPJSyGdzPzwxjzIvCiz7HveH1tgK+5//k+diuwYjSDamjvpTQ7M+h5Ock2Gu19GGM4pZU7SqkJcOXCbPJS4/jdO2e4YfmcqR6OX/vOhL7jV0SuyG3q7PPbUtlXTrKN/kEnrd0DQ5uha3pHKRVOURZh49oCtpc3D60FCtXhmnbeqmieoJG95y8H60I+N+KCvsNpGHSakHL6nrLNBnsvVc1dJMZayUyMmeghKqVmmdvWFGC1CJt2ja588/5N+/jYL9/hrt+8S8UE7sh1sLot5HMjLugPOFwXS7JDCPq5Ka5PA/X2Xk61dFOYGa/lmkqpsMtOtnHpggxePxb6Qq3ath4qm7pYV5LBrqpWrnt4G//8l6PYewfCOrZBp+FwjT3k8yMv6DtdQX+kFgwe2Z7+O+2umX5Rpm5orpSaGJeXZnKiodNvS3d/dpS70joP3riU1x/YwIcvzOe/d5ziqofeYNO7Z8LW3qGiqZOegcGQz4+4oO8YdL0Roeb0wVW5U93aTVGG1ugrpSbG+lJXccnfT4aWo99R3kxmYgyLcpLISorlRx9ZweYvXU5RZgLffO4QNz+6PSzbMh44G3pqByIw6A8MOhGBrMTgQT/GaiEjIYbdp8/hNNpoTSk1cZbkJpOREMP2k01BzzXGsL28hcuKM8/b0Gl5fgp/uOd9/OjDyzlcY2drGMpAD9W0kxATtLvNkAgM+obMxFisUaENLTvZxt4zrr90GvSVUhPFYhEuL81ke3kLziCpmRMNnTR39nF5yfDScxHhwxfmE2O1UB6Gi7sHq9u5IC8l5PMjLug7Bp0hpXY8cpNj6Xdf/NU++kqpiXR5SSbNnX0cq+8Y8bzt7nz+ulL/642sURYWZCZQ3ji+oN/vcHK0zs6K/Gkc9AecZsTumr48ef2UuGjSErRcUyk1cdaXujrKbC8fOcWzo7yZoswE8lLjAp5TnJ047qB/oqGDfoeTFfmpIT8m4oK+Y9BJzgh99H15gr6mdpRSEy03xUZpduKIF3MHBp28XdnCupKMEZ+rJCuRs63d9I6i8sbXoZp2gOk903eMcaavlTtKqcmwvjSLd0+dCxis959to7t/0G8+31tJdiLGQGXT6Fb5ejtY3Uayzcq89NDjX8QFfQitXNPDs0BLZ/pKqcmwvjSTPoeTXVXn/N6//WQzIvC+BcGDPjCui7kHq9tZkZ86qkWpERr0Q5/pF6S5/sKVZidN1HCUUmrIJQvSiY4StgdI8ewob2ZFXgop8dEjPk9RZgIWYcx5/d6BQY7Xd4wqtQMzIOiX5iTxzOffx/UX5E7giJRSyiU+xspF89P85vU7egfYd7aNdUFSOwC26CgK0uOpGGPQL6uz43Ca6R/0l85JpjRndO0U1halE2XRnjtKqcmxvjSLo3V2mjr6zjv+7qlzDDpN0Hy+R0nW2Ct4PBdxl4+icgciMOhHWYToEBdmKaXUVPC0ZPBtm7y9vJlYq4UL56eF9Dwl2Ymcau7CMYZduQ5Wt5OZGMPcUVQ7QgQGfaWUinTL5qaQGh/NthPnB/0d5c2sLUofcX9vb8XZifQPOjnb2jPqMRysbmN5XsqoOwtr0FdKqVGKsgjrSjLZXt6Ea+NAaLT3cqKhM6R8vsdQBc8oUzxdfQ7KGztHndqBEIO+iFwvIsdFpFxEvhngnNtE5KiIHBGRp3zuSxaRGhH56ahHqJRSEWh9SSYN9j5OugP2DneqJ9R8PrwX9E82jtzWwdfROjtOAytHeREXQtgjV0SigEeBa4BqYJeIbDbGHPU6pxT4FrDOGNMqItk+T/ND4M1Rj04ppSLU5V6tlhfmJLH9ZAup8dEsnZMc8nMk26LJTood9Uz/YLX7Iu4oGq15hDLTXwuUG2MqjTH9wCbgZp9zPgc8aoxpBTDGDG0vIyIXATnAy6MenVJKRaj8tHgWZCbw95OuFM+O8mbW+bRSDkVJduKoyzYPVreRm2wLaYdBX6EE/TzgrNftavcxbwuBhSKyQ0TeFpHrAUTEAvw78I1Rj0wppSLc+tJM3qk8R1ldB/X23lHl8z1KsxOpaOoaujYQikPV7SwfQ2oHQgv6/v5s+Y7OCpQCG4CNwK9EJBX4IvCiMeYsIxCRu0Vkt4jsbmoKvkGBUkpFgstLs+gZGOQnr5103R5D0C/JTqSzz0F9iNsw2nsHqGzuGlM+H0LI6eOa2Rd43c4Hav2c87YxZgA4JSLHcf0ReB+wXkS+CCQCMSLSaYw572KwMeYx4DGANWvWhGfjSKWUmmCXLkjHahFeOlxPQXoc88bQ+LHYq4JnTkrgVsweh6vHtijLI5SZ/i6gVESKRCQGuAPY7HPOC8BVACKSiSvdU2mM+bgxZp4xphB4AHjCN+ArpdR0lWSLZvU8V/AdyywfRl+2ebBm7BdxIYSgb4xxAPcCW4Ay4BljzBER+YGI3OQ+bQvQIiJHgdeBbxhjWopfGZQAAAonSURBVMY0IqWUmkY8G6uMJZ8Prv3Ak23WkIP+oep2CtLjSB/jplGhpHcwxrwIvOhz7DteXxvga+5/gZ7jceDxsQxSKaUi1YcvyudUcxcbFvlWqodGRCgZxS5aB6rbWDnG1A7oilyllBqXvNQ4Hr59FYmxIc2h/SrJTqQihL7657r6qW7tGXPlDmjQV0qpKVeSnUhzZz9t3f0jnjeW7RF9adBXSqkpFurF3INn2wC4YIwXcUGDvlJKTbmSLNfOf0GDfk07CzITSLaNvCvXSDToK6XUFMtLiyPWagka9A9Vt48rtQMa9JVSaspFWYQFWYlDHTv9OVLbTr29l9XzQtugJRAN+kopFQGClW3+7PUKkmKt3LLat/XZ6GjQV0qpCFCSlUhNWw/d/Y5h95U3dvDi4To+dVkhKXFjz+eDBn2llIoIpTmuCp7Kpq5h9/3s9Qps1ig+fXnRuF9Hg75SSkWAQGWbZ1q6+dOBWj5+ybwxt17wpkFfKaUiQGFGAlEWGRb0f/5mBVEW4XNXLAjL62jQV0qpCBBjtTA/Pf68oF/X3sOze85y+5oCcsawS5Y/GvSVUipCFGcnUu7Vg+cXb1ZiDHz+yvDM8kGDvlJKRYyS7ESqmrsYGHTS1NHHpl1nuHV1Hvlpo9+cJZCxt4VTSikVViVZiTichtMt3Ty7p5p+h5MvbCgO62voTF8ppSKEp4Jnz+lzPLmzig+smMuCrMSwvoYGfaWUihCe/XIf2nKcrv5BvnRVSdhfQ4O+UkpFiMRYK3NSbDR39nPt0hwW5SaF/TVCCvoicr2IHBeRchHxu7G5iNwmIkdF5IiIPOU+tkpEdrqPHRSR28M5eKWUmmk8KZ57rw7/LB9CuJArIlHAo8A1QDWwS0Q2G2OOep1TCnwLWGeMaRURz2aR3cA/GGNOishcYI+IbDHGtIX9O1FKqRng45fMY3VBKivGsQ/uSEKp3lkLlBtjKgFEZBNwM3DU65zPAY8aY1oBjDGN7v+e8JxgjKkVkUYgC9Cgr5RSflx/wRyuv2DOhD1/KOmdPOCs1+1q9zFvC4GFIrJDRN4Wket9n0RE1gIxQIWf++4Wkd0isrupqSn00SullBqVUIK++DlmfG5bgVJgA7AR+JWIDH02EZE5wJPAXcYY57AnM+YxY8waY8yarKysUMeulFJqlEIJ+tVAgdftfKDWzzl/MsYMGGNOAcdx/RFARJKBvwIPGmPeHv+QlVJKjVUoQX8XUCoiRSISA9wBbPY55wXgKgARycSV7ql0n/888IQx5g/hG7ZSSqmxCBr0jTEO4F5gC1AGPGOMOSIiPxCRm9ynbQFaROQo8DrwDWNMC3AbcAVwp4jsd/9bNSHfiVJKqaDEGN/0/NRas2aN2b1791QPQymlphUR2WOMWRPsPF2Rq5RSs4gGfaWUmkUiLr0jIk3A6akexyTKBJqnehBTTN8DF30f9D3wGMv7MN8YE7TmPeKC/mwjIrtDycPNZPoeuOj7oO+Bx0S+D5reUUqpWUSDvlJKzSIa9KfeY1M9gAig74GLvg/6HnhM2PugOX2llJpFdKavlFKziAb9SSIivxaRRhE57HUsXUS2ishJ93/TpnKMk0FECkTkdREpc++o9mX38VnzXoiITUTeFZED7vfg++7jRSLyjvs9+L27d9WMJyJRIrJPRP7ivj3r3gcRqRKRQ+5WNbvdxybkd0KD/uR5HPDdZ+CbwKvGmFLgVfftmc4BfN0YswS4FPiSiCxldr0XfcDVxpiVwCrgehG5FPgR8LD7PWgFPjOFY5xMX8bV18tjtr4PVxljVnmVak7I74QG/UlijNkGnPM5fDPwW/fXvwVumdRBTQFjTJ0xZq/76w5cv+x5zKL3wrh0um9Gu/8Z4GrgWffxGf0eeIhIPnAj8Cv3bWEWvg8BTMjvhAb9qZVjjKkDVzAEsoOcP6OISCGwGniHWfZeuFMa+4FGYCuuHeXa3F1twf8OdTPRI8A/Ap7NlTKYne+DAV4WkT0icrf72IT8ToSyR65SYSciicAfga8YY+yuCd7sYYwZBFa5d5h7Hlji77TJHdXkEpEPAI3GmD0issFz2M+pM/p9cFvn3kc8G9gqIscm6oV0pj+1GtxbSXq2lGyc4vFMChGJxhXwf2eMec59eFa+F8aYNuANXNc3UkXk/2/vbkLjqsIwjv8fE6pJtIs2QRQtVaSLIkHwq5IoCqUiiFiwglRo3bmQEkpBFNEguHAjItVC8QsX/RItdiNUMLVSWr9SjRWhC60uDOpCRIq48XVx3sFhTMLEdjJ0zvODw9x7z7333Dlk3rk5l3lP40Zsrhnqes0YcJ+kM8A+yrDOi9TXD0TET/n6C+Um4BY69Jlw0O+uQ8CWXN4CvNfFa1kSOWb7GvBtRLzQVFVNX0gaacwhLWkAWE95tjEFPJC79XQfAETEExFxVUSspszI92FEbKayfpA0JOmyxjKwAThFhz4T/nHWEpG0lzJx/DDwM/AMZZrJA8Aq4EdgU0S0PuztKZLGgY+Br/l3HPdJyrh+FX0haZTyYK6PcuN1ICKelXQt5Y53BXASeDgi/urelS6dHN7ZERH31tYP+X4P5mo/sCcinpO0kg58Jhz0zcwq4uEdM7OKOOibmVXEQd/MrCIO+mZmFXHQNzOriIO+mVlFHPStSpK2StrZgfMekbTghNaSJiQNnu+2zdrhoG89T8U5/a03pQU4HyYAB33rCgd96wmStks6lWVC0uqcqOUVYBq4WtIjkk5L+oiS96Vx7IikdyR9lmUst09K2i3pMPDWPO0OSNonaUbSfmCgqW6XpM9bJkrZBlwJTEmaym0bJB2XNC3p7UxGZ9YZEeHickEX4EZKWoch4FLgG0rK5r+BdbnPFZSfso8Ay4BjwM6s2wOM5/IqSl4ggEngC2Bggba3A6/n8ihlkpibcn1FvvZRkqqN5voZYDiXh4GjwFCuPw483e0+dend4tTK1gvGgYMRcRZA0rvA7cAPEXEi97kVOBIRv+Y++4E1WbceWNuU3nl5IwEWcCgi/lyg7TuAlwAiYkbSTFPdg5kbvZ/ypbMWmGk5fl1uP5btLwOOt/vGzRbLQd96wXzJ+M+2rM+XaOoi4LbW4J5BuPUcc/nPeSVdA+wAbo6I3yS9CVwyx7ECPoiIh9pox+yceUzfesFR4H5Jg5madiMlk2ezT4A7Ja3MfP6bmuoOA481ViTdsMi2N+dx11OGeACWU74wfpd0OXBP0zF/AI3/JE4AY5Kuy3MMSlqDWYf4Tt8ueBExnXfSn+amVykTajfvMytpkjJ0Mkt5uNuX1duAl3Nopp8SyB9ts/ldwBt57JeNa4iIrySdpDxf+I7yDKFhN/C+pNmIuEvSVmCvpIuz/ingdJvtmy2KUyubmVXEwztmZhXx8I5ZGyTdDTzfsvn7iNjYjesx+788vGNmVhEP75iZVcRB38ysIg76ZmYVcdA3M6uIg76ZWUX+AYmgfiKMDeT5AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.groupby(df['order_date'].dt.week)[['is_paid_order']].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "5dNkBM8ALs-K"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fe9cf3bc400>"
      ]
     },
     "execution_count": 37,
     "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": [
    "df.groupby(df['order_date'].dt.month)[['is_paid_order']].mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PolyCollection at 0x7fe9cee01ac8>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjIAAAGsCAYAAADHSE33AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAMTQAADE0B0s6tTgAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3Xd81PX9B/DXjey912WQnZAAYYS9K1AUHAiooKBUKdpSi1tsrVaUDimODn42gi0qKiiVOrCKYFiSEIIZZJF12Xvnkhuf3x9fchJI4BLuc9/vJe/n43EPyOXuPu+D5O59n/F+yxhjDIQQQgghVkgudgCEEEIIIcNFiQwhhBBCrBYlMoQQQgixWpTIEEIIIcRqUSJDCCGEEKtFiQwhhBBCrBYlMoQQQgixWpTIEGJB69evx9q1a0WNobS0FDKZDEVFRVwev6enB6tWrYKrqytkMhmXMaSG978pIWRwlMgQYgbz5s2DTCaDTCaDg4MDIiIisH79epw/f77f7V577TX89a9/FSlKQXBwMKqrqzFmzBguj//BBx/g6NGjOHnyJKqrq7mMITW8/01vxKxZs/C73/1O7DAI4YYSGULM5NFHH0V1dTXy8/ORkpICrVaLKVOm4NChQ8bbuLm5wc3NTcQoAYVCAX9/fygUCi6PX1xcjLi4OCQkJMDf35/LGJbW09Nzze/z/jcdyPViImS0oESGEDNxcnKCv78/QkJCMG/ePLz77ru47777sGnTJmi1WgBXLy2FhYXh1VdfxcqVK+Hk5IS4uDgcOXLE+P28vDwsXboU3t7ecHd3x9KlS1FSUmL8/tGjRyGTyfDVV18hPj4ejo6OWLVqFTQaDd58800EBgbC19cXf/zjH433GWgZ5H//+x+mTp0Ke3t7+Pn54eGHHx70eXZ2duJnP/sZPDw84OzsjBUrVqC2ttb4/F544QV89913kMlkmDdv3oCPodPp8Nvf/hYhISGwt7dHXFxcv4TvX//6FyIjI2FnZ4fExER88cUXVz3nb775BvHx8XBxccFtt92G5uZmAMBf/vIXxMfH9xtPq9XC09MTBw4cAAB0dXXhkUcegY+PD9zd3XHLLbegtLTUePv169djzZo1ePrpp+Ht7Y0777wTjDE888wzCAoKgr29PcLDw7Fr165B/01v5DkM5He/+x1mzZqFv/zlLwgKCsLkyZMBAK+88gri4uLg6OiIqKgovP766/2ex4kTJ/DCCy9AJpMhLCzM+L19+/YhPj4eDg4OSEhIwP79+43fa2xsxMqVK+Hp6QknJyeMHz8ep06dGjQ2QkTFCCE3bO7cuWzr1q1XXZ+ZmckAsNOnTzPGGFu3bh1bs2aN8fuhoaHM19eXvfPOO6ygoIA9+OCDLCAggPX09DDGGEtLS2MpKSnswoULLCsri916660sOTnZeP9vv/2WAWDz5s1jaWlpLDU1lXl5ebGbbrqJPfDAA+zChQts9+7dDAA7f/48Y4yxkpISBoAVFhYyxhjLyclhNjY27Nlnn2W5ubns7Nmz7PXXXx/0uT744IMsMjKSHTt2jJ09e5ZNnTqV3XTTTYwxxlpaWtijjz7Kpk+fzqqrq1ljY+OAj/Hss88yf39/duDAAVZUVMQ+//xz9sUXXzDGGDtx4gRTKBTstddeY3l5eew3v/kNs7W1ZSUlJVc95++//56lpaWx8PBwtmXLFsYYYxUVFUwmk7EffvjBON5nn33GnJ2dWVdXF2OMsXvvvZfddNNNLC0tjeXl5bH777+fJSQkMJ1OZ/x/cnJyYps3b2Z5eXmsoKCAffDBBywkJIQdP36clZaWsiNHjrBPPvlkwH/TG30OA3n++eeZk5MTW716NcvOzma5ubmMMcZeffVVlpqayoqLi9kHH3zAnJyc2GeffWb8/0hOTmaPPfYYq66uZnV1dYwxxr755hvm7e3NPvzwQ3bx4kX27rvvMgcHB3bq1CnGGGObNm1iixcvZllZWayoqIgdOHCApaenDxobIWKiRIYQMxgskdFoNAwA27dvH2Ns4ERm06ZNxq+rqqoYAJaVlTXgONXV1QwAKysrY4z9+Ib4/fffG2+zceNG5unpyTQajfG6mJgYY3Jy5Zvufffdx26++WaTnmdbWxtTKpXGN0rGGLtw4QIDwLKzsxljjG3dupXNnTt30Mfo6upidnZ27KOPPhrw+6tXr2YrV67sd93UqVPZ448/Puhzfvnll9mkSZOMX8+ePbvf/8fl/+4lJSXM1taWNTU1Gb/f29vLHB0dWWpqqvH24eHhTK/XG2/z5z//mS1cuJAZDIarYr7y39Qcz+FKzz//PHN2dmbt7e2D3oYx4f///vvvN349c+ZM9vzzz/e7zfz589kbb7zR77oHH3yQbdiwgTHG2C233MJefPHFa45DiFTQ0hIhHLFLzeWvdXonMTHR+Pe+PSV1dXUAgNbWVjzyyCOIioqCq6srIiMjAQBqtXrQx/Dz8zMuaVx+XX19/YDjZ2dnD7oEdKXi4mLodDpMmzbNeF1sbCzc3d2Rn59v0mMUFRWhp6dn0DHz8/P7PT4ATJ8+/arHv/Lfre/fDADuuusufPDBBwCA3t5e/Oc//8Hq1asBADk5OdBqtQgODoazszOcnZ3h4eGB7u5uFBcXGx9j/PjxkMt/fIlcsWIFcnNzERcXh1//+tc4duzYoM/RHM9hIFFRUXB2du533WeffYZZs2bBz88Pzs7OePvtt6/6+bhSVlYWnnjiCePzd3Z2xp49e4zP/8EHH8TLL7+M2bNn48UXXzT5/5YQMSjFDoCQkSwvLw8A+u1NuJKNjY3x730Jj8FgAAA89thjOH36NHbu3IkxY8ZAp9Nh/Pjxxj03gz3G5V/3Xdf3mFfqS7ZMMZTbDvcxTB3jyud8+fO78847sXnzZmRkZKCyshIAsHjxYgBAR0cHHBwckJmZedVj+vr6Gv/u6OjY73thYWEoLCzEF198gcOHD2PZsmVYt24d3njjDS7PYSBXxlRcXIw77rgDTz31FHbu3Ak3Nzf84Q9/uO4x8I6ODvz5z382/pv0cXBwAAAsX74cxcXFOHToED7//HNs27YN//rXv4zJICFSQjMyhHD0xhtvIDg4GBMnThzW/U+fPo2f/exnuPnmmxEfH4/W1lYzRyjMChw9etSk20ZERECpVOL06dPG6/Ly8tDS0oLY2FiTHiMqKgp2dnaDjhkbG9vv8QHg1KlTJj8+ICQk8+fPxwcffIAPP/wQt99+O2xtbQEIMy1dXV3o7u5GZGRkv4urq+s1H9fJyQl33nkn3nrrLfzzn/9ESkoKt+dgioyMDDg4OODFF1/E5MmTERUV1W8zOCAkS3q9vt9148ePR3Fx8VXPPygoyHibgIAAPPTQQzh48CA2bNiAd955x6yxE2IuNCNDiJl0dnaipqYGvb29KC4uxltvvYX9+/fjk08+gVI5vF+1iIgI7N+/H4sWLUJTUxOeeOIJM0cNPPXUU5gwYQK2bt2Ke++9Fz09PUhNTcUvfvGLq27r4uKCBx54AI8++ihcXFzg5OSEhx9+GDfddNNVJ4UG4+DggMceewybN2+GXC5HUlISCgsLYTAYsGTJEmzevBlz5szBm2++iUWLFmHv3r04d+4c9u3bN6TntXr1avz+979HS0sLPvroI+P1sbGxuOOOO3DXXXdhx44diI6OhlqtxkcffYTf/e538PLyGvDx3nnnHTDGMHXqVCgUChw8eBAxMTED3tZcz+F6IiIi0NbWhj179mDWrFnYt28f0tLS+iXOoaGhOH36NCorK+Ho6AgPDw88++yzWLVqFVQqFW6++WZ0d3cjNTUVPj4+WL16NZ5//nlMmTIF8fHxaGpqwokTJ7BgwQKzxk6IudCMDCFmsnPnTgQEBCA6OhoPPPAAbGxskJaWhqVLlw77MV999VUwxjBp0iQ89NBDePHFF80YsSA+Ph6HDh3Cl19+ifHjx2Px4sXX3BPx6quvYvbs2Vi2bBnmzJmDoKAg/Pvf/x7SmC+88ALuv/9+/OIXvzDuOembNZgxYwbefvtt7Ny5EwkJCfjkk09w8ODBay7PDWTFihWorq6Gra3tVW/C7777LpYsWYIHHngAsbGxWL9+PbRa7VVLN5dzc3PD3/72NyQnJyM5ORlNTU2DJibmeg7Xk5SUhG3btuHJJ5/ExIkTUVpaio0bN/a7zeOPP47GxkaEh4cjKSkJgLB09P777+Pf//43EhMT8ZOf/AT//e9/ERoaCgBQKpV4/PHHER8fj5tvvhnJycl46aWXzBo7IeYiY+ZY9CaEEEIIEQHNyBBCCCHEalEiQwghhBCrRYkMIYQQQqwWJTKEEEIIsVqUyBBCCCHEao3oOjJ2dnbw8fEROwxCCCGEDEF9fT16enpMuu2ITmR8fHxQUVEhdhiEEEIIGQKVSmXybWlpiRBCCCFWi3siU1hYiBkzZiA6OhrJycnIzc296jbbt2/HhAkTjBdXV1ds2bIFgNB87YknnsDYsWMxbtw4zJ8//7oN0QghhBAyOnBPZDZu3IiHHnoIBQUFePLJJ7Fhw4arbvP0008jMzMTmZmZOHPmDGxtbbFmzRoAwKefforvvvsOmZmZ+OGHH7Bw4UI8++yzvMMmhBBCiBXgmsjU1dUhIyMDa9euBSD0PikpKUFpaemg9zl48CBUKhUmTZpkvK6npwcajQaMMbS1tQ1p7YwQQgghIxfXzb5qtRqBgYHGzr8ymQwhISEoLy8ftHlaSkpKv1mbZcuW4ejRo/D394eLiwuCgoJw7NgxnmETQgghxEpwX1qSyWT9vr5Wj0q1Wo3jx48bl5UAICMjA3l5eaisrERVVRUWLlyIX/ziFwPef8eOHVCpVMZLR0eHeZ4EIYQQQiSJayITHByMiooK6HQ6AEISo1arERISMuDtd+/ejeXLl8PT09N43Z49ezB//ny4u7tDLpdj3bp1+Pbbbwe8/5YtW1BRUWG8ODs7m/9JEUIIIUQyuCYyvr6+SEpKwt69ewEABw4cQFhY2IDLSowx7Nmz56rNwOHh4fjmm2+g1WoBAIcOHUJCQgLPsAkhhBBiJbgvLe3atQu7du1CdHQ0tm/fjpSUFADA0qVLkZ6ebrzdkSNHwBjDwoUL+93/kUceQUhICBITEzFu3Dh8++23+Otf/8o7bEIIIYRYARm71qYVK6dSqaiyLyGEEGJlhvL+TZV9CSGEEGK1KJEhhBBCiNWiRIYQQgghVosSGUIIIYRYLUpkCCHcGQxiR0AIGam4tigghBDGgNxcQKEAPDyEi52d2FERQkYKSmQIIVw1NgI9PcLfu7qAykrA0ZGSGkKIeVAiQwjhxmAAqqquvr6ri5IaQoh5UCJDCOGmvh641F1kUAMlNZ6egK2tZWIkhFg3SmQIIVzo9UBNzdDuc2VS4+kpJDaU1BBCBkOJDCGEi9pa4FLj+2HpS2oqKgAnpx+XnyipIYRcjhIZQojZ6XRCImMunZ3ChZIaQsiVKJEhhJhddTW/2jGU1BBCLkeJDCHErHp7hU2+lkBJDSGEEhlCiFlVVQlF8CztyqTG0xNwd6ekhpCRjhIZQojZdHcLBfDE1pfUqNWAs/OPMzU2NmJHRggxN0pkCCFmM1DxO7F1dAgXa0lqDAbhotf3/9NgEGaXHB3FjpAQaaFEhhBiFp2dQEuL2FFcm7mTmsuTjIESj+Fcdy22tkBCAiCTDS9eQkYiSmQIIWZRWSl2BENzZVLj5ibs7RlK4mFpfRupfX0tPzYhUkWJDCHkhrW1Ae3tYkcxfH1JjTWorga8vQG5XOxICJEG+lUghNwwa5uNsWbmLjZIiLWjRIYQckOam4VWAsRybrT9AyEjCSUyhJBhY0yaJ5VGOr2eZmUI6UOJDCFk2BobAY1G7ChGp7o6QKsVOwpCxEeJDCFkWAwGmo0Rk8EgbPwlZLSjRIYQMiz19TQjILaGBqCnR+woCBEXJTKEkCHT64GaGrGjIIzRrAwhlMgQQoaMTs1IB+1TIqMdJTKEkCGhOibSQ3V8yGhGiQwhZEiqq8Upz08G19Ii9LoiZDSiRIYQYrK+Xj9EeugEGRmtKJEhhJisqkrYYEqkx9r7XREyXNwTmcLCQsyYMQPR0dFITk5Gbm7uVbfZvn07JkyYYLy4urpiy5Ytxu9nZWVh3rx5iIuLQ0xMDD7++GPeYRNCrtDdLWwsJdJFe2XIaCRjjO/nqwULFuC+++7D+vXrsX//frz66qs4derUoLfv7e1FYGAgDh8+jEmTJqGrqwuJiYl45513MGvWLOh0OjQ3N8PHx+e6Y6tUKlRUVJjz6RAyal28KOzFINIWEQG4u4sdBSE3Zijv31xnZOrq6pCRkYG1a9cCAFasWIGSkhKUlpYOep+DBw9CpVJh0qRJAID33nsP06dPx6xZswAASqXSpCSGEGI+nZ2UxFgL2itDRhuuiYxarUZgYCCUSiUAQCaTISQkBOXl5YPeJyUlBRs2bDB+nZubC3t7e9xyyy2YMGEC7rvvPtTTbkNCLIqWLKxHdzfQ1CR2FIRYDvc9MjKZrN/X11rJUqvVOH78ONasWWO8TqvV4vDhw9i1axfOnTuH4OBgPPLIIwPef8eOHVCpVMZLR0eHeZ4EIaMYbSK1PrQpm4wmXBOZ4OBgVFRUQHepBChjDGq1GiEhIQPefvfu3Vi+fDk8PT2N14WGhmL+/PkICgqCTCbDmjVrcObMmQHvv2XLFlRUVBgvzs7O5n9ShIwyNBtjfXp6hD5MhIwGXBMZX19fJCUlYe/evQCAAwcOICwsDGFhYVfdljGGPXv29FtWAoBVq1YhLS0NbW1tAIAvv/wS48eP5xk2IeSS5magq0vsKMhwUOFCMlpwX1ratWsXdu3ahejoaGzfvh0pKSkAgKVLlyI9Pd14uyNHjoAxhoULF/a7f0hICJ555hlMnz4d48ePx9dff42//vWvvMMmZNRjjDaOWjOtlooXktGB+/FrMdHxa0KGr6EBKCsTOwpyI5RKICEBUCjEjoSQoZHM8WtCiHUyGISlCWLdqMEnGQ0okSGEXKW+XuirRKxfXZ2Q0BAyUlEiQwjpR68HamrEjoKYC/1/kpGOEhlCSD+1tfQJfqSprxc2/xIyElEiQwgxoj0VI5PBQCfQyMhFiQwhxIhqj4xcjY1CoTxCRhpKZAghAITNvVR3ZOSiukBkpKJEhhACgPrzjAZNTUJTSUJGEkpkCCHQaKhj8mhBvbPISEOJDCEElZU0GzNatLYCnZ1iR0GI+VAiQ8go19kJtLTwe3ytXo8T6irUdVL3SamgWRkykijFDoAQIi7eb2r7LxRh/4UiAECUpztmqAIwTeUPXydHvgOTQbW3A21tgKur2JEQcuMokSFkFGtrE97UeGnR9OC/BSUIcHZClKcb0qrq8M4PF/DODxeEpCY4ANNVAfBxdOAXBBlQVRUlMmRkoESGkFGM92zMgQtF0Oj1uHdcLKYG+aNXr0dmTT1OVlQjraoW75xvwTvnLyD6sqTGm5Iai+hbUnR3FzsSQm4MJTKEjFLNzUAXx20rtZ1d+OpiGaI83ZEc6AcAsFUokBzkj+Qgf/T0JTXqaqRX1aKgqQV7zl9AjJc7ZqgCMV3lDy9KariqrKREhlg/SmQIGYUsURztw5wC6BjDmsQYyGSyq75vp1BgapA/pl6W1JxQV+NsVS3yG3Ox+3wuYrw8jDM1Xg72fAMehTQaoeKvl5fYkRAyfJTIEDIKNTYKb2K8lLe241hZJcb5eiPR1/u6t++X1Oj0OFdTh5MV1UivqkN+YzN2Z+Yi9lJSM42SGrOqrgY8PYEBck1CrAIlMoSMMowJb148vZedDwZgTWLMkO9rp1RgmkpIWHp0emTU1OGkuhpnq+uQ15fUeHtcOv0UAE9Kam5ITw/Q0AD4+IgdCSHDQ4kMIaNMXZ3QV4mX/MZmpFXVYlqQPyI9b2wDhp1SgekqYWlJo9PhXN/yU3UtLjQ04+3MXMR6ewrLT0H+8KCkZliqq4XlJTlVFiNWiBIZQkYRvR6oqeH3+IwxvJuVBzmAuxOGPhtzLfZKZb+kJqO6HicrqnC2ug4XGprw9rkcxF1Kaqap/OFhT0mNqbRaIcH19xc7EkKGjhIZQkaR2lpAp+P3+Jm1Dcipb8KCMBVUrs7cxrFXKjEjOAAzggPQrdMho1pYfsqorkNuQxNSzuUg3scTM1QBmEpJjUlqa4XlJYVC7EgIGRpKZAgZJXQ64c2KFwNjeC8rD0q5HKvGRvMb6AoOSiVmBgdiZnAgunU6nK0SNgqfq65DTr2Q1MT5eGJGcCCmBfnD3d7OYrFZk76fj8BAsSMhZGgokSFklKiuBgwGfo9/qqIaxS1tuCVqjGiVeh2USswKCcSskEB0a3VIr67FqYpqZFTXC0lNRjbifbyE5acgf7iJmNToDAZodHpodLpLFz26L/2pGeRPQFiyc7a14RJTbS3g6wso6Z2BWBH6cSVkFOjtBerr+T2+3mDAvuwC2CsVuCMugt9AQ+Bgo8TskCDMDgkyJjUn1dU4V1OP7PpG/DMjG2N9vYzLT252Ayc1jDFo+yUdwp/dVyQhA/3ZfdnXVyYpumFmlUq5HPdPiL+Rf5pBGQxCwhsczOXhCeGCEhlCRoGqKuHYNS/fllagqqMTq+OjBk0IxHR5UtOl1SK9b/mpph5ZdY14KyMb0V4eUMhlAyYdhmH+48llMjgoFbBXKmGvVMDNzhZ+Tg6Xvlb2+97lf159vfD3balncPhiGZZFj+HWyqG+HvDzA2xtuTw8IWZHiQwhI5xGAzQ18Xv8Hr0eH+QWwtXWFsuiw/kNZCaONjaYExqEOaFB6NRqkV4lzNTk1DdBIZcZEwkvB/trJBZXJxqX38bh0p9KuXzAqsbDdU9iDF5KTcNHuYXYNHmc2R73cn11hkJDuTw8IWZHiQwhI1xlJd/ZmMNFZWjq1uD+8fFwsLGulxQnGxvMDVVhbqhK7FBMMsHPB3HenjhSWoFbY8IR6MLnZFhjozArQ4e9iDWg8keEjGB9HY65Pb5WiwN5RfB2sMeiiBB+AxEAgEwmw5rEGBgYw76cQm7jWKIXFyHmQokMISNYZSXfx/80vxgdvVqsHhsNWypAYhFx3p6Y6O+DE+oqlLa0cRuHd3d0QsyFEhlCRqi2NqC9nd/jt2h68N+CEgS5OGFuaBC/gchV+qomv5edz3UcmpUh1oASGUJGKN5vQgcuFEGj1+OehBgoqEmPRYV7uGGGKkBopNnAbyd3ayvQ0cHt4QkxC3r1IWQEam4W9sfwUtfZha8uliHSww1Tg6hBjxjuSoiGXCbDu1n5YBx3c/NeniTkRnFPZAoLCzFjxgxER0cjOTkZubm5V91m+/btmDBhgvHi6uqKLVu29LuNRqNBfHw8Jk+ezDtkQqyaJTZqfpBTCB1jWJMYa9bjxcR0QS7OmB+mQm5DE87XNnAbp6NDWKYkRKq4JzIbN27EQw89hIKCAjz55JPYsGHDVbd5+umnkZmZiczMTJw5cwa2trZYs2ZNv9ts3boV06dP5x0uIVavsVGoHcNLeWs7jpVVINHXC+P8vPkNRK5rZXwUlHI53sumWRkyenFNZOrq6pCRkYG1a9cCAFasWIGSkhKUlpYOep+DBw9CpVJh0qRJxutSU1NRWFiIe++9l2e4hFi9vmJmPL2fnQ8GYG1iLN+ByHX5ODpgcUQILja34nRlDbdxurqE5UpCpIhrIqNWqxEYGAjlpQ5kMpkMISEhKC8vH/Q+KSkp/WZtOjs78eijj+Lvf/87z1AJGRHq6oS+SrwUNDbjTFUtpgb5I9LTnd9AxGQrYiNhr1Dg/ex86DnOyvBuc0HIcHFfWrpy/fxa059qtRrHjx/vt6z0xBNP4JFHHkFQ0PWPd+7YsQMqlcp46aDt9mQU0euBGn4fysEYw96sPMgB3JMQzW8gMiRu9na4JXoMKts78V0ZvzUg3q0uCBkurolMcHAwKioqoNPpAAgvhGq1GiEhA1cA3b17N5YvXw5PT0/jdcePH8eLL76IsLAw3HXXXcjKysLYsWMHvP+WLVtQUVFhvDg78ynfTYgU1dYCl37VuDhf24Cc+ibMC1NB5erCbyAyZMujw+FsY4MPcgqg1eu5jUOzMkSKuCYyvr6+SEpKwt69ewEABw4cQFhYGMLCwq66LWMMe/bsuWoz8A8//IDS0lKUlpZi3759SExMRE5ODs+wCbE6Op2wrMSLgTG8l50PpVyOVWNpNkZqnGxtcFtsBOq7uvG/EjW3cXp7he7YhEgJ96WlXbt2YdeuXYiOjsb27duRkpICAFi6dCnS09ONtzty5AgYY1i4cCHvkAgZcaqrhaUlXk5X1OBicyuWRITCx9GB30Bk2JZGhsHd3g4Hcoug4Tg1V10NGAzcHp6QIZMxnmf2RKZSqVBRUSF2GIRw1dsLZGfzm/LXGwx49PB3aNJo8Lel8+FmZ8dnIHLDviwqxVvncrAmMQZ3xEZyGycwEAgI4PbwhAzp/Zsq+xJi5dRqvvsWvi2tQFVHJ5ZHh1MSI3ELw0Pg6+iAg3kX0dGr5TZObS3fGUBChoISGUKsmFoNtLTwe/wevR4f5hbCxdYGy6LH8BuImIWNXI7VY6PRqdXhP/kXuY3D+4QcIUNBiQwhVqq2lu8GXwA4XFSGxm4NVsRFwtHGhu9gxCxmhwZB5eqMzwpL0aLp4TZOXR2g5TfpQ4jJKJEhxAo1NwO8t391abX4OK8I3g72WBwRyncwYjYKmQz3JMSgR6/HgQtF3MYxGPhXkSbEFJTIEGJl2tuBkhL+43xaUIL2Xi1WjY2GrULBf0BiNsmBfojwcMNXxeWo6+ziNk5DA99K0oSYghIZQqyIRgNcvMi/KFmrpgeH8osR5OKEeaHXr6pNpEUmk2FNYgx0BgM+yi3kNo4lOq0Tcj2UyBBiJbRaoLDQMqdFDuQVQaPX4+6EGCjk9DJhjcb5eiPBxwtHSytQ0cavXUtTE99u64RcD71CEWIF9HohibHENH5dZxcOXyxHhIcbpgX58x+QcCGTyXBPYgwMAPblFHAbhzGgkl+LJ0KuixIZQiSOMaC4GOjutsx4H+YWQmcwYE1i7FVNX4l1ifHywOQAX5yqqEZxcyu3cVpagC5+W3Fg7KOdAAAgAElEQVQIuSZKZAiRuLIyoK3NMmOp29pxrLQCib5eGO/nbZlBCVd3J8RABuC97Hyu49CsDBELJTKESFhlJdDYaLnx3s/OhwHAmsRYyw1KuApzd8WskECcq6lHTj2/H6a2NuFEHSGWRokMIRJVX2/Z6qkFjc34vrIWU4P8EOXpbrmBCXerx0ZDLpPh/ex88GyvRyeYiBgokSFEglpbhfYDlsIYw7tZ+ZBDWIogI0uAsxMWjgnGhYZmZNTUcxuno4NvywxCBkKJDCES09kpbO61ZF/6H+oakF3fiLlhKgS7ulhuYGIxK+OjYCOXC8uHHH+4ysroODaxLEpkCJGQnh6gqEgo/24pfbMxSrkcq+KjLDcwsSgvB3v8NDIUJS1tOFXBr7eATif8DFMfJmIplMgQIhE6nVArRqez7LinK2twsbkViyNC4OvkaNnBiUXdHhsJe6UC+7ILoOeYLYuRkJPRixIZQiTAYBBe+Hv4NSsekN5gwHvZ+bBXKLAiNtKygxOLc7WzxfLocFR1dOLbMr5dR7u6LNNOgxBKZAgRGWNCE8jOTsuP/W1ZBaraO7EsJhxu9naWD4BY3LLoMXCxtcFHOYXo5dzvoq1N2DNDCE+UyBAiMrVanJMevXo9PswphIutDZZHj+E6VnQ0EB4OeHgA1LpJXI42NrgjNhIN3Rp8VVzOfbzGRjqWTfiilxRCRFRTI9SLEcOXF8vQ2K3BHXGRcLSx4TaOuzvg4iIkMeHhwPjxlNSIbXFkKDwd7HHgQhG6LbApq7oaaGjgPgwZpehlhBCRNDWJV9a9S6vFxxeK4OVgjyURodzGkcmAoKD+18nlVyc17u6U1FiSnUKBlfFRaOvpxWcFJRYZs7xcqI9EiLnRSwcRTW3t6D2i2d4OlJaKN/6nBSVo79Vi1dgo2CoU3Mbx9ATs7Qf/fl9SExEhJDVjxlBSYykLwlTwd3LEf/KL0W6Btup9zU+puSQxN3q5IKKorAQqKoCcHKCubnSdbOjuFvc0R6umB4cKihHo4oT5oSpu48hkQGCg6beXy4XEh5Iay1DK5bgrIRpdOh0O5hVbZEyDQSgxYOnTeWRko5cIYnF1dT/2ENLrhc2ueXmj45Nab6/wQs75sMg1Hcgrgkanxz0JMVBwzBJ8fABb2+Hd9/KkZty4H5Mamcy8MY52M4MDEeLmgs+LStDcbZlyvGLVSyIjFyUyxKJaWgbuIdTVBVy4IKyji/kmz5NeL37F07rOLhy+WI4IDzdMC/LnNo5CAQQEmO+xBpqpoaTmxsllMtyTEINevQH7LxRZbFwqmEfMiRIZYjGdnUK9lGuprxeWm5qbLROTpTAmLCd1d4sbx0e5hdAZDFiTGAMZx0zA1xdQKs3/uJTUmN/kAF9Ee7rjf8XlqOmw3LRo3+vBaFpWJnxQIkMsQqMx/ROYVitsChxJa+mlpcIGXzGp29pxtLQCCT5eGOfrzW0cpRLw8+P28EZXJjVhYYCbGyU1QyWTyXBPYgz0jOGj3AKLjj3YDC0hQ0GJDOGur4ncUNfE29qA3FyhBoU1f2qrqBCOWovt/ewCGADuszH+/kKSYUkKBeDlBURGUlIzHIm+3hjn641jZZUob7Vsxl1f/+OeOUKGgxIZwtWNnlIwGISqoLm54s9oDEddnXDMXGyFTS34vrIGyYF+iPby4DaOra2wrCQmSmqGZ01iDBiAfTn5Fh+7slKoAEzIcHBYxSZEYM66ERoNUFAgvEGpVHz2X5iblKbN383KgxzA3QkxXMcJCJBWwtCX1Hh5CZutW1qE/VdtbdYxyyeXCxeFQvhTrxdOvvEQ6emOqUF++L6yFkVNLYj0dOcz0CDKygAbG8DV1aLDkhHACt4OiLXiUcmzsVF4zKAgwJvfNo8b1tFx/Y3NlnK+tgFZdY2YF6pCiJsLt3Hs7YWEQap4JzWXJxx9f5p63WDfu1J3tzA7yctdY2NwprIW72Xn47dzpvIbaAB9H3xiYgAHB4sOTawcJTKEC569VXQ64dNbYyMQEiK9Fz2NRjihJIWjpYwxvJuVB6VMhtVjo7iOFRgordmYaxkoqelLuoebjFiCg4OwwZnXnqsQNxfMCQ3CsbJKZNU1IJHjpvCB6PXCUnRs7PBrEJHRh/uvX2FhIWbMmIHo6GgkJycjd4CPE9u3b8eECROMF1dXV2zZsgUAcOTIEUydOhXx8fFISEjA1q1bwaxhTngUs1S3244OofZMZaU0kgZAOHE1nI3NvJyurMHF5lYsigiFr5Mjt3EcHYVWA9aoL6kJDxcuoaFAcLCQmPn7C3t+vLyE5+fqCjg7C8/Xzk5YCrF05WHeCePqsdFQymR4LytflNdarVb8opHEunD/Fdy4cSMeeughFBQU4Mknn8SGDRuuus3TTz+NzMxMZGZm4syZM7C1tcWaNWsAAB4eHnj//feRm5uL9PR0HDt2DO+//z7vsMkwtbUJsyWWwphw4iE3V/yGdAaDkMRI5ci43mDA+9n5sFcocGdcJNexrmwMSfixs+O7rOrn5IifhIegoKkF6dV1/Aa6hr5yDfSZlZiCayJTV1eHjIwMrF27FgCwYsUKlJSUoPQa3fIOHjwIlUqFSZMmAQCSkpIQHh4OALC3t8eECRNQXGyZviBkaLq6hDVuMV58+iqFFheLUzlXig3xjpZVorK9E7dEj4GbvR23cVxcaIOmpQUE8J0JujMuErYKOd7LzodBpGxCSvvMiLRxTWTUajUCAwOhvHTERCaTISQkBOXl5YPeJyUlZcBZGwCoqanB/v37sXTpUi7xkuHr7RUSCbGng5ubxWlEyWNj843o1evxYU4BnG1tsDwmnOtYNBtjeTY2Qi8rXjwc7LE0cgzKW9txQm2BdeJBNDcLdZgIuRbuS0tXFt661pqrWq3G8ePHjctKl2tra8OyZcvw5JNPYuLEiQPef8eOHVCpVMZLR0fHjQVPTNK3QU/MHkKXu7wRZWcn//F4bmwersMXy9DQrcGK2Eg42dhwG8fdHXBy4vbw5Bp4Fx68LTYcjkol9mUXQCfiJrTaWmnUYiLSxTWRCQ4ORkVFBXSXdj4yxqBWqxESEjLg7Xfv3o3ly5fD09Oz3/Xt7e1YsmQJli9fbtwEPJAtW7agoqLCeHF2djbfkyEDYkyYidFYpnHukHR1CckMz0aUltrYPBRdWi0+vnARng72WBwZym0cmYxmY8TEuxWEi60tbo0JR01nF46UijstUlEx8vqvEfPhmsj4+voiKSkJe/fuBQAcOHAAYWFhCAsLu+q2jDHs2bPnqmWljo4OLFmyBIsXL8ZvfvMbnuGSYSgpEdaypYxXI0pLb2w21aGCErT19mJ1fBTsOH5k9/QUascQ8fj58S0OeXPUGLja2eKj3EL0iLxubA2vNUQc3JeWdu3ahV27diE6Ohrbt29HSkoKAGDp0qVIT0833u7IkSNgjGHhwoX97v/aa6/hzJkz+OSTT4zHs7dt28Y7bGICa/qUZO5GlF1dQq0YqZ2qaO3pwacFxQh0dsL8MBW3cWQy4RgwEZdcLiwx8eJgo8SKuEg0dWtwuEjcrF3Ks79EXDI2gouyqFQqVNBOMS7q6qRTfn+o+l78/f2HV4+jt1dYspLKnqDL7c7MxX8LS/DYtImYERzAbRxfX6HWChEfY0B2Nr/WBb16PX75xVH06PX429L5cOS458oUtrZCwTyRwyCcDeX9m5pGkiFrbrbeJAa4sUaUOp20NjZfrr6rG19eLEO4uyumqfh9TJfLheO/RBpkMr7/H7YKBVaOjUJ7rxb/LRD/PHRvLxXMI/1RIkOGpKMDuEYZIKvS14iytNS0SrwGg7CcJNWp7Q9zhNMlaxJjIedY+pX3vgwydF5efPcrzQ9VIdDZCZ8WlKCth9PUzxB0d4tXs4pIDyUyxGRS6iFkTo2Nwmbg6x2hLi2V7mbDirZ2HC2twFgfT4z341f2lfdJGTI8vGdlFHI57kqIRrdOh0/yivgNNARtbSPnQxW5MZTIEJP09T+RSg8hc+trRJmfL3zau5JaLe2Nze9lF8AAYG1i7FW1m8yJd+0SMnyennwbqE5XBWCMuyu+KCpDY9cAvyQiaGoSeq2R0c3kRCYzMxPvvfceAKC5uRnV1dXcgiLS0tdDiNdmQikZqBFlba2wuVmqippa8H1lDaYE+iHai1/nRltbvtVkyY3jWddHLpPh7oQYaA0G7L8gjVkZQOi1Vl8vdhRETCYlMv/4xz+wbt06Yx2XpqamAavvkpGHMWE5SUo9hHjra0SZkyNsCpb6wbe9WXmQAbgnIYbrOLz7+5Ab5+YmdOfmZaK/D+K8PfBNiRrVHRYom22i8nKgpUXsKIhYTHpZ2rVrF06fPg3XS53hIiIiUCflj6jEbMrLhbXo0ai3V2g/IGWnKqqRVdeIuaFBCHFz4TaOvb2woZRIH89ZGZlMhnsSYqFnDB/kFPAbaBhKSizTkoRIj0mJjK2tLRyuWHxV0rGFEa+qSno9hMiP8hub8fr3mXC1s8XdnGdjAgOHV3OHWJ6zM99u5PE+nkjy98Hx8iqUtkjnU07fErg5Cl4S62JSIuPj44OCggLjJsJ///vfCKZqWCNaQ4P0ZyNGs6r2DrxyPA0ymQxbZ02BtyO/XZ6OjoAHv603hAPePbDuSYgBA7BPYrMyUq7zRPgxaVpl586duOeee5Cfn4+wsDA4Ojri0KFDvGMjImltFZaUiDS1anrwUmoaOnu1eGrmZER6unMdjxpDWp++5JPXSbtwDzdMVwXgVEU1ChqbuW4yH6qeHmFmJiaG9nSNFib9N0dGRuL06dNIT0/H559/jqysLERERPCOjYigq4sKTUmZRqfDy8fTUNvZhQcnJmByIN+iLi4ufJcpCD+8lwPvGhsNOYB3s/IhtU439Do2upiUyNx2222Qy+WIi4tDfHw8FAoFbrvtNt6xEQvr+yQz0grejRR6gwE7Tp9DUXMrVsRFYlFEKPcxaTbGetnbC7VleFG5OmNemArZ9Y34oa6R30DDRDPLo4dJiUz5AD8NFy9eNHswRDw6nZDE0NqyNDHG8H8Z2ThbXYd5oUG4e2w09zHd3QEnJ+7DEI54z8qsjI+CUi7He1l5kpuVAWiv32hxzUTmrbfewpQpU1BQUIDk5GTjJSYmBmPGjLFUjIQzqfcQIsCBvCJ8XaLGeD9v/HzyOK7Ve/vQbIz1413E0NfJEYvCQ1DU3IozVbX8BroBdPpy5LvmZt9FixYhKioKmzZtwp/+9Cfj9a6urhg3bhz34IhllJRIt4cQAb4trcD72QUIc3PF49MnwsYCOxh5NyEkluPvL7yR81oyXhEXiW9K1HgvOx+TA/2gkOA5/fJywMZGKBhIRp5rJjKhoaEIDQ3FhQsXLBUPsTC1mipiStn52nr8Pf0HeDvYY+vsKXC0seE+pkwmLEmQkcHGBvD1FapV8+Bub4ebo8Lwcd5FpJZXYl6ois9AN4AxYfNvTIxwoouMLCZ9tGtoaMAvf/lLzJkzp98SE7FuUu8hNNqVtLTiTyczYK9U4Lk5yfB0sMwUiY+PsCRBRg7ezT5vjYmAk40SH+QUQCvR0wJUMG/kMimReeCBB6BSqVBTU4Pf/OY38PX1xeLFi3nHRjhqbpZ+D6HRrK6zC9tS06A1GPDUzMkIduXXfuBycrnQU4mMLAqFkMzw4mxrg9tiI1DX2Y1vStT8BrpBWq2QzOh0YkdCzMnkU0tPPfUU7O3tsWzZMnz88cc4efIk79gIJ+3twr4YIk0dvVpsS01Ds6YHm5PHY6yP5Zoc+fkB1H1kZPL1FZaZeFkaGQZ3Ozt8lFsIjYQzBY1GWGYiI4fJvZYAwM7ODk1NTVAqlaigj/NWSaMRTihJ8KQkAdCr1+MPJ9JR0d6BdePjMDPYcptVlEohkSEjk1zOd1bGXqnEyvhItGh68PqZ89BL+EWmvZ2OZY8kJiUyMTExaGpqwtq1azFt2jRMnToVSUlJvGMjZqbVCn1I9HqxIyEDMTCGN86cR25DE26OCsOyKMuWOOC9j4KIj/f+p0URoZgZHIjvK2uwJzNXkrVl+lRVCQkNsX4yNsSftBMnTqC5uRk//elPoZD4q55KpaKZo0v0eiA/H+juFjsSMpg953NxqKAE04L8sWX6RIseY7W1BcaOpd40o0FjI1Bayu/xtXo9fp96Bjn1TbhvXCxujZFuOxsbGyA+npZTpWgo79/XfdnS6/VITEw0fj1z5kzccsstkk9iyI/6jh5SEiNd/y0owaGCEsR5e+BXUydYvBZHQAAlMaOFpyffGkE2CgWemjEZwa7O+NcPeUgtr+Q32A3Samm/4Ehw3ZcuhUIBlUqFbnoXtFplZUBbm9hRkMGcqqjGnvO5CHJxwlMzJ8PWwh8S7O2FAnhkdLBEnSAnWxs8N1soGfDmmfPIqpNuad22NtovY+1MmlCLjo7G7NmzsWrVKjg7Oxuvf/jhh7kFRsyjqkqYSibSdKGhCa99nwk3ezs8NzsZLiIUcOHdj4dIj4eHUBiuq4vfGN6ODnhu9hQ8d+QU/njiLF5aMB2hbtJspV5dLXR6v+ztjVgRkyaT29rakJiYiAsXLiAtLQ1paWlIT0/nHRu5QdQwTdoq2jqw/Xg6FDIZts6aAl8ny5ccdXQU3tTI6GOJXlqhbq54cuYk9Or1eCk1DY1d0pzZ71t+l/CpcXINQ97sa01G82bflhbhF3Pk/u9at+ZuDZ45chKN3Ro8O2sykvx9RYkjKgpwleaHZGIB+fmW6bOWWl6Jnd9nIsTVBS/Nnw4nW/6tNobD1VX4nSDiM+tmX2J9Ghqkn8Scq6nHL784ik/yLkIn0ZLmvHRrddh2PA31Xd3YNClRtCTGxYWSmNHOUh3OZ4cE4d7EWJS3teMPJ9OhlWgNiLY2fj2pCD+UyIwwlZXC5l4pJzHFza3488mzqOroxN6sPDz+v1TkNTSJHZZF6AwG/PlUBkpa2rB6bDQWjAkWLRZLvYkR6XJ2tlxH6FtjwvHTyFDk1DfhzbQfYJDoi1RVlWVmqYj5UCIzQhgMwiyM1D9N9PUQ0jOG5+dMxar4KFR3dGHrt6fw9/Qf0N7TK3aI3DDGsOtsFjJr67FwTDBWxkWKFou7O+DkJNrwREIs1elcJpPh/gljMTXID8fVVdiblWeZgYeIMeFINu2XsR6UyIwAOp1Qsbe5WexIrq29txe/Tz2D1p4e/GrqBIzz88bqsdHYsWg2En298HWJGpu/PIajpRWSrgg6XB/mFuJIaQWS/H3w0MQEyEQ8KmSpNy8ifZbc8K2QyfCrqUmI8fLAf/KL8XmhNIu49PbyLRpIzMukRCYjIwNLlixBdHQ0wsPDjRciPo0GyMuT/lRor16P7cfTUdXeifUT4jFd9WOL5SAXZzw/Zyo2J48HZMAbaefx/LHTqGiT+JMagq9LyvFhbiEiPNzw2PSJUIpYfc7LC3BwEG14IkFBQZY7gm+nUOCZmZMR6OKEtzNzcbpCmkcrW1uB2lqxoyCmMOnVdN26dbj99tuxf/9+HDp0yHgxRWFhIWbMmIHo6GgkJycjNzf3qtts374dEyZMMF5cXV2xZcsW4/dTUlIQFRWFiIgIPPTQQ9DRnB8AIXnJzwd6esSO5Nr0jOG17zOR19iMZdFjcMsAPYRkMhnmhqrwxpJ5WBQegpz6Jjz21Xd4PzsfPRLdGGiqs9V12HU2G75ODnh21hQ4iFgP3RLF0Ij1sbOzbFFEFztbPDc7GW52dsJrg0T3yFVWAp2dYkdBrsek49cTJkxAZmbmsAZYsGAB7rvvPqxfvx779+/Hq6++ilOnTg16+97eXgQGBuLw4cOYNGkSSkpKMHPmTJw7dw6+vr649dZbcfPNN2Pjxo3XHXskH79uahKmPqW+AsMYw9uZufi8qBQzVAH49bQkyE346Jff2Ix/nM1CeWs7/J0c8eDEBEzw97FAxOZV1NSC3x49DRuFHC8vmIEgF3Erbvn6AsHi7S8mEtbbC2RnW/Y15WJzK3777Sko5XJsWzADKlfpVaSztRX6MVFXHssy+/HrvkRiqOrq6pCRkYG1a9cCAFasWIGSkhKUXmPx8eDBg1CpVJg0aRIAYP/+/bj99tvh5+cHmUyGn//853j//feHHMtIUl0tbEaTehIDAIcKS/B5USnivT3xy+TxJiUxABDj5YE//WQW7hsXi2ZND36fegY7TmeguVvDOWLzqenowsvH08AYwzMzJ4uexMjlQk8lQgZiayt0x7akCA83PD5jIrp0OryUegbNGun9ftN+GekzKZE5efIkpk2bhnHjxiE5Odl4uR61Wo3AwEAoL02ly2QyhISEoLy8fND7pKSkYMOGDcavy8vLERoaavw6LCzsmvcfyRgTfqGqqsSOxDQn1FV45/wFqFych9VDSCmX49aYCLy2ZC6mBPrhhLoam788hi+KSqGXeBbX3tOLbaln0NbTi19NnYBYb0+xQ4KfH3X5JdcmRvPQJH9fbJqUiPqubmxLTUO3VnpbB1pagLo6saMggzHpZW3nzp3DHuDKkxnXWslSq9U4fvz4VTMulz/Gte6/Y8cO7Nixw/h1h9R3wA6BXg9cvAi0t4sdiWly6hvx+pnz8LC3w9bZU+B8A5U8fRwd8PTMyThTWYOUczn457kcHC2twMZJiQj3sFARjCHo0evxyol0VHV0YsOEeExTiT8NolQKiQwh19L3c2Lp1iYLxgSjoVuDD3IK8OdTGXhm1mRRN8QPpKJCKFlAZQukx6REZu7cuQCA2tpayGQy+PqaVok0ODgYFRUV0Ol0UCqVYIxBrVYjJCRkwNvv3r0by5cvh6fnj59eQ0JC+i1FlZWVDXr/LVu29NskrFKpTIpT6np7hePVEpx1HZC6rR1/OJEOpVyGrbPN10MoOcgfiX7e+DCnAP8tLMVTXx/H0qgxuGtsNBxspDHVIGxsPof8xmbcGhOOpQNsbBaDvz+t8RPT+PkB9fWWr6OyMi4SjV3d+LpEjX+czcIjk8eJWqLgSn31ZeLi6HdJakxKeS9cuIDExETExsYiJiYG48aNQ17e9YsZ+fr6IikpCXv37gUAHDhwAGFhYQgLC7vqtowx7Nmzp9+yEiDsq/nkk09QW1sLxhj+8Y9/4K677jIl7BGhs1M4Xm0tSUxTtwYvpaZBo9PjiemTMMbdvDMmDkol1o2Px59+MguRnu74b2EJNh8+htMV1aLXnmGMYXdmDr6vrMXM4ECsTYwVNZ4+Yux9INZLoRBn9k4mk+GhiQmYFOCLb0srsC+nwPJBXEdPD+2XkSKTEpmHH34YzzzzDJqbm9Hc3Ixnn30WmzZtMmmAXbt2YdeuXYiOjsb27duRkpICAFi6dGm/DtpHjhwBYwwLFy7sd//w8HC88MILmDlzJiIiIuDr63tVsjNStbQABQWAVit2JKbp0mqxLTUNDV3d2DR5HNdTRmHurti2YAY2TkxAj06HP53KwCsn0lHX2cVtzOv5tKAYXxSVYayPJ345ZZzJG5t5E2PfA7Fuvr6AjQh9HRVyObZMS0KEhxv2XyjCV8XS2w9J+2WkZ9jHr5OSkoZ1ksmSrPn4dW2tsCZrLXQGA14+nobztQ24e2w07oy3XAvZFk0P3jmfi+/Kq2CrkGNVfDSWRY+x6Br78fIq/OX7cwh2dca2+TMk093X3l44OiqRnIpYkfp6QKxzFa2aHjxz5CTqO7vw1MzJmBworQ1eMhkQGytURSZ8mP34tUKh6FfILj8/H3L6iMcFY8KLhzUlMYwx/D39B5yvbcBPxgRjhYV7CLnb2+FXU5Pw/Jyp8HJwsHgjyuy6RryRdh6e9nZ4bnayZJIYQCh+R0kMGQ5vb6FQnhjcLv0uOdvaYsfpcyhsahEnkEEwJvS2s/JanSOGSdnItm3bMHfuXCxatAiLFy/GnDlz8Morr/CObdQxGISTSfX1YkcyNPtyCnC0rBKTAnxF7SE0zs8bOxbNxmoLNqIsb23HH06mw0Yux9bZyfB2lE7tf0v20CEjj0wmbt2hQBcnPDNrMhgYXj6ehuoOaZXY7ekBysrEjoIAJi4tAUJxuzNnzoAxhunTp8Pb25t3bDfMmpaWtFqgqAjoEm+Lx7B8VVyOXWezEOHhhhfnTYO9RAqVVLV34P8yspFV1whXW1usGx+HuaFBZk2yGrs1eOabE2jR9OC52ckY5yet34moKMDVVewoiLXLzQW6u8UbP62qFn88kQ5fZ0e8vGAG3MSaJhpESAhtpudhKO/fJicy1shaEpnubiGJ6eU3ccBFelUt/nAiHT5OwguMu720XmAYY0gtr8Ke87lo7enFWB9PPDQx0Sxl0Du1Wvzm21Moa23HL5PHY16otI76u7gA0dFiR0FGgpYWYaZYTIcvluH/MrIR5emOF+ZOg51SOuef5XJhvww1YjUvs+2R6TtB5OPjA19fX+Ol72ty49rahMaP1pbEFDW1YMfpc3CytcFzs6dILokBhOOcc0KD8LqZG1FqDQb86eRZlLW2Y01CjOSSGEDoZkyIObi7i18EbnFEKO6IjUBhUwt2nM6A3mAQN6DLGAzCfhkJhTTqXHNGprq6GgEBASgbZCHw8tYBUiT1GZmGBmFjr7XNidV0dOKZIyeh0erwu3nTEONlHRsx8hubsetsFspuoBElYwyvnzmP78orsTgiBA8mibcnaDDu7kBEhNhRkJGkrU0oyikmxhjeTDuPo2WVWBQeIup+vIF4egJjpFH/ckQw24xMwKWdXgEBAQgNDUVoaCj0ej2ysrJGTNVcsVRWChvFrC2JaevpxUupaWjv6cWj05KsJokBfmxEuW5cHFqG2Yjyvex8fFdeiSmBftggwSQGEE4qEWJOrq7CcqWYZDIZfj55HMb5euOr4nJ8nCfyetcVmpqED6fE8kzuft3e3o7GxkbMnj0br7zyCh555BHesY1IfdOQNTViRzJ0PTo9Xrl0emBD0lhMDfIXO6QhU8jlWPVFZ5YAACAASURBVB4Tjp3DaET55cUyfJx3EVGe7vj11CQoJJjEeHnRWj3hQwrLlTZyOZ6YMRFj3F3xXnY+jpZKa8ZdrRZ3Y/RoZVIio9Pp4OLigs8++wzr1q3DiRMncPLkSd6xjTg6nVCpt7lZ7EiGTs8Ydn5/DgVNLbgtJhw/jQwTO6Qb0teI8qkZk+Boo8Q/z+Xg2W9OoLi5dcDbp1XVIiUjG/7Ojnhm1mRJbTbsI5PRbAzhx8lJWLYUm6ONDZ6dNQXejg74W/oPyKyRTr0K2i8jDpMSmZ6eHgDA0aNHsWDBAuGOVBBvSDQaoWdSp7RKIZiEMYa3z+XgTFUtZgUHYo1EegiZQ3KQP3YumYvl0eEobmnDU18fx+7MXHRrf+yYV9DYjB2nM+BsZ4vnZidL7vhnHx8foa8SIbxIJVH2dLDHc7OnwF6pwJ9OnR30A4gYNBrxKiKPViZlIwsWLEB8fDyOHz+OuXPnorm5GUqJ1AuxBh0dwsmkS/mg1flPfjG+vFiGBB8v/EJCPYTMRWhEGTdgI8rqjk68clzoCfbsrCkIcBb5+MYg5HJxi5eR0cHBQdjUKgXBri54euYU6A0M246nidpn7UqNjcKFWMZ168gwxlBTU4Pa2lqEh4fD1dUVDQ0NUKvVSEpKslScwyKFU0tNTUK3VGvb1NsntbwSO7/PRIirC16aP11S5fd5MDCGr0vU2PvDBXRqdbBXKNCr1+PJmZMxRWL9Xi4XECCdT8tkZOvpAXJypPOadqqiGq+eykCgixO2LZgBF4lMS8rlQFyc0O+MDJ3Zey399Kc/xYQJE+B6qUyot7e35JMYKaiuBkpKpPMLP1RZdQ1488x5eDrYY+vsKSM+iQEAuUyGReEheH3JPMwJCUKvXo8HJyZIOolRKgE/6YZHRhg7O6EPk1RMVwVg/YR4VLZ34g8n0tErkQZItF/Gcq6byMhkMkRERKCR5slMxpgwC1NVJXYkw1fW2oY/njgLW4UCz82eIqkeQpYgNKKcgL13LMGiCGnXSwoMBBTS23tMRrCAAGHGQSpuiRqDZdFjcKGhGa99nwmDRD49dncLJ5kIXyZtdHFyckJSUhJuueUWODv/WN79j3/8I7fArJVeL5Tzbm8XO5Lha+zqxrbUNPTq9XhuTjJC3UZvwx47iWcIvr7U54VYno2N8HNXWyt2JD+6b1wcmro1OKGuxp7zubh/fLwk6jw1NADOzkJpBMKHSYlMREQEIqhU6HX19grVLzWm11eTnE6tFi8dT0Njtwa/Sp6ARF8JzSGTfjw8gOBgsaMgo5W/v/AmLZGVHMhlMvxyyni0aHrwWWEpvB0csDwmXOywAAinmJycaL8ML9Q00kw6O4WZGK3WIsNxoTUYsC31DLLqGrEmMQZ3xEaKHRIZhLOz0BRSAh84yShWXS29JfSOXi2e+/Yk1G0d+PXUJMwKkcYueAcHobmklJbkpMzsm30rKytx2223YdKkSQCAzMxM7Ny5c/gRjjAtLUKhO2tOYhhj+FvaeWTVNWJxRAhuj6EZOKmytwciIymJIeLz9RU2m0uJs60NnpudDE97O7yRdh7ZddLY30n7ZfgxKZHZuHEj7rzzTuh0QpGwhIQEpKSkcA3MWtTWCjMx1r4zXeghVCXpHkJE2JsQFUWbe4k0KBTCEpPUeDs6YOvsZNjK5fjDyXSUt0pj02JDg1CSg5iXSYlMTU0N1q5da6zmq1QqR31BPMaEdU8JN9c2mTX0ECLCm0ZkJFXvJdIi1YrSYe6ueHLmJPTq9Hgp9Qwah9AclqfycustjipVJiUySqUSl2+laW5uhsHapyBugMEgzMLUS6fFx7CdqayRfA8hIiwjhYcDjo5iR0JIf1KuKp3o641HpoxHY7cGL6WeQacE1v/1eqG+zMjdnWp5JiUyK1euxM9//nO0t7djz549WLx4MTZs2MA7NsmqqwNapdPaY9gKGpvxl+/PwdnOFr+RcA8hAoSGAq6j9xQ8kTgvL6FQnhTNCQ3C2sRYlLe2408nz0IrgQ/hXV20X8acTEpkHnvsMcybNw+TJk3C559/js2bN2Pz5s28Y5OskZBJV7X37yHkL9EeQkQoeEc1KIiUSb3z+m0x4VgSEYqsukb8Ne28JArm1dcDzc1iRzEymLTRpampCXfffTfuvvtu3vEQC2jV9GBb6hl09PbiyZmTEeXpLnZIZBDe3tKdtifkcp6eQE2NcDpHamQyGR5IGoumbg1Sy6vg5eCAe8fFih0WysqE5WKpzmZZC5NmZKKiorBy5Up88cUXGMFlZ0YFjU6Hl4+noaazS/I9hEY7NzcgJETsKAgxXVCQ2BEMTiGT4dFpSYjxcsfB/Is4Wir+SQ3aL2MeJiUy5eXluPnmm7F9+3YEBwfjmWeeQUFBAe/YiJnpDQbsOH0ORc2tuCM2QvI9hEYzJydhcy8dICPWxM1NKNYoVXYKBZ6aORmeDvb4v4xsVLSJfyy7q2tknH4Vk0mJjJOTE9avX49jx47h2LFjaGhoQFxcHO/YiBkxxvDPczk4W12HuaFBuCchRuyQyCDs7IRj1lQBlFijsDDpFcm7nJudHX49NQlagwGvnjqHHp34PRbq6oTCqmR4TH6p1Ol0+Pjjj/Hoo4/i008/xaZNm3jGRczsk7yL+Kq4HON8vbFp8jgqeCdRSqVQ8E7KbwSEXIudnfAzLOVEPN7HE3eNjUZ5WzvezswROxwAQGkp1ZcZLpN+1DZv3gyVSoW33noLa9euhVqtxptvvsk7NmImR8sq8G52PkLdXPDEjImwkfIrzCgmlwszMbTxj1g7R0fpL43eHhuB8X7e+LpEjdTySrHDgV4PlJTQfpnhMOlzn5+fH86ePYsgKe/kIgM6X1uPv6X9AG8He2ydnQxHGxuxQyKDGDNG2Bvz/+3dfViUdb4/8PfNDDDggDyDOMwMoiMIyIOCiU+lbg92tDrWb12jrLWwX5vtycrcWg2tzMzjrlvryT2X69a6v9pcd+vkbnryAS1FBBVT8QEVhAkRUAR5fprfH7ewi6KAznB/7+H9ui6uLp1h5s2QzIfv/fl+P0TOoL1Z/fx5pZN0zUWS8PPkeLz8zbf46OBRRPgORKiXsg0+tbXADz8ABoOiMVSnR7+av/HGGyxiVKjwSjXe33cI7loN3piQDH8PzpAXldEI+HAXPDkZ0Y8PGKhzx3+MSUBTSyv+M/MwmlqV75e5eJH9Mr3Vo0Jm69atiIyMhJubGzQaDVxcXKDh1DqhldfV451vD6C5rQ2vpYyGcaCX0pHoJkJC5Hk1RM4oNFQuaEQVE+SPx6ItKKyqxobcPKXjAJD7ZZqalE6hHj3ukVmzZg0qKipQXV2Nq1evorq6ukdPkJ+fj5SUFFgsFiQnJyMvr+v/UXbv3o2kpCRER0cjMjISmZmZAICGhgY89dRTiI2NRUxMDGbMmIGKiooefnn909XGJrzz7QFcbmjE/KQ4xATxWFhR+fmJffYGkT0YjWKP2JgZNRSxQf7433NF2FtconQctLYCZ84ALS1KJ1GHHhUy3t7euO++++Dt7Y0BAwZ0fPTEvHnzkJaWhtOnT2PhwoVdzmgqKSnBnDlz8Mknn+D48ePIzc3t2N69bt061NTU4Pvvv8exY8cQHByMlStX9uJL7F8u1NTiFzv3obi6Bk+OjMR4o8DnhvdzXl7yVlUiZydJQESEuENPNZKEn4+Jh4+7O/4r5ygu1NQqHQn19XIxI8DVLuH1qJB58MEHsWXLll4/eFlZGQ4dOoTU1FQAwMyZM1FQUIDCwsJO91u7di1SU1M7ihedTgeff2kYqKurQ3NzM1paWlBTUwMDO6G6dLLiMn6xYy8u1tTimYRoPDQ8QulIt+TjI/YWTUfy8JB/sIu8q4PInkTfleer0+HnY+LR0NKC/8w8JES/TG2tXMwIMOdSaD16G1m7di1mzJgBb29vBAUFITAwEEFBQd1+XnFxMUJDQ6G9diiGJEkwGo0oKirqdL+8vDzU19dj6tSpiI+Px/z581FXVwdAXtFpf97g4GBUVVXhhRde6O3X6fT2FpcgfXcWmlvb8Nq40XhgqFnpSLcUGCi/kUdHy7sb+hM3N/kHOtvMqL9xdZX/3xf1nKSRwQF4dMQwFFypxsdHTigdBwBQUwOcPctt2bfSo0ImJycHBQUFOHr0KLKzs5GTk4Ps7OwePcH1B691NaupubkZGRkZ2LRpE3JyclBVVYX09HQAwPbt2yFJEkpLS3HhwgX4+Phg2bJlXT7X6tWrYTAYOj5qamp6lFHNbDYb/nriDFbvPwy9myveumcsRgs+P8nH558zhNrf1IcMkX/IOTuNRv563dyUTkKkDJ1O/iVG1NXYx0YMQ3SgH7aePY9M6wWl4wAAqqs5k+lWevS/kslkQlBQEEpLS3Hx4kUEBQXBZOp+Tk9YWBisVitarnUs2Ww2FBcXw3jdJDyTyYQHH3wQvr6+0Gq1mDVrFg4cOAAA+Oijj/DII49Ap9PBzc0Njz/+OHbt2tXl8y1YsABWq7XjQy/y0A87aGlrw38dPIo/HTsF40AvrJgyDkN8xV7eGDBAPi/ler6+8upMUJDzXm5p7xPw8FA6CZGy9Pqufw6IQCNJ+I8xCfB2d8Pa7O9RWlOndCQA8pbs67oy6JoeFTL79u1DREQEnnvuOaSlpWHo0KEdu4puJSgoCAkJCdi4cSMAYPPmzTCbzTBf1+E4e/Zs7Nq1C43XzmfeunUr4uLiAABDhgzBtm3bYLPZYLPZsGXLFsTExPTma3RKtc3NWP5dNnYUFCM+OBDv3DMWAZ5iv0N2N0NIowHCwoDISOc8GM5slht8iajzyqxo/Dx0+HlyPOpbWrB6/yE0C9AvAwCXLwPXdWYQeljILFiwAJs2bcLhw4eRm5uLTZs24aWXXurRE6xbtw7r1q2DxWLBihUrsH79egDAtGnTkJOTAwBISUnB9OnTER8fj9jYWJSXl3dcPkpPT0dVVRWio6MRExODiooKvPXWW7fztTqN8rp6/HJnJo5crMCPhhjxi/GjhT+xtzczhDw95WImLMx5+kgGD5a3WhPRPwUGyucoiSg+JBCPREbgbGUV/nj0pNJxOpSXc1r29SRbV00r14mPj0dubm63fycag8EAqwO+4xcuACUKHTVwtrIKy7/LxpWGRjwRG4mHhg8RfgCkiwtgsdzeKktzM1BcDFRW2j9XXwkMFPc3TyIRFBTIqw2iaW1rw5u7s3Ci4jIWpozCmMHiVF2hoWKfmnynevP+3aMVGU9PT2zfvr3jzxkZGfAU9UAAJ5ZdchGLd2WitqkZL9+ViIcjI4QvYiRJbuS93UtFrq7y5w8bJu62zVsRefmcSBSiXnbVuLjgpTHx8HJzxW+zj6CsVox+GUD+ZfriRaVTiKFHKzI5OTmYOXMm3N3dIUkSGhsbsXnzZowaNaovMt42Z1qR+Xt+ATbk5sHLzQ2Lxo/GcH/fvg1wm4xG+x2/39YGlJbKH2ro3h8wQF6JEnV3BpFIWluBU6fkg+BEc+hCGd75LhvD/Hzw1j1j4SrQP2qTSewRELerN+/fPSpkAHmL9KlTp2Cz2RAZGQlXwXsyAOcoZFptNvwhNw//OFOIUP0AvDEhCSF6dXTChoQ45vj9hga54e3qVfs/tr24u8t9PqKel0EkouZm4ORJMecM/fH7E/ji1DnMsIRjTtwIpeN0Eh7ufD14dr+09NVXX6G2thYxMTGIjY1FTU3NbZ30S73T0NKC9/cdxD/OFGJEgB+WT0lRTRHj7++4GUI6nbzSYTaLWSj0prGZiP6p/cA8EZv8fxIzHMP9ffE/pwuQUyLWNZ3Cwv49MbtHhczixYs7jQzw8fHB4sWLHRaKgMr6BizJ2I/skouYaAzFkonJ8FLJKWre3vJyp6P5+8tnz4i0rOriot5+HiIRiDq+Q+vigpfuSoDe1RUfZB9BRZ0418BsNvnAvB7OcnY6t3WhT5IktHH4g8MUVV3Fop37cLayCo+NGIYXk+PhKuKvKF3w8JCbc/vqh5BWKxdNw4crf9Bce2Mz++CJ7oyXl5gH5gV6euCF5DjUNDVj9f7DaBHofdBmk0cZ9IMD7W/Q4+nXWVlZHX/ev38/vERsMXcCRy6W442d+3ClvgEvJMVhVrRF+J1J7dzc5NUIJWouvR6IipIvZynVh2c09r+5UUSO4usLiDgfOCk0GNMt4Th1qRKfHjutdJxO2trkIZN14myu6hM9uor/3nvv4eGHH0Z0dDQA4MSJE/jb3/7m0GD90faCIvzu4DG4azX45cRkxAYJdM2kG+0zhJTsAZckucHYz09uBq6q6rvnHjRIrEtcRM4gOFhu/C0rUzpJZ4/HRuJkRSW+OHUW0YF+SBzU/RDlvtLaCuTny32ESq9S95Ue71qqrKzsGEuQkpLSqWdGVGrZtdRms+HTY6fw15NnEeTpgTcmJMHgrZ4VL0mSV2JEW6SrrJQP02tuduzz+PvLjcdE5Bjnzol3KGZZbR1e+eZbaCQXrLp3Avw9dEpH6sTVVb7krtZ+PYdsv1YjNRQyTa2t+DD7CPYWX8AwPx8sGjcaPjp1/Z8n8ta/1lb5e1Ve7pizZ7y95ZUolVz9I1KltjZ5lUG0/o+sH0qxct9BRAX4YemkMdAIdL4MIF/uHz5c/q/a2H37NTlGdWMTlu7Owt7iCxgzOARLJ92luiLGYBC3iAEcO4jS07NvG5uJ+isXF/kXBp1Yix4YMzgE04aacaLiMj47Lla/DCBfljt92vGr0kpjIaOQkqu1+MXOvTh5qRIzLOF4ZWwi3LXq2JnULihIvoatBu2DKI1G+zQju7mJe94FkTPSaORL2KKdxfrkyEhE+A7E306eRW5pudJxbtDYKK9mtbQoncRxWMgo4ETFZfxi516U1dTh2cQYzIkbAReV/Vrv4yOvdKhNYKB89ozvHUx4EPUHKpGzE/EXCFeNBgvuSoSHVos1B3Jxub5B6Ug3qK+XdzO1tiqdxDFYyPSxb4t+QPruLLS0tmHR+CTcH9EHJ8fZmV4v5hkPPXUngyglScwlbqL+QsRLuiF6TzyfNBLVjU34ddZhtArYelpbKxczAh19YzcsZPqIzWbD5hNn8OusXHi7ueLtyWMxSqAtez2l08mnbgrW03ZbvL2BESPkrdM9/aEYHi4XckSknL46Pbw3xhoG4f4IE46XX8amvHyl43SppkY+NE/AOuuOOMHbkfha2tqwNud7/L9jp2Aa6IUVU8Yh3Ed9J6e1z0FxphlCLi5AaKhc0HS3fdxguLNLUkRkP/7+8r9dkcyJi0K4jzf+kpeP7y9WKB2nS9XV8nZ2ZypmWMg4WG1TM97+9gB2FlqREBKId+5Jgb+n+k4pat81oNYzCbrT3SBKNTU2E/UXgwbJfW+icLvWL+Ou1WBNVi4qG8TrlwHkAZPnzyudwn5YyDhQWW0d3ti1D0fLLuHeIUb8YtxoeLiqbzmjP80Q6moQpa+vOhubifqDsDCxRoOEeg3A/x01ElcaG7EmK1fIfhkAuHRJPgHdGbCQcZAzl6/gFzv3obi6Bk+OjERaYoxwhyX1VH+bIfSvgyj9/dXd2Ezk7Np/0bL3OVF3YrwxFD8aYsTRskv464kzSse5qfJywAFnxvY5db6zCu7AD6VYkrEfdU3NeGVsIh4aHqGawY/X688zhPR6+VKTSr91RP2GiJe+n44fAdNAL3x+/DSOl19SOs5NXbwon1avZixk7GxLfgFW7jsId60GS+++C2MNg5SOdNtEbKYjIuqKVisfqSDKZgR3jQYvj02Em0aDX+0/jKqGRqUj3VRJiVzQqBULGTtptdmw/vBxbMjNQ6jXALw7eRws/urd4iLi9kYioltxd5dXZkS5ij/YS4+0UTGobGjEmgO5aBO0XwaQLzFViLnRqluCfLvVrb6lBSv35uAfZwoRHeiH5ZPHIUSv3s5YT0/5rBheUiEitRkwQKwD8yaZDJhsNuDIxQp8cfKs0nFu6fx54PJlpVP0HguZO1RZ34AluzKRc6EMk0yDsXjiGOjd1Ht2ffsR4KL8RkNE1FsDB8qbFETxTEIMwrz1+PT4aZyoELtSKCyUt2erCd+u7sD5qmos2rEX565U48cjhmF+UhxcVVwBcIYQETmLgAB5s4II3LVyv4yriwtW7z+M6sYmpSPdlM0mH5hXXa10kp5T77uuwnJLy/HGzkxcaWjE/OQ4/J9oi2p3JgGcIUREzic0VN60IIIwby88kxiNy/UN+EDwfhmbTR5lUFOjdJKeYSFzG/7nWBHe+S4bLhKweOIY3G0yKB3pjnGGEBE5I5NJ3rwggsnmMNxtGoxDpeX4n1PnlI5zS21t8pDJujqlk3SPhUwvfZ5TjHd3HEWApw7LJ49DTJAg5f4dCAvjDCEick6SJG9eEOVk8mcSYzDYawD+dOwUTgreL9PaCuTnA/X1Sie5NRYyvXRfdAgeiBqMFZPHweCt/iWM4GB5jhARkbNqPzDPzU3pJICHVotXxo6C1kXCr/YfxlWB+2UAoKVFLmYaxT0Gh4VMbw30cMWSe+MxUCfQEZK3yddXnuhMROTsXF3FOTDPONALcxOiUVHfgA+zj8AmcL8MADQ3A6dPA02C1lwsZPopvZ4zhIiof9Hp5MtMImwunWIOw0RjKHIulOGr/AKl43SrqUkuZpqblU5yIwG+ndTXdDp5mVXFm6yIiG6LXi/GqeWSJCEtMRah+gHY+P1JnL5UqXSkbjU2ypeZWlqUTtKZwwuZ/Px8pKSkwGKxIDk5GXl5eV3eb/fu3UhKSkJ0dDQiIyORmZnZo9uod9qXVzUapZMQESnDz0+MYbgerlq8PDYRLpKE1fsPo6ZJwOWO69TXy7uZWluVTvJPDi9k5s2bh7S0NJw+fRoLFy7E3Llzb7hPSUkJ5syZg08++QTHjx9Hbm4uoqKiur2NekejEafhjYhISWFhgIeH0ikAs483fho/AuV19fitCvplAKC2Vi5m2tqUTiJzaCFTVlaGQ4cOITU1FQAwc+ZMFBQUoLCwsNP91q5di9TU1I4CRafTwcfHp9vbqOckSZ4/IsoWRCIiJbm4yD8TReiX+dEQI8aFDcKBkov4x5lCpeP0SE0NcOmS0ilkDv0WFhcXIzQ0FNprbeKSJMFoNKKoqKjT/fLy8lBfX4+pU6ciPj4e8+fPR921U3hudRv1nEiHQhERiUCnE2MmkyRJeG5ULEL0nvjkyAmcuayOYUeiLB45vBa9/tj+rpbNmpubkZGRgU2bNiEnJwdVVVVIT0/v9rbrrV69GgaDoeOjRi3nKzuYSMd0ExGJxN9fjJ+Pnq6uePmuROBav0ytiNuDBOXQQiYsLAxWqxUt11qcbTYbiouLYbyuBDaZTHjwwQfh6+sLrVaLWbNm4cCBA93edr0FCxbAarV2fOh55r5Qg9OIiERkNIoxZ26I70A8FReFi7V1+K+c71XRLyMChxYyQUFBSEhIwMaNGwEAmzdvhtlshtls7nS/2bNnY9euXWi8dnTg1q1bERcX1+1tdGuijbInIhKRSP0y90eYMNYQgkxrKT47flrpOKrg8G/bunXrsG7dOlgsFqxYsQLr168HAEybNg05OTkAgJSUFEyfPh3x8fGIjY1FeXk5li1b1u1tdHOenvI/TJ4VQ0TUPQ8PeSeT0iRJwvOjR2Ko70D85cQZfHHqrNKRhCfZnHjtymAwwGq12v1xL1wASkrs/rB24+0tFzE8K4aIqHcKCoDLAsxyvNrYhCUZ+1FUfRXPJkTj/qFmpSPdICzMcbP6evP+LcBCGtlTQIB8VgyLGCKi3jOZxOiX8XJ3w5JJyRikH4D/PnwcGYX2/6XcWbCQcSKDB8v/CHk5iYjo9ojUL+Or0+HNSWMQ4OmB32YfQab1gtKRhCTAt4ruVPthdyEhSichIlI/UfplACDQ0wNvThwDb3d3/Hr/YRy6UKZ0JOGwkFE5rRawWABfX6WTEBE5j4AAeSaTCEK9BuDNSWOgc9Xi/X0HcbxckCN1BcFCRsXc3YHISHmaKxER2ZfRKP+cFYFxoBcWT0iGxsUFy7/LRr5KTv/tCyxkVEqvl4sYUf6RERE5G41GrGMshvr54PXxSWiz2fD2ngMovFKtdCQhsJBRIT8/+XLStRFWRETkIJ6e4vTLAMCIQD+8ljIaDa2tWLYnCz9c5SgeFjIqM2gQEB4uzm8IRETOLjBQrD7E+JBALLgrAVebmrF0dxbKavv3IGUWMiohSYDZLA+AJCKivmUyiXUpf8zgEMxPisPl+gYs3Z2FyvoGpSMphoWMCmg0wLBhYkxoJSLqj0TrlwGAiabBSBsVi9LaOqTvyUJ1Y5PSkRTBQkZwbm7A8OGAl5fSSYiI+jdPT8BgUDpFZ/cOMWJOXBSs1TV4a08WapublY7U51jICMzTU96Z5OGhdBIiIgLk2UI+Pkqn6GyGZQh+PGIYzl2pxvJvs9HQ0qJ0pD7FQkZQPj7ySoyrq9JJiIjoX5nNYvXLAMBjI4ZhhiUcJy9V4r29B9HU2qp0pD7DQkZAQUFARIQYsz6IiKgzjUa83aOSJOHJkVG4d4gR35dVYPX+w2hpa1M6Vp/gW6VAJEk+r0CkMwuIiOhGAwbIg3pFIkkSnk2MwUTjYGSXXMQHB46g1WZTOpbD8Ug1QbRPXB04UOkkRETUE8HBQE0NcEWgaQEukoQXkkaisbUF3xWXQKfV4LlRsZBEWj6yM67ICMDVVe6HYRFDRKQuZrO8u1QkGhcXvDQmAfHBgdheUIw/HDkBmxOvzLCQUZiHh7wzydNT6SRERNRbIp4vAwCuGg0WpoxCVIAftuQX4M95+UpHchgWMgry9pZXYkSr5omIqOdE7JcBAHetBq+PH40I34HYlJePL06dVTqSQ7CQUUhAADB0qFzNExGRugUHi9ke4OnqisUTkmH09sIfvz+JrWfPKx3J7ljIKGDwYHluh2hLkUREdPtE7JcBAC93XSF6PgAAFNNJREFUNyyZlIwQvSf++9AxZJy3Kh3JrljI9KH2nUkhIUonISIie9NqxTtfpp2vTof0SXchwNMDvz1wBPutF5SOZDcsZPqIVisPfhRpFDwREdmXXg+EhiqdomuBnh54c+IYeLu741f7D+NwaZnSkeyChUwf0OnknUl6vdJJiIjI0UJC5M0cIgr1GoA3J42BzlWLlXsP4nj5JaUj3TEWMg6m18s7k0Sby0FERI4THi7urDzjQC8snpAMjYsLln+XjfzLAp3odxtYyDiQnx9gsciXlYiIqP/QasU8X6bdUD8fvD4+CW02G97ecwCFV6qVjnTbWMg4yKBB4jZ9ERGR4+n18nuBqEYE+uG1lNFoaG3Fsj1ZKLlao3Sk28JCxs4kSd6CJ2qzFxER9Z1Bg8TtlwGA+JBALLgrAVebmrF0dxbKauuUjtRrLGTsSKORdyb5+yudhIiIRCFyvwwAjBkcgvlJcbhU34Clu7NQWd+gdKReYSFjJ25u8s4kLy+lkxARkUjaz5cR2UTTYKSNikVpbR2W7slCdWOT0pF6jIWMHQwYIBcxOp3SSYiISEReXmL3ywDAvUOMmDMyCsXVNXhrTxZqm5uVjtQjLGTukI+PvDNJ5GVDIiJSXmio+Kv2M4YPwY9HDMO5K9VY/m02GlpalI7ULRYydyA4GIiIkEcPEBERdSc8XPwjOR4bMQwzLOE4eakSK/cdRFNrq9KRbsnhb8H5+flISUmBxWJBcnIy8vLyurzf7t27kZSUhOjoaERGRiIzM7PT7eXl5QgODsajjz7q6MjdkiTAaAQMBqWTEBGRmri6it8vI0kSnhwZhXuHGHHkYgVW7z+MlrY2pWPdlMMLmXnz5iEtLQ2nT5/GwoULMXfu3BvuU1JSgjlz5uCTTz7B8ePHkZubi6ioqE73ef755zFt2jRHx+2RoCAgMFDpFEREpEbe3uL3y0iShGcTYzDRGIrskov4MPsIWm02pWN1yaGFTFlZGQ4dOoTU1FQAwMyZM1FQUIDCwsJO91u7di1SU1M7ihedTgcfH5+O2//0pz8hODgYkyZNcmTcHuOlJCIiuhODBok/f89FkvBCUhySQ4PxbVEJfnfwKGwCFjMOfUsuLi5GaGgotNcuCEqSBKPRiKKiok73y8vLQ319PaZOnYr4+HjMnz8fdXXyoTwlJSVYvXo1VqxY0e3zrV69GgaDoeOjpkadpxQSEZFzkyR5hIHo/TIaFxcsuCsBccEB2F5QjD8cOSFcMePwtQXpujP6u3oBmpubkZGRgU2bNiEnJwdVVVVIT08HADz77LNYuXIl9D0oXRcsWACr1drx0ZPPISIiUoIa+mUAwFWjwWspoxEV4Ict+QX4c16+0pE6cWgtGBYWBqvVipaWFmi1WthsNhQXF8NoNHa6n8lkQkJCAnx9fQEAs2bNwsqVKwEAmZmZHX01NTU1qK+vx3333Ydt27Y5MjoREZHDeXsDISFAaanSSW7NXavB6+NHI313Fjbl5cNDq8ELYRFKxwLg4BWZoKAgJCQkYOPGjQCAzZs3w2w2w2w2d7rf7NmzsWvXLjQ2NgIAtm7diri4OADA5cuXUVhYiMLCQqxatQoPPPAAixgiInIaoaHi98sAgKerKxZPSIbR2wuffH8Sm4+cVzoSgD64tLRu3TqsW7cOFosFK1aswPr16wEA06ZNQ05ODgAgJSUF06dPR3x8PGJjY1FeXo5ly5Y5OhoREZHiJEkd58sAgJe7G5ZMSkaI3hMrdxzD/x5XfilJsonWtWNHBoMBVqtV6RhERETdqqoCzpxROkXPlNfVY/OZE1g1KxYDPex/tH1v3r+5kZiIiEgAAwfKJ8arQaCnB955MNEhRUxvsZAhIiISxODB8iBi6jkWMkRERIJoP19Go1E6iXqwkCEiIhKIm5s8kPi6Y9joJljIEBERCcbLS16ZYTHTPRYyREREAvLxAa47do26wEKGiIhIUH5+gMmkdAqxsZAhIiISWEAAEBamdApxsZAhIiISXFCQPMqAbsRChoiISAUGDZIHTFJnLGSIiIhUYvBgIDBQ6RRiYSFDRESkIkYj4O+vdApxsJAhIiJSGZMJ8PVVOoUYWMgQERGpjCQB4eHyoMn+joUMERGRCrXPZfLyUjqJsljIEBERqZSLizyXqT9PzGYhQ0REpGIaDTBsGODhoXQSZbCQISIiUjmNBrBYAJ1O6SR9j4UMERGRE9Bq5ZUZNzelk/QtFjJEREROws1NXplxdVU6Sd9hIUNERORE3N3lYkarVTpJ32AhQ0RE5GR0Ovkyk0ajdBLHYyFDRETkhDw9gaFD5S3azszJvzwiIqL+S6+Xz5mRJKWTOA4LGSIiIifm7S2fAOysxQwLGSIiIifn4wOYzUqncAwWMkRERP2An588NdvZsJAhIiLqJwICgLAwpVPYFwsZIiKifiQoCAgNVTqF/bCQISIi6mcGDQKCg5VOYR8sZIiIiPohgwEIDFQ6xZ1jIUNERNRPGY2Av7/SKe6MwwuZ/Px8pKSkwGKxIDk5GXl5eV3eb/fu3UhKSkJ0dDQiIyORmZkJAPjzn/+MhIQExMTEIDY2Fh988IGjIxMREfUbJhPg66t0itvn8JFS8+bNQ1paGp566in85S9/wdy5czuKlHYlJSWYM2cOvv76a0RFRaGhoQENDQ0AAIPBgK+//hohISGoqqrCqFGjkJiYiHHjxjk6OhERkdOTJCA8HGhrA6qqlE7Tew5dkSkrK8OhQ4eQmpoKAJg5cyYKCgpQWFjY6X5r165FamoqoqKiAAA6nQ4+Pj4AgHHjxiEkJAQAMHDgQERGRqKgoMCRsYmIiPoVSZJP//XyUjpJ7zm0kCkuLkZoaCi012aJS5IEo9GIoqKiTvfLy8tDfX09pk6divj4eMyfPx91dXU3PF5eXh4yMzMxefLkLp9v9erVMBgMHR81NTX2/6KIiIickIuLPJdpwAClk/SOw3tkpOuGO9hsthvu09zcjIyMDGzatAk5OTmoqqpCenp6p/tYrVY89NBD+OijjxB6kw3wCxYsgNVq7fjQ6/V2+zqIiIicnUYDDBsGeHgonaTnHFrIhIWFwWq1oqWlBYBcxBQXF8NoNHa6n8lkwoMPPghfX19otVrMmjULBw4c6Li9pKQEU6dOxS9/+Us89thjjoxMRETUr2k0gMUC6HRKJ+kZhxYyQUFBSEhIwMaNGwEAmzdvhtlshvm6yVWzZ8/Grl270NjYCADYunUr4uLiAAAXLlzAlClT8Nprr2HOnDmOjEtEREQAtFp5ZcbNTekk3XP4paV169Zh3bp1sFgsWLFiBdavXw8AmDZtGnJycgAAKSkpmD59OuLj4xEbG4vy8nIsW7YMALBkyRIUFRVhzZo1iI+PR3x8PDZs2ODo2ERERP2am5u8MuPqqnSSW5NsXTWtOAmDwQCr1ap0DCIiItWqrwdOnwaudYl0CAuT5zY5Qm/ev3myLxEREd2Uh4d8mUmjUTpJ11jIEBER0S15egJDh8pbtEUjYCQiIiISjV4vnzNz3akqimMhQ0RERD3i7S2fACxSMcNChoiIiHrMxwcwm8UpZhw+NJKIiIici5+fPGRSBFyRISIiol4TpfFXkBhEREREvcdChoiIiFSLhQwRERGpFgsZIiIiUi0WMkRERKRaLGSIiIhItVjIEBERkWqxkCEiIiLVYiFDREREqsVChoiIiFSLhQwRERGpFgsZIiIiUi0WMkRERKRaLGSIiIhItVjIEBERkWpJNpvNpnQIR3F3d0dgYKDSMRRTU1MDvV6vdAxV42toH3wd7YOvo33wdbQPR76O5eXlaGxs7NF9nbqQ6e8MBgOsVqvSMVSNr6F98HW0D76O9sHX0T5EeR15aYmIiIhUi4UMERERqZYmPT09XekQ5Dhjx45VOoLq8TW0D76O9sHX0T74OtqHCK8je2SIiIhItXhpiYiIiFSLhQwRERGpFgsZJ9PQ0ICHH34YFosF8fHxuP/++1FYWKh0LNVaunQpJEnCsWPHlI6iWo2NjXjhhRcwbNgwREdHIzU1VelIqrRt2zaMGjUKCQkJiImJwccff6x0JFV48cUXYTabb/h3nJ+fj5SUFFgsFiQnJyMvL0/BlOLr6nUU5f2GhYwTSktLw6lTp5Cbm4t/+7d/Q1pamtKRVOnQoUPYv38/jEaj0lFUbdGiRXBxccHp06dx/PhxvP/++0pHUh2bzYbZs2djw4YNOHz4MLZs2YJ58+bh6tWrSkcT3qOPPorvvvsOJpOp09/PmzcPaWlpOH36NBYuXIi5c+cqlFAdbvY6ivB+w0LGyeh0OkybNg2SJAEA7rrrLpw7d07hVOrT2NiIn/3sZ1i7dm3Ha0m9V1tbiw0bNmD58uUdr+OgQYMUTqVeV65cAQBUV1fD398f7u7uCicS38SJE2EwGDr9XVlZGQ4dOtSxOjhz5kwUFBRw9foWunodRXm/YSHj5H7zm99g+vTpSsdQnSVLliA1NRXh4eFKR1G1s2fPwt/fH2+//TZGjx6NCRMmYMeOHUrHUh1JkvD555/j3//932EymTB+/Hh8/PHHcHNzUzqaKhUXFyM0NBRarRaA/PoajUYUFRUpnEzdlHq/YSHjxJYvX478/Hy88847SkdRlczMTGRnZ+P5559XOorqNTc349y5cxgxYgRycnLw4YcfYtasWSgvL1c6mqq0tLTg3XffxZdffonz589jx44dmDNnDi5fvqx0NNW6fqWVJ5HcGSXfb1jIOKlVq1bhr3/9K77++mt4enoqHUdVdu/ejZMnTyI8PBxmsxlWqxX33Xcfvv76a6WjqY7JZIKLiwsef/xxAEBcXBzCw8Nx/PhxhZOpS25uLkpKSjBu3DgAQFJSEkJDQ3HkyBGFk6lTWFgYrFYrWlpaAMhFTHFxMfvhbpPS7zcsZJzQ6tWr8emnn+Kbb76Bj4+P0nFUZ9GiRSgpKUFhYSEKCwthMBiwbds2PPDAA0pHU52AgABMmTIF27ZtAwCcP38eBQUFGD58uMLJ1KX9jffUqVMAgDNnzuDs2bOwWCwKJ1OnoKAgJCQkYOPGjQCAzZs3w2w2w2w2KxtMhUR4v+HJvk7GarUiLCwMQ4YMgZeXFwDA3d0dWVlZCidTL7PZjC1btiAmJkbpKKp07tw5/PSnP8WlS5eg0Wjw5ptv4pFHHlE6lup8+umnWL58OVxcXGCz2fD6669j1qxZSscS3s9+9jN8+eWXKC0tRUBAAPR6Pc6cOYNTp07hqaeewqVLl+Dt7Y2PP/4Y0dHRSscVVlevY0ZGhhDvNyxkiIiISLV4aYmIiIhUi4UMERERqRYLGSIiIlItFjJERESkWixkiIiISLVYyBDRHZEkCTU1NUrHuKmPPvoIv/rVr3r9eU899RQ+/PDDG/6+pKQE99xzT68e6/e//z1iY2Oh1WpveMzXX38dUVFRiIuLQ3JyMnbu3NnrrET9mVbpAERELS0tHXNv7P24zz33nF0fMzQ0FLt27erV54waNQqff/453n333RtumzBhAhYvXgwPDw8cOXIEd999Ny5cuACdTmevyEROjSsyRE5MkiS89957GDNmDMLDw7Fhw4aO28xmM44dO9bx59GjRyMjIwMAcPfdd+PVV1/FxIkTERYWhvfffx+fffYZUlJSYDKZ8Nlnn3V6nlWrVmHcuHGwWCz49NNPO/4+OzsbkydPxujRo5GYmIjNmzcDAAoLCxEQEIBly5ZhwoQJ+OCDD7rMnp6efsePm56ejldeeQUA0NraildeeQUxMTGIiYnB/Pnz0dTUBAD44YcfMGXKFIwcORIPPfQQKioqunxN258DAOrr6/HjH/8YI0aMQFxcHO69994uPycuLg5RUVFwcbnxR+4DDzwADw8PAEBsbCxaW1tv+txEdCOuyBA5OZ1Oh6ysLJw4cQLJycl44oknerT6UVRUhIyMDJSWliIiIgIvv/wy9u3bhwMHDuDhhx/udKqsJEnYu3cvzp07h+TkZIwfPx5eXl6YN28e/v73v2PQoEGoqKjAqFGjOuYFXbp0CUOHDsWSJUtumsEej5uent7xeL/73e9w8OBBHDx4EBqNBjNmzMCaNWvw6quv4sUXX8TEiRPx5ptv4ty5c4iLi8P9999/y9do69atqKysRF5eHgDc8RDHDRs2ICIiAgaD4Y4eh6g/YSFD5OTaBzZGRUVBq9WitLS0R2+Ujz32GFxcXBAaGoqAgAA8/PDDAOTLJBcuXEBDQ0PH5Y9nnnkGADBkyBCMHz8e3377LXx8fHDu3LlOM6psNhtOnToFk8kEnU6Hn/zkJ7fMYO/H3b59O+bOnQt3d3cAwLPPPouPPvoIr776Knbt2oXf/OY3Hc83ZcqUbl+juLg4nDx5Es8//zwmTZqEadOmdfs5N7Njxw4sXboU33zzzW0/BlF/xEKGyMn9a6+FRqPpmPir1WrR2tracVtDQ8MtP6/9zxqNBgA6HqcrkiTBZrNh5MiR2LNnzw23FxYWYsCAAZAkqVdfy50+rs1mu+G23mb4V0OGDEFeXh527tyJ7du3Y+HChcjNzYWvr2+vHmf37t14+umn8dVXX3GgJlEvsUeGqJ+KiIjoGO524MCBjsnKt+P3v/89ALmQ+O677zB+/HikpKQgPz+/0y6c3Nzcjp4UJR73Rz/6Ef7whz+gqakJLS0tWL9+PaZOnQoAmDx5cqfn27FjR7ePZ7VaIUkSZsyYgVWrVsFms6G4uLjHXx8A7NmzB0888QS+/PJLxMXF9epziYgrMkT91jvvvIM5c+Zg/fr1SExMvKPJv+7u7hg3bhzKy8vxwQcfICwsDADw1Vdf4dVXX8VLL72E5uZmGI1GfPHFF4o9blpaGs6ePYvExEQAclPziy++CABYs2YNnnzySWzatAkWi6WjwLmVo0ePYtGiRbDZbGhra8MTTzyBkSNH3nC/jRs3YtGiRaisrMSXX36JFStW4KuvvkJCQgLmzp2LxsZGPP300x33/+Mf/4jY2NgevUZE/R2nXxORkCRJwtWrV6HX65WOQkQC46UlIiIiUi1eWiIiIXGxmIh6gisyREREpFosZIiIiEi1WMgQERGRarGQISIiItViIUNERESqxUKGiIiIVIuFDBEREanW/wclUwyFzZd0ywAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "alpha = 0.05\n",
    "number_sample = 12\n",
    "\n",
    "mean_cr = df.groupby(df['order_date'].dt.month)['is_paid_order'].sum() / \\\n",
    "        df.groupby(df['order_date'].dt.month)['is_paid_order'].count()\n",
    "\n",
    "se_cr = np.std(mean_cr, ddof=1) / np.sqrt(number_sample) \n",
    "\n",
    "t_score = st.t.ppf(1-alpha/2, number_sample -1)\n",
    "\n",
    "lower = mean_cr - t_score * se_cr\n",
    "upper = mean_cr + t_score * se_cr\n",
    "\n",
    "plt.figure(figsize=(8, 6), dpi=80, facecolor='w', edgecolor='r')\n",
    "\n",
    "\n",
    "plt.xlabel('number period is {}'.format(number_sample))\n",
    "plt.ylabel('conversion rate')\n",
    "plt.title('Dinamic of conversion rates')\n",
    "\n",
    "plt.plot(mean_cr)\n",
    "plt.fill_between(range(1,13), lower, upper, edgecolor='none', facecolor='b', alpha=0.2, interpolate=True)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [
    "rgsImuK1xlp9",
    "1ZvHLLlq6quU",
    "c6hjIKh_8FGN",
    "jJ8HpZg_PoPE"
   ],
   "name": "2.3.7: Statistical Tests, part1: Significance testing; AB Testing; Simpson paradox || Exercises.ipynb",
   "provenance": []
  },
  "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
}