421 lines
31 KiB
Plaintext
421 lines
31 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "markdown",
|
||
"source": [
|
||
"# Konum Ölçüleri\n",
|
||
"Konum ölçüleri, analiste verinin merkezinin veya başka nir konumun bulunduğu yerin nicel değerini elde etmeye yarar."
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 10,
|
||
"metadata": {
|
||
"collapsed": true
|
||
},
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"[10, 13, 14, 15, 17, 19, 20, 22, 24, 35, 39, 40, 43, 53, 53, 54, 55, 78, 81, 82, 82, 83, 86, 87, 90, 92, 94, 96, 99, 99]\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"import random\n",
|
||
"import datetime\n",
|
||
"\n",
|
||
"sd = datetime.datetime.timestamp(datetime.datetime.now()) * 1000\n",
|
||
"random.seed(sd)\n",
|
||
"\n",
|
||
"veri = [random.randint(10, 100) for _ in range(30)]\n",
|
||
"veri.sort()\n",
|
||
"\n",
|
||
"print(veri)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 11,
|
||
"outputs": [],
|
||
"source": [
|
||
"veri = [10, 14, 14, 17, 21, 21, 24, 27, 28, 30, 33, 35, 38, 41, 41, 42, 49, 51, 53, 54, 54, 60, 62, 71, 74, 76, 77, 77,\n",
|
||
" 79, 87]"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 12,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "45.333333333333336"
|
||
},
|
||
"execution_count": 12,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Ortalama\n",
|
||
"import statistics\n",
|
||
"\n",
|
||
"statistics.mean(veri)"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 13,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "41.5"
|
||
},
|
||
"execution_count": 13,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Medyan\n",
|
||
"# Örneklemin merkezi eğilimini aşırı (aykurı) değerlerden etkilenmeyecek şekilde yansıtmaktır.\n",
|
||
"# Öncelikle küçükten, büyüğe sıralama\n",
|
||
"# n tek ise -> ortadaki gözlem\n",
|
||
"# n çit ise -> ortadaki iki gözlemin aritmatik ırtalması\n",
|
||
"\n",
|
||
"statistics.median(veri)"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 14,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "[14, 21, 41, 54, 77]"
|
||
},
|
||
"execution_count": 14,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Mod: tepe noktası. En çok tekrarlayan değer\n",
|
||
"statistics.multimode(veri)"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 15,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "[26.25, 41.5, 64.25]"
|
||
},
|
||
"execution_count": 15,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Çeyrekler / Kartiller / Quartiles\n",
|
||
"# Q1: en küçük %25 i ayıran değer = .25 * (n+1) inci değer\n",
|
||
"# Q2: en küçük %50 yi ayıran değer = medyan = .50 * (n+1)\n",
|
||
"# Q3: en küçük %75 i ayıran değer = .75 * (n + 1)\n",
|
||
"\n",
|
||
"# Nasıl hesaplıyoruz\n",
|
||
"# örneğin veride 12 rakam var\n",
|
||
"# bu durumda .25 * (12 + 1) = 3.25 inci değeri bulmalız\n",
|
||
"# varsayalım 3. değer : 65, 4. değer ise 67 olsun\n",
|
||
"# Q1 = 65 + .25 * (67-65) = 65.5 olacaktır\n",
|
||
"\n",
|
||
"statistics.quantiles(veri)"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"source": [
|
||
"Bir örneklemin beş sayılı özeti\n",
|
||
"* min\n",
|
||
"* Q1\n",
|
||
"* Q2\n",
|
||
"* Q3\n",
|
||
"* max"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 52,
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Ortalama: 73.17 :: Medyan 73.5 :: 5 Sayılı Özet: 60 65.5 73.5 81.5 85\n",
|
||
"Ortalama: 144.46 :: Medyan 75 :: 5 Sayılı Özet: 60 66.0 75.0 83.0 1000\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"veriSeti = [60, 63, 65, 67, 70, 72, 75, 75, 80, 82, 84, 85]\n",
|
||
"\n",
|
||
"qs = statistics.quantiles(veriSeti)\n",
|
||
"med = statistics.median(veriSeti)\n",
|
||
"print(\"Ortalama: \", round(statistics.mean(veriSeti), 2),\n",
|
||
" \" :: Medyan\", med,\n",
|
||
" \" :: 5 Sayılı Özet:\", veriSeti[0], qs[0], qs[1], qs[2], veriSeti[-1])\n",
|
||
"\n",
|
||
"# Veriye extrem bir değer / aykırı bir gözlem ekleyelim\n",
|
||
"veriSetiExtrem = [60, 63, 65, 67, 70, 72, 75, 75, 80, 82, 84, 85, 1000]\n",
|
||
"qs = statistics.quantiles(veriSetiExtrem)\n",
|
||
"med = statistics.median(veriSetiExtrem)\n",
|
||
"print(\"Ortalama: \", round(statistics.mean(veriSetiExtrem), 2),\n",
|
||
" \" :: Medyan\", med,\n",
|
||
" \" :: 5 Sayılı Özet:\", veriSetiExtrem[0], qs[0], qs[1], qs[2], veriSetiExtrem[-1])"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 45,
|
||
"outputs": [],
|
||
"source": [
|
||
"# standard numpy and matplotlib library imports\n",
|
||
"import numpy as np\n",
|
||
"import matplotlib.pyplot as plt\n",
|
||
"\n",
|
||
"def dotplot(input_x, **args):\n",
|
||
" # Count how many times does each value occur\n",
|
||
" unique_values, counts = np.unique(input_x, return_counts=True)\n",
|
||
"\n",
|
||
" # Convert 1D input into 2D array\n",
|
||
" scatter_x = [] # x values\n",
|
||
" scatter_y = [] # corresponding y values\n",
|
||
" for idx, value in enumerate(unique_values):\n",
|
||
" for counter in range(1, counts[idx]+1):\n",
|
||
" scatter_x.append(value)\n",
|
||
" scatter_y.append(counter)\n",
|
||
"\n",
|
||
" # draw dot plot using scatter()\n",
|
||
" plt.scatter(scatter_x, scatter_y, **args)\n",
|
||
"\n",
|
||
" # Optional - show all unique values on x-axis.\n",
|
||
" # Matplotlib might hide some of them\n",
|
||
" plt.gca().set_xticks(unique_values)"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 30,
|
||
"outputs": [
|
||
{
|
||
"name": "stdout",
|
||
"output_type": "stream",
|
||
"text": [
|
||
"Nitrojen Yok Ortalama, Mod, Medyan 0.399 0.4 0.43\n",
|
||
"Nitrojen Var Ortalama, Mod, Medyan 0.565 0.505 0.26\n"
|
||
]
|
||
}
|
||
],
|
||
"source": [
|
||
"nitrojen_yok = [.32, .53, .28, .37, .47, .43, .36, .42, .38, .43]\n",
|
||
"nitrojen_var = [.26, .43, .47, .49, .52, .75, .79, .86, .62, .46]\n",
|
||
"\n",
|
||
"print(\"Nitrojen Yok Ortalama, Mod, Medyan\", statistics.mean(nitrojen_yok), statistics.median(nitrojen_yok), statistics.mode(nitrojen_yok))\n",
|
||
"print(\"Nitrojen Var Ortalama, Mod, Medyan\", statistics.mean(nitrojen_var), statistics.median(nitrojen_var), statistics.mode(nitrojen_var))"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"source": [
|
||
"Her iki veri setinin de mod, medyan ve ortalama değerleri çok benzer. Bir de görsel olarak bakalım"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 50,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "<Figure size 640x480 with 1 Axes>",
|
||
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAGsCAYAAADew6NRAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA+LklEQVR4nO3de1yUZeL//zcgGIqCaJ4PpQaZ4obH1jTWtMxj0Wk7uKaJtqtZuummVrtp7kf7amZp7qqlqWlZiZableaW5abmqgVKSp7yhCcQUBE5ze8PfnPvDMwMMMzAwP16Ph48HjPc99zXNfdp3nPd132Nn8VisQgAAMCk/Cu7AgAAAJWJMAQAAEyNMAQAAEyNMAQAAEyNMAQAAEyNMAQAAEyNMAQAAEytRmVXwBd06dJFOTk5uv766yu7KgAAoJTOnz+voKAg/fe//y3XcghDkq5du6b8/PzKrgYAACiDvLw8eWLsaMKQpIYNG0qStmzZUsk1AQAApdWnTx+PLIc+QwAAwNQIQwAAwNQIQwAAwNQIQwAAwNQIQwAAwNQIQwAAwNTKfGt9Xl6edu7cqW3btunHH3/U0aNHdenSJQUHB6tly5bq0aOHHn30UTVr1swjFbx06ZKWL1+uTZs26eTJk5Kk5s2b6+6779bw4cMVEhLikXIAAIA5+VnKOFrR4MGDlZyc7HKe4OBg/e1vf1NsbGy5Knfo0CGNGjVKp0+fdji9adOmevvtt9WmTZtylWMdp4BxhgAAqDo89fld5pahK1euyM/PTz169NA999yj6OhoXX/99bp06ZK+/fZbvfnmm0pPT9eUKVMUHh6umJgYtyp2+fJljR49WqdPn1atWrU0ceJEuzc9Z84cnT59WqNGjdKnn35KCxEAAHBLmcNQv3799NBDD6l169Z2/w8LC9Pjjz+u2267TQ8++KCysrI0e/Zst8PQkiVLdOrUKUnS/Pnz1bNnT2Pa448/rpYtWyouLk6nTp3S22+/rfHjx7tVDgAAMLcyd6B+/vnniwUhW23atNH9998vSfrll1+cXuJyJT8/Xx988IEkqWfPnnZByKpXr17G/z/44AN+Www+Jb/AosRDF7R1z0klHrqg/ILy/3YOAMA7vPLbZDfddJPx+OzZs2ratGmZXr97926lp6dLkgYOHOh0vgEDBmjbtm26ePGi9uzZo65du7pVX8CTvk84rcXrE5WakW38r37odRp9X5R6dCzbsQAA8D6v3FqfmppqPHanL8/+/fuNx9HR0U7ns51m+xqgsnyfcFozl++yC0KSlJqRrZnLd+n7hLK3lAIAvMsrYWjTpk2SpNDQUJeX1Jw5cuSIJMnf31/Nmzd3Ol/z5s3l5+dn9xqgsuQXWLR4faLLeZZ8so9LZgDgYzx+mWzt2rU6cOCAJOnhhx9WQEBAmZeRlpYmSapbt64CAwOdzhcUFKS6desqIyNDFy9edLlM651ojqSkpKhJkyZlridgK+lIarEWoaIupF9V0pFURbVtUEG1AgCUxKMtQ7/88otmzJghSWrWrJmeeuopt5Zz9epVSVLNmjVLnNc6T1ZWlltlAZ6Sluk6CJV1PgBAxfBYy1BaWprGjBmjrKwsBQYGas6cOapTp065lmm9BFbeeSTXAzK5ajUCSiu87nUenQ8AUDE80jJ0+fJlxcXF6fjx4/L399fs2bPVqVMnt5cXHBwsScrOLvkb9LVr1yRJtWrVcrs8wBNuaV1f9UNdB50GYcG6pXX9CqoRAKA0yh2Grl69qtGjR2v//v3y8/PTjBkz1L9//3Its169epKkzMxM5ebmOp0vJydHGRkZkgoHfQQqU4C/n0bfF+VynlH3dlCAf+laMwEAFaNcYSgnJ0djx47V7t27JUkvvviiHnjggXJXynoHWkFBgTEKtSOnTp2S9afV3LlrDfC0Hh2basoTXYu1EDUIC9aUJ7oyzhAA+CC3+wzl5ubqmWee0X/+8x9J0nPPPaehQ4d6pFIdOnQwHv/444+64YYbHM63d+9eh68BKlOPjk3VvUMTJR1JVVpmtsLrXqdbWtenRQgAfJRbLUP5+fmaOHGivv76a0nSH//4R40ePdpjlercubNx2etf//qX0/k2btwoqfCyWnn6KAGeFuDvp6i2DRTTqbmi2jYgCAGADytzGLJYLHrhhRf0xRdfSJKeeOIJTZgwwaOVCggI0COPPCJJ2rZtm77//vti8/znP//Rd999J0l65JFH3BrPCAAAoMxhaMaMGVq3bp0kafDgwXr22Wd15coVp395eXnFlnHy5ElFRkYqMjJSkydPdljOqFGj1KxZM1ksFo0dO1arV6/WmTNndObMGa1atUpPP/20pMLxjOLi4sr6NgAAACS50WfovffeMx5v2LBBGzZscDn/zJkzjV+xL4uQkBAtXrxYcXFxSklJ0bRp0zRt2jS7eZo2baolS5a49ftnAAAAkpd+td5T2rZtqw0bNmj58uXatGmTTpw4IUlq0aKF7r77bj3xxBPlHtgRAACYm5/Fem+6iVlHoHY1SjUAAPAtnvr89sqv1gMAAFQVhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqNdx50YkTJ5SQkGD8JSUlKTs7W5K0YsUKde/e3WMV/Pnnn7V69Wrt2rVLZ8+eVW5urkJDQ3XzzTerX79+io2NVWBgoMfKAwAA5uJWGOrbt6+n6+HQP//5T73xxhsqKCiw+/+FCxe0bds2bdu2Te+9956WLFmiRo0aVUidAABA9VKuy2Th4eGKiYlRTEyMp+pj2LRpk15//XUVFBSoWbNmevXVV7Vp0yYjAPXu3VuSdPDgQU2YMMHj5QMAAHNwq2Vo3rx5ioqKUvPmzSVJ8fHx2rp1q0crtmzZMklSrVq1tHz5crVo0cKYdv3116tLly4aPXq0vv32W+3evVsHDhzQzTff7NE6AACA6s+tlqH+/fsbQchbDhw4IEnq1KmTXRCy8vPz05AhQ4znR44c8Wp9AABA9eSzd5MFBQVJKgw9zgQEBBiPw8PDvV4nAABQ/fhsGGrfvr0kae/evTp79qzDeTZu3CipMAh16NChwuoGAACqD58NQ2PGjJG/v78uX76sESNG6KuvvlJqaqqysrKUlJSkiRMnavPmzQoMDNS0adMUEhJS2VUGAABVkFsdqCtCly5dtHDhQk2dOlWHDx/W2LFj7aYHBASoX79+iouLU8eOHUtcXp8+fZxOS0lJUZMmTcpdZwAAUPX4bMuQJPXu3Vvz58932IE6Pz9fZ86c0cmTJyuhZgAAoLrw2Zahq1evasqUKfr888/Vpk0bvf766+rcubNq1qypY8eO6b333tOGDRs0YcKEUo01tGXLFqfTXLUaAQCA6s1nw9CECRP09ddf68Ybb9SHH35o1yfo1ltv1a233qpGjRrp7bff1j//+U916dJFvXr1qsQaAwCAqsgnL5Pt27dPX3/9tSTpqaeecto5+umnn9Z1110nSfroo48qrH4AAKD68MkwtHv3buNxVFSU0/mCg4PVtm1bSdKhQ4e8Xi8AAFD9+GQYys7ONh67GnTRdnpJ8wEAADjik2GoYcOGxuPExESn8129etVoEWratKnX6wUAAKofnwxDt912m/z9C6u2ePFiXblyxeF8Cxcu1NWrVyWJztMAAMAtbt1Ndvz4caWlpdk9tzp06JBq1qxpPA8PD1fLli2LLSMyMlKS1K1bN61cudJuWpMmTfTQQw9pzZo1Onz4sB555BGNGTPG7tb61atXa/369ZKkZs2a6aGHHnLnrQAAAJNzKwwtXLhQ69atczht+vTpds9jY2M1a9asMpfx4osvKj09XV9++aWSk5M1fvx4h/O1atVK//jHPxQcHFzmMgAAAHx2nKGgoCC9+eab+u6777R+/XolJCTo3Llzys/PV926dRUZGam+ffsqNjZWtWrVquzqAgCAKsrPYrFYKrsSlc06ArWrUaoBAIBv8dTnt092oAYAAKgohCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqhCEAAGBqNdx50YkTJ5SQkGD8JSUlKTs7W5K0YsUKde/e3aOVzMvL06effqovvvhCBw8eVGpqqurUqaNGjRqpY8eO6tu3r+644w6PlgkAAMzBrTDUt29fT9fDqcOHD2v8+PFKTk62+39aWprS0tL0888/6+jRo4QhAADgFrfCkFV4eLiioqIkSVu3bvVIhWwdOXJEw4YN04ULFxQaGqq4uDjFxMSoUaNGunbtmpKTk7Vp0yZdvHjR42UDAABzcCsMzZs3T1FRUWrevLkkKT4+3uNhKD8/X5MmTdKFCxfUrFkzrVq1Sk2aNLGbp1GjRurVq5dHywUAAObiVhjq37+/p+tRzPr167Vv3z5J0quvvlosCAEAAHiCz95N9sEHH0iS2rdvr65du1ZybQAAQHXlk2EoLS1NCQkJkqSePXvaTcvNza2MKgEAgGqqXB2ovSUxMdF43LZtW6WmpmrhwoXavHmzzp07pxo1aujGG29Uv3799MQTT6hOnTqVWFsAAFCV+WQYOn36tPE4NTVVgwcPVmpqqvG/3NxcJScnKzk5WfHx8VqyZInatGnjcpl9+vRxOi0lJYU+SQAAmJRPXia7dOmS8fi1115TWlqaRowYoc2bNysxMVFfffWV4uLi5Ofnp1OnTmnMmDG6evVqJdYYAABUVT7ZMlRQUGA8zs3N1TPPPKOxY8ca/2vRooUmTZqkunXrau7cuTp27Jg+/PBDPfHEE06XuWXLFqfTXLUaAQCA6s0nW4Zq1aplPK5bt67i4uIczjdixAiFhYVJkjZv3lwRVQMAANWMT4ahevXqGY87duyomjVrOpwvKChIHTt2lCT98ssvFVI3AABQvfhkGGrbtq3xODQ01OW81ulXrlzxap0AAED15JNhqE2bNsalsvT0dJfzWn+XjNvrAQCAO3wyDAUFBSkmJkaS9NNPPzm9Uyw7O9sYnPGWW26psPoBAIDqwyfDkCQNGzZMknT58mX94x//cDjPokWLlJmZKUkaNGhQhdUNAABUH27dWn/8+HGlpaXZPbc6dOiQXYfn8PBwtWzZstgyIiMjJUndunXTypUri03v1KmT7r//fsXHx2vRokVKT0/XY489piZNmujMmTP64IMPtHr1aklSdHS0Bg8e7M5bAQAAJudWGFq4cKHWrVvncNr06dPtnsfGxmrWrFnuFKNp06bp8uXL2rRpk9asWaM1a9YUmyc6OlpvvfWWatTwySGTAACAj/PpBBEUFKT58+dr06ZNWrt2rfbv36/09HSFhITolltu0aBBgzRkyBCCEAAAcJufxWKxVHYlKpt1BGpXo1QDAADf4qnPb5/tQA0AAFARCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUarjzohMnTighIcH4S0pKUnZ2tiRpxYoV6t69u0craWv79u0aPny48XzmzJm6//77vVYeAACo3twKQ3379vV0PUrl6tWreumllyqlbAAAUD2V6zJZeHi4YmJiFBMT46n6uPTmm2/qxIkTatGiRYWUBwAAqj+3wtC8efO0ZcsWbd++XYsXL9Y999zj6XoVs2/fPi1fvlyhoaH685//7PXyAACAObh1max///6erodLeXl5euGFF5Sfn6/nnntO9evXr9DyAQBA9VUl7iZ75513dODAAXXq1EkPP/xwZVcHAABUIz4fho4ePaq33npLgYGBeuWVV+Tn51fZVQIAANWIT4chi8Wil156SdeuXdPIkSPVtm3byq4SAACoZnw6DK1Zs0a7du1Sy5YtNWbMmMquDgAAqIbc6kBdEc6ePas5c+ZIkqZNm6aaNWuWa3l9+vRxOi0lJUVNmjQp1/IBAEDV5LMtQ9OmTdOlS5c0ePBg9ejRo7KrAwAAqimfbBn6/PPPtWXLFoWGhmrKlCkeWeaWLVucTnPVagQAAKo3n2sZysnJ0d///ndJ0qRJkxhTCAAAeJXPtQxlZ2fr/PnzkqQXX3xRL774osv5p0yZYrQerV+/Xu3atfN6HQEAQPXhcy1DAAAAFcnnWoZq166t9evXu5xn3759RovRuHHjjD4/N954o7erBwAAqhmfC0MBAQElXurKzMw0Hjdt2pRLYwAAwG1uhaHjx48rLS3N7rnVoUOH7MYECg8PV8uWLYstIzIyUpLUrVs3rVy50p1qAAAAlJtbYWjhwoVat26dw2nTp0+3ex4bG6tZs2a5UwwAAIDX0YEaAACYmp/FYrFUdiUqm7UDtquBGQEAgG/x1Oc3LUMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUCEMAAMDUarjzohMnTighIcH4S0pKUnZ2tiRpxYoV6t69e7krlpaWpn//+9/asWOHfv75Z50+fVq5ubkKCwtTZGSk+vfvryFDhigoKKjcZQEAAPNyKwz17dvX0/Wwk5CQoEcffVR5eXnFpp0/f17nz5/Xtm3btHz5ci1cuFAtWrTwan0AAED15VYYsgoPD1dUVJQkaevWrR6pkCRdvXpVeXl5qlu3rgYOHKiYmBhFRESodu3aOnHihN5//32tXbtWycnJGjFihDZs2KDg4GCPlQ8AAMzDrTA0b948RUVFqXnz5pKk+Ph4j4ahkJAQTZw4UUOHDi0WcsLCwoyy33jjDZ04cUKrV6/WyJEjPVY+AAAwD7c6UPfv398IQt7Qvn17jRo1ymVrz6hRoxQWFibJs61SAADAXKrs3WSBgYG64YYbJElnz56t3MoAAIAqq8qGIUlKTU2VVHhZDQAAwB1VNgwlJSXpxIkTkqTo6OhKrg0AAKiqynU3WWWxWCyaOXOmJMnf31+///3vS3xNnz59nE5LSUlRkyZNPFY/AABQdVTJlqG33npLP/zwgyRp6NChuummmyq5RgAAoKqqci1DX3zxhRYsWCCp8K6zSZMmlep1W7ZscTrNVasRAACo3qpUy9D27ds1adIkWSwW3XDDDVq8eDE/xwEAAMqlyoShvXv3asyYMcrJyVHTpk21bNkyNWjQoLKrBQAAqrgqEYaSkpI0atQoZWVl6frrr9e7776rpk2bVna1AABANeDzYcj6+2OXLl1SWFiYli1bplatWlV2tQAAQDXh02Ho2LFjevLJJ5Wenq46deronXfe4c4xAADgUT4bhk6dOqXhw4fr/PnzqlWrlhYvXqwOHTpUdrUAAEA149at9cePH1daWprdc6tDhw6pZs2axvPw8HC1bNmy2DIiIyMlSd26ddPKlSvtpl24cEEjRoxQSkqKAgMDNXv2bEVGRurKlStO61S7dm133goAADA5t8LQwoULtW7dOofTpk+fbvc8NjZWs2bNKtPyv/32W/3666+SpNzcXI0dO7bE1xw8eLBMZQAAAEg+fJkMAACgIvhZLBZLZVeisllHoHY1SjUAAPAtnvr8pmUIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYGmEIAACYWg13XnTixAklJCQYf0lJScrOzpYkrVixQt27d/dYBXNycrRq1Sp99tln+vXXX5Wbm6smTZqod+/eGj58uBo2bOixsgAAgPm4FYb69u3r6Xo4dO7cOY0cOVLJycl2/z9y5IiOHDmitWvXasGCBeratWuF1AcAAFQ/5bpMFh4erpiYGMXExHiqPob8/HyNGTNGycnJqlGjhp555hl99dVX+u677zR79mzVr19f6enpGjt2rE6dOuXx8gEAgDm41TI0b948RUVFqXnz5pKk+Ph4bd261aMVi4+PV2JioiTpb3/7mx5++GFj2pAhQxQZGakHHnhAGRkZmjdvnmbPnu3R8gEAgDm4FYb69+/v6XoU895770mSbrzxRrsgZBUZGan7779fa9as0caNG/XCCy8oLCzM6/UqLUtBvrJP/Kz8yxcVEFJP17VoJz//gFK/Pr/AoqQjqUrLzFZ43et0S+v6CvD382KNfbfckurki3XOy8vTwR92KivtgmqFN1Bkt+6qUcOtw60YV/tWecot7z7r7rIra/v5Km9uB5ReZW2H6rb9q8r78czZ2cNOnTqlAwcOSJIGDhzodL4BAwZozZo1ysvL0zfffKP77ruvgmro2pUDO3Rh01LlX0o1/hdQp74a3P2kat98W4mv/z7htBavT1RqRrbxv/qh12n0fVHq0bGpV+rsq+VKclknX6xzzZQf5b/rA4X6XVHw/z8t8d+1VdD1EXXuN6Bc5bratw78muZ2ueXdZ91d9k85LStl+/kqb24HlF5lbYfqtv2r0vvxyVvr9+/fbzyOjo52Ot+tt94qf//Ct7Bv3z6v16s0rhzYobNrZ9ttfEnKv5Sqs2tn68qBHS5f/33Cac1cvsvuw0GSUjOyNXP5Ln2fcNrjdfblcl1NW7Zhn8/V+bPVHyts1zuqqyt20+rqisJ2vaPdX250u1xX+9aZtbPdLre8+2x56vzZ6o8rfPv5Km9uB5ReZW2H6rb9q9r78ckwdOTIEeNxy5Ytnc533XXXqUGDBpKko0ePer1eJbEU5OvCpqUu57mweaksBfkOp+UXWLR4faLL1y/5ZJ/yCyxu17GqlevK+q2HXU6v6Dr7qUD319pV+LjIVR7rc/9dHygvL6/M5Za4b1nsyyltueXdZ10pTZ3vr7VLfipwONkb289XeXM7oPQqaztUt+1fFd+PT4ahtLQ043H9+vVdzmsNQxcvXnQ5X58+fZz+paSklL/SUuF10SIpuKj8zFRln/jZ4bSkI6nFviUXdSH9qpKOuC6jrHy5XFdK+pys6Dq3qXFO9QKyigUSKz8/KdTvig7+sLPM5Za0b/n5FQ9CpSm3vPusK6Wpc72ALLWpcc7hdG9sP1/lze2A0qus7VDdtn9VfD8+GYauXr1qPK5Zs6bLea3Tr1y54nK+ipB/2XUgK2m+tMzSBYPSzldavl6uL5Xhanl1/a86nWYrK+1Cmcst7b5V1nLLu8964jWu1ltF7CO+wJvbAaVXWduhum3/qvh+fLIDtS0/Z193SzndasuWLU6n9enTp0x1ciYgpF655guve12pXl/a+UrL18v1pTJcLS+zINjpNFu1whuUudzS7ltlLbe8+6wnXuNqvVXEPuILvLkdUHqVtR2q2/aviu/HJ1uGgoP/d3K0/syHM9euXZMk1a5d26t1Ko3rWrRTQB3Xl/UC6tbXdS3aOZx2S+v6qh/q+uTfICxYt7R2XUZZ+XK5rpR093VF1/lwXkNdzK8li5PLdxaLlGGprchuZf+5mpL2LYtFbpVb3n3WldLU+WJ+LR3Oc/yTOt7Yfr7Km9sBpVdZ26G6bf+q+H58MgzVq/e/tGjbf8iRCxcKm/59YYwhP/8ANbj7SZfzNLjrSadjLAT4+xm3kzsz6t4OHh+DxZfLdeW+mDYup1d0nS3yV3xW4U/DFA0m1ucFXR9xa7yhEvctP/tySltuefdZV0pT5/isrrI4OQ15Y/v5Km9uB5ReZW2H6rb9q+L78ckw1Lp1a+Px8ePHnc6XnZ2t8+fPF3tNZap9821q9MCkYqk4oG59NXpgUoljK/To2FRTnuharPWhQViwpjzR1Wtjr/hqua6mjRjcwefqPPCxB5XedaQyZd9SmanaSu86slzjDLnatxo/MMntcsu7z5anzgMfe7DCt5+v8uZ2QOlV1naobtu/qr0fP4vFWeN66cXHx2vKlCmSPPOr9adOndKdd94pSRo3bpyefvpph/Pt3LlTw4YNkyS9+uqrbg+6aO0z5KpfUVkxArXnymUEanuMQF29VZURe6s7RqD2DG+/H099fvtkB+pmzZopMjJSBw8e1MaNG52GoY0bCweSCwgI8MqPxZaHn3+Aglt1cPv1Af5+impb9o625eWL5ZZUJ1+sc40aNdS+x+1eKdfVvlWecsu7z7q77Mrafr7Km9sBpVdZ26G6bf+q8n588jKZJP3hD3+QJB0+fFgfffRRsem//PKL4uPjJRX+ZIdtPyMAAIDScqtl6Pjx43Ydm2379Rw6dMhubKDw8HCHo0hHRkZKkrp166aVK1cWm279EdbExES9/PLLunDhggYPHqyaNWtqx44dmjVrlnJychQaGqrx48e78zYAAADcC0MLFy7UunXrHE6bPn263fPY2FjNmjWrzGUEBARo4cKFGjlypJKTkzVv3jzNmzfPbp6wsDAtWLBAzZo1K/PyAQAAJB/tM2TVsGFDrV27VqtWrdJnn32mY8eOKS8vT40bN1bv3r01YsQINWzoeIwSAACA0vDI3WRVnTfuJgMAAN7lqc9vn+1ADQAAUBEIQwAAwNQIQwAAwNQIQwAAwNR8+m6yinLu3Dnl5+cbHbEAAIDvS0lJUUBA+X/eg5YhSTVr1vTY70bZSklJUUpKiseXi6rBm9vf1bLLU25l1Rn2WFe+obK2Q3Xb/t58PzVq1LAb6Nld3FrvRdyyb27e3P6ull2eciurzrDHuvINlbUdqtv2rwrvh5YhAABgaoQhAABgaoQhAABgaoQhAABgaoQhAABgaoQhAABgatxaDwAATI2WIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqe/6n2Kmb79u1avXq1fvzxR128eFH16tVTdHS0Hn30Uf32t791e7lnzpzRli1btHPnTh08eFBnzpxRQUGBwsPD1b59ew0ePFj9+vWTv3/p8ujVq1f18ccfa/PmzTpy5IjS09MVFhamxo0bKzo6WgMGDFB0dLTb9TUrb21/Z8u+/vrrtX//fmOeevXq6fLlyw7L3bVrl3bu3Kl9+/bp2LFjunjxoi5fvqzAwEAFBgYqLy9Pubm5xmubNWumpUuXGstesWKFunfvLkn6+OOP9cILLziclpeXp08//VRffPGF9u7dq8zMzDK9z5kzZ+r++++3+9/OnTu1Zs0a7dmzR6mpqbJYLMrNzS1xWb/73e+0aNEih9M8fUyVlzf3HSt3j3tfW1fe5I3tsHPnTg0bNqxMr6lXr566detmV+7kyZO1bt26El9r3e/dKdf2WLZlsVi0ZcsWxcfHKyEhQenp6apTp45uuukmDR48WA888IDd9vf2/rxr1y6tWbNGP/74o86fP6+CggLVq1fP2B/79+8vPz+/Epdje746ePCgUlNTVadOHTVq1EgdO3ZU3759dccdd5S5fqa+tf7VV1+1+/Ao6sknn9Tzzz9f5uV+8cUXGj9+vEpatV27dtX8+fNVr149l/Pt3r1bkyZN0qlTp5zOExsbq1mzZpW5rmbmre1fmmW7Yi334Ycf1k8//eTWMqT/nSTPnz+vgQMHKiMjo9i0w4cPa/z48UpOTna7nPj4eLVv316SlJ+fr2nTpmnNmjVuLctZGPL0MVVe3tx3rNw97n1tXXmTt7aDO6HEUbneDkM1atTQ1q1b1aBBA7v/Z2VlaeLEiS5/Jb5Tp05avHix6tSp49X92WKxaPr06Vq9erXL+W677Ta99dZbCgkJcTpPac5X3bp108qVK8tcT9O2DK1YscLY+D179tS4cePUsmVL/frrr1qwYIG2bdumpUuXqkmTJmU+KLKysmSxWNSgQQMNGTJEvXr1UuvWrVWzZk0dPnxYS5cu1ZYtW7Rr1y796U9/0urVq51+Q9u9e7fi4uKUlZWlxo0ba9SoUerRo4fq16+vy5cv6+eff9bnn3+uWrVqlXudmIk3t7+zZY8fP147d+60m/evf/2r+vfv77Dchg0b6oEHHlCXLl10ww03aNu2bXrrrbckSTfccIMuXrxoBJy6des6bdF55ZVXlJGRoRYtWujEiRPG/48cOaJhw4bpwoULuu6665SdnS1J6tKlix5//HFdu3ZNc+bM0YULF+yWt3DhQk2ePFmZmZmKiIgwgpAkvfbaa0YQ6t69u55++mnddNNN+utf/6pNmzbJ399fBQUFatOmjVatWqWgoCC7Zdeo4fiU5Mljqry8ue9Ylee496V15U3e3A5dunTRnj17HE57//33NXv2bLv/3X777Ro3blyxcq06d+6sJUuWOC3Put+7KtcqNTVV99xzj/Lz89WrV69iQUiSpkyZYgShfv36KS4uTi1btlRaWpo2bNigxYsXa8+ePRo3bpx69+7t1f353XffNYJQZGSkxo0bp3bt2ikwMFDJyclauHCh9uzZox07dujll1/WnDlzHC7H9nwVGhqquLg4xcTEqFGjRrp27ZqSk5O1adMmXbx4scx1lCRZTCgtLc3SuXNnS0REhOWxxx6z5OXl2U3Py8uzPProo5aIiAhL586dLRcvXizT8r/99lvLmjVrLDk5OU7nmTp1qiUiIsISERFh+fzzzx3Oc+nSJcvvfvc7S0REhCU2NtaSkZFRpnrAMW9uf2fL3rx5syUiIsISGRlpbPeIiAjLV199VapyHS33+PHjlt/85jfGsjp06GC37B07dhjl9unTx7Jq1Spj2vfff2+5//77LREREZY77rjDEh0d7bTOd955p6Vv377Ga+fMmWM8fvvtt406Hjt2zNKuXTtLRESEZdiwYZbc3Fxj2vPPP2+JiIiwDBkyxHLzzTdbIiIiLIsWLSr1evXUMVVe3j53WCzlP+59ZV15U0Vsh5LKvfPOO4utw6LlTpgwwRIREWEZOnSoR8q3WCyWpUuXutx2O3bsMKb/5S9/cbiMdevWGfN07NjRq+vxjjvuMM4z6enpxaZfu3bNMmTIEEtERISlXbt2lrS0tGLz5OXlGeer3r17W06fPl3mepSk6n0d8IBPPvlEly5dkiRNmjRJAQEBdtMDAgI0adIkSdKlS5f06aeflmn5vXr10sMPP6zAwECn8zz33HPGt7GtW7c6nGfp0qU6ffq0AgMDNXfuXNWtW7dM9YBj3tz+jpZ96dIlTZs2TZKKXbr4/vvvS1Wuo+W2aNFCXbp0MeZ57rnn7JadlZVllPvyyy/ruuuuM6Zt27ZN+/btkyT17dtXV65ccVrnadOmacCAAcZrN27caNR3yJAhdv/Pz8836uKoladu3bqKiYmRJK1cubLESzlWnjqmysvb5w6p/Me9r6wrb6qI7VBSudYyw8LCdOeddzos17Yl1lPWr19frFxbGzZskCT5+/sXOydY3XfffYqIiJAko0XYG+sxLS1NZ86ckSTFxMQoNDS02DxBQUHq37+/pMLL7MePHy82z/r1643z1auvvmrX6uYppgxD1ubDZs2a6dZbb3U4T3R0tJo2bSpJ+uqrrzxeh/DwcNWvX1+SdO7cuWLT8/Pz9dFHH0mSevfurRtuuMHjdTArb25/R8uePXu2zp07Z2xv2z4au3fvLlW5zup89OhRSYUnlOHDhyssLMyYtmbNGp07d06DBg1Sz5497cr597//LUlq3769cf3dUZ2tr23VqpXx2pMnT0oqbFK//vrrjf///PPPkqRatWopKirK6Trq1q2bpML9/scff3Q6X1mVdEx5grfPHRV13FfEuvKmyjqHW8tt2LChfv31V0nSgAED7C732pZ7+vRpj5RrdeDAAR04cMBhuVbW47B169Zq2LCh02VZj0NJaty4sVfWo239XHWOtr1MGx4eXmz6Bx98IKnwfNW1a9cy1aG0TBmGkpKSJKnEu6+s023v/vGU3Nxco79H7dq1i01PSkoyTlJFP8hKc1cOnPPm9i+67B9++EEffvihQkNDjW9gLVu2NOY/duxYqcp1VOevvvrKCCadO3cutuxvvvlGoaGhmjp1arEyjhw5Iqlw37Jddm5url2dHb3W6r777rN7bv3GXKdOHZcnPtvAlpiY6HS+sirpmPIEb587Kuq4r4h15U2VdQ63lmu7zmJjY52Wm56ebvf/goIC5eXluV2+bWdsR+VK/zsOS2pRtD0Omzdv7nJed9djSEiIEei/++47owXaVl5enjZt2iRJatu2bbG6pKWlKSEhQZJ3jwnThaGzZ8/q8uXLkqQWLVq4nNc6/fLlyx7/9vTNN98oJydHkuMD2vZDom3btjpx4oSmTJmi3/72t+rQoYOioqL04IMP6t1339W1a9c8WrfqzJvbv+iyr127ppdeekkWi0WjR482TgS233yuXr1abNlFy7VdbrNmzXTq1Cl9+OGHevbZZyUVdr78y1/+UmzZFotFEydONFoAHLn++uuNZX/99dfq0KGD/vCHP8hisahTp04Ov3lKhSe5vn37FvufVHgydnb5Kzk5Wa+++qrxfPbs2YqLi9Mnn3xSrg8JqeRjqrwq4txRUce9t9eVN1XWOdy2XOulnzZt2qhjx45Oy7Xu08nJybrrrrvUvn17tW/fXt27dy/zfp+Xl6d//etfLsuV/nccljRERkpKivHY2Y0LRd+PO+tx3LhxkqRTp05p5MiR2r59u9LT03X58mXt2bNHf/zjH5WYmKiQkBC98sorxb5IFT0mUlNT9corr+iOO+5QVFSUOnTooMGDB2vBggVGEHSH6cKQbU9zR73wbdl+iLjdQ92BnJwco8d8rVq1dO+99xabx7Z5NTk5WUOGDFF8fLzS0tKMZSQmJmrmzJn6/e9/r/Pnz3usftWZN7d/0WUvWLBAx44dU+fOnXX77bcb04reOlp02UXLtZ2+aNEi3XnnnXrppZeUl5en4OBgrVq1SrfcckuxZbdq1UoPPfSQyzrb3rZd9Fvb119/rXvvvVeHDx8u9rqePXsWC0pt27aVVNhXyVmLT3p6ut235ZycHH333Xf6y1/+okceecT4kCmr0hxT5VUR546KOO4rYl15U2Wdw21ff/XqVUnFW0cdlSsV7vfHjx9XQUGB8bys+/22bduMOzudlSv97zg8cuSIy/3jv//9b4llWpVnPQ4aNEgzZsxQ7dq1tXfvXg0fPlzdu3dX586d9eijj2rXrl2KjY3VRx99pE6dOhV7ve0xkZqaqsGDB+u9997T2bNnjfHLkpOTNX/+fKfnq9IwXRjKysoyHtesWdPlvLYdTm1fV15/+9vfjMsjzz77rMNrpLap/u9//7tyc3M1YcIEffPNN0pMTNTGjRuNge5+/vlnPfPMM6XujGpm3tz+tvOkpaVp6dKlCgwM1PTp042Tp6RiHVuLLrtoua7KHj16tG6++WaHyyrNIGYrVqwwHg8dOlQ1atRQjRo19MADD8jPz0+nTp3SmDFjjFYEK9twZ3X33Xcbj19//XWjM7VUeDIdOXKkpk6dalenW265xRjQLTExUSNHjnTrWCvNMVVeFXHuqIjjviLWlTdV1jm86Ov9/f2dBknbcgcNGqR3331XW7ZsUWJionbu3KnFixeXeb+3XiJzVa70v+OwoKBAc+fOdTjPp59+avR5klRi61R51+ODDz6o2bNnO9zXrl27pjNnzjgNhLatPa+99prS0tI0YsQIbd68WYmJifrqq68UFxdnd76yPd+WlunCUFmUZjTMslq2bJni4+MlFXaQfOKJJxzOZ3uCy83N1YwZM/THP/5RTZo0UVBQkNq0aWN8O5SkPXv2uBxgC2VXnu2/bt065eXlaeTIkcY3NU+UO3XqVGN5tWrV0htvvKFHH31UKSkpys/PtxujxLZzszO2geXf//638vLyFBcXp//7v//ThAkTJBX2ayo6PtJNN91UbFnt2rXToEGDJBXeJTdy5Ejt2rVL6enpevTRR9WwYUO98cYbdh9g9evX17JlyzR06FBJ0qFDh/Tuu++WWG9bpT2mKpK7+463j3tfXFfe5I1zuFWPHj3UqFGjEssdOnSofvvb36p58+YKCgpSWFiYYmJiyrTfZ2ZmGjc9uCpXkvr06WP0IYyPj9eECROUkJCgjIwMHT16VPPnz9fUqVNL3bm5NNNdSUtL07BhwzRmzBi1atVKixYt0n/+8x9t375dy5cvV8+ePbV9+3bFxcXp/fffL/Z6a2uaVHhMjBs3TpMnT1bLli0VFBSkFi1aaNKkSXbnqw8//LDM9TRdGLIdpKyka+7WDq9FX+eu9evXG/0lbr31Vs2dO9fpTmZbXuvWrZ02iz7zzDNGT/zNmzeXu47VnTe3v+08p0+fVqtWrTRmzJhi04p2+iu67KLl2k7fuXOnDh06pFatWmnDhg1q1qyZkpKSNGbMGC1btszuEpTttzlnbDuCFq3ziBEjjE6WRe96c7Y+XnnlFfXq1UtS4fD+Q4cOVffu3dWnTx/NnDlTeXl5mjFjhjF/aGio/Pz89Pzzzxsn+LLcvluWY6q8KuLc4c3jviLXlTdV1jm86OuddWAubbll2e83btxotM66Kte63DfffFMdOnQwXvvQQw+pW7duuueee7RgwQIFBwdr9OjRxmuc9Q0sy/txxPrl6ocfflDXrl21cuVK/e53v1ODBg0UHh6u2267TUuWLNGAAQOUn5+vV155Rb/88ovdMmzLq1u3ruLi4hyWZXu+cuez0HRhyPa25tTUVJfz2o68a9vz3h2bN2/W1KlTZbFY1K5dOy1ZssTlTmVbT1e3EjZo0EA33nijJBXbiVCcN7d/0Z81ePnll41WENtpRfvmFF120XJtX2sdE+bll19W8+bNjeCSlJSkN954w245roa1t7L2NXJU56CgIKOTZtFRqJ2tj1q1amnx4sWaO3euevbsqXr16ikwMFCNGjVSbGys1q1bZzc2UuPGjY2yrL8ndPTo0VI1c5f1mCqvijh3eOu4r+h15U2VdQ63LTcwMLDYDQTulFva/d46tpCjGxccadCggd5//31NmzZN3bp1U2hoqAIDA9WsWTM9/vjj+te//qXIyEhj/qLjC7n7forasmWLcQfaM88843DsKz8/P/35z3+WVNhSbW25tLJd7x07dnR6adT2fOXOZ6Hpfo6jUaNGql27tq5cueJwcCdb1tuWQ0JCXDZLluTbb7/VhAkTlJ+frzZt2mjp0qUl3vZoe2nF0UBVtqzTHd22CHve3P6NGjVScHCwcUIbMWKEw/mszd1W1pPhrl27VLduXYflWutsvbbvaNlF+/VY7zZzZdeuXXY/xeGszraX0xzV2Za/v78GDhyogQMHOlzWt99+azy2vSPGtpNmZmamgoODndbbnWOqvCri3OGN474y1pU3VcY5XJJdUGnSpInLlteylFvSfn/s2DHt3btXUmE/wNK0+EqF4eCRRx7RI4884nC6bZAsqd+Zu+vRtkXZ1dhjLVq0UL169XTx4kUdOnTIblpFfRaarmVIkvFbSiUN+GbdAW1/e6msduzYoXHjxik3N1ctWrTQsmXLStVhMSoqymjCLjpWRVHW3v116tRxu55m4s3t36ZNG7fr5arc8uyDJbG2MFQUaxiqWbOm3a9t29754uqk5+4x5QnePnd4+rivzHXlTRV5Dreyts5IKrHlsizllrTf25Zb0iWysrD9UmIdvNUZd9ej7eW10vZLKjpfmzZtjFZMb34WmjIM9enTR1Jh2nX2q+A//fSTcduxdf6y2rt3r/70pz8pOztbjRs31rvvvlvqVN2oUSMjSf/www9Ok/vZs2eNuwKKXvKAY97c/vfcc4/x+P/9v/+n9evXG39PPvlksfkHDBhgTK9du7bTcm0f2y7X9vq5tfOy1dixYzV37lzNnTtXa9euNcb7kGT0JwgODrZrdrdd9po1axwOyFe0zmWRlpZm9I0YMGCA0eR+7do14+TcunVrp99+y3NMeYK3zx2ePO4re115U0Wdw60sFotdn57z5897pNyS9nvbclu1amV0jC6vw4cP67vvvjOee2s92o6Abf05DUdOnTplDB9hHe3aKigoyPgJn59++slpEM3OzjYGZ3Tns9CUYejee+81+lPMmTOn2CWA/Px841eJ69SpY/f7S6WVlJSkUaNGKSsrSw0aNNC7775b4iifRVnv9HDVO/61114zetsX/TCEY97c/g8++KCx7I8//lgRERFq166d2rVrp9GjRxfrxzNw4EBjuiSj3Nq1a+uuu+5yWGfrckNCQuz6EhQd+r979+7G5aoOHTrYnWSs3zCvXr2qzMxMh3XeunWrw+Zm2zqX1NdAKvzgyM/PV05OjiZNmqSMjAyFhIRo/PjxkgrvFpkxY4bxDdnZ+vbEMVVeFXHu8MRx7wvrypsqYjvY2rlzpxEIrJ2NSyo3JCTE5Tm5NPu9bbmeGg8qMzNTEydOVG5urho3bmx8ofHGeuzRo4fx+M0333R6C7/tEADWGzBsDRs2TFLhoI//+Mc/HC5j0aJFxtAU7nwWmjIM1atXz+hP8cMPP2j06NH66aeflJaWpp9++kmjR4/Wrl27JBV2+iraMfbkyZOKjIxUZGSkJk+eXGz5hw8f1siRI3Xp0iWFhIRowYIFatiwoa5cueLwz1nSHThwoDEWxbRp0/T666/ryJEjysjIUGJiop599ll98sknkgpbJGx/awbOeXP7u1r28ePHi/1W0Llz5xyWe+XKFfXr108zZszQ119/rcuXLxvX/q13ZgwZMsTo2NikSRO72+qLOnnypKZMmWI8v+mmm4zxapYvX27cKv/DDz/o8ccf17hx47Rw4UKHyyraBG27PmJjYzVu3Dh98cUXOnHihDIzM7Vq1Sr17NlTd9xxh7Zt2yZ/f3+9+OKLslgs+vLLL/X4448bH/oREREaPnx4sTI9dUyVl7fPHVL5j3tfWVfeVBHbwZZ1ffv5+Rl3YZVU7m233aaHH35Y8+bN0/bt25WSkqKDBw8a5d5+++0l7ve25boaaLGo6dOn6/nnn9c333yjlJQUZWZm6ujRo1q1apUGDRqkpKQk1apVS3PnzjW+lHhjPXbq1MkYk2znzp36wx/+oK1btyo1NVUXL17Uzp079dRTTxkja//mN79x+OOznTp1Ms5XixYt0l//+lcdOHBAGRkZOnjwoKZNm2acr6KjozV48OBSrysrP4uJR+qbNWuWli1b5nT6k08+qeeff77Y/0+ePGk0F8bGxmrWrFl20+fPn68FCxaUuh7NmjUr1qnW6tKlS3rqqaeK3dps684779Rrr71WZe8OqSze2v6lWba39OrVy2j+XrFihdEnx7bO1mnR0dF67rnnjN8FKi3b5RZd9s0332z8kGRZdevWTa+99prDH5f05DHlCd7cd6TyHfe+tq68ydvbQSocZPD2229XVlaWunfvrhUrVpSq3EaNGmnmzJklvgdn+72jcktr8uTJdr9jVlTTpk01e/Zs485Ob67HjIwMjR071ghVzrRv316LFi1yOj5aTk5Oieer6OhovfXWWy5/gsgZ091NZmvy5MmKiYnRqlWr9OOPPyo9PV1hYWGKjo7WY489Znw7q0x16tTRe++9p/j4eH366af65ZdfdOnSJYWGhhq/U9SnT58qOV5IZfPm9ne17BYtWuidd96RVHib6pUrV4xpd911lyZNmiSp8CR50003KSEhQefPn1dqaqoxRlHNmjVVo0YN5eTk2NU5JSXFri+AK0FBQZo/f742bdqktWvXav/+/UpLSzPGr8nPzzcuxXTq1Mlly5NV06ZN1aNHD/33v//V6dOnlZGRoeDgYIWEhKh27drKy8tTRkaGMjMzFRQUpIYNGyoqKkqDBg1STExMldmPvX3u4LgvnYo4h2/atMkYddl6ebk05Vovb+3du1eHDh1SWlqaMjIyjEtRzZs310svveR0v3dUbmk9/PDDql27tnbv3q2zZ88a+07r1q1111136aGHHrK7a82b6zE0NFQrVqzQl19+qc8++0z79+/XhQsXZLFYFBYWpvbt26t///4aOHCgw1vvrRydr9LT0xUSEqJbbrlFgwYN0pAhQ0r8nTVnTN0yBAAAYMo+QwAAAFaEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGqEIQAAYGr/H9eYtg5v496bAAAAAElFTkSuQmCC"
|
||
},
|
||
"metadata": {},
|
||
"output_type": "display_data"
|
||
}
|
||
],
|
||
"source": [
|
||
"dotplot(input_x=nitrojen_yok)\n",
|
||
"dotplot(input_x=nitrojen_var)\n",
|
||
"\n",
|
||
"# Grafikten görüldüğü üzere, nitrojen_var veri setinin değişkenliği nitrojen_yok'tan çok daha fazla."
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "markdown",
|
||
"source": [
|
||
"# Değişkenlik Ölçüleri (Measures Of Variation)\n",
|
||
"\n",
|
||
"Örneklem değişkenliği veri analizinde önemli bir rol oynar.\n",
|
||
"\n",
|
||
"Küçük veri analizi problemlerinde bile, belirli bir istatistiksel yöntemin başarısı, örneklemdeki gözlemler arasındaki değişkenliğin büyüklüğüne bağlı olabilir. Bir örneklemdeki konum ölçüleri, bir veri kümesinin uygun bir özetini sağlamayabilir. Örneğin, bir önceki örnekte, örneklem değişkenliğini hesaba katmadan azot kullanımının büyümeyi arttırdığı sonucuna varamayız.\n",
|
||
"\n",
|
||
"Pek çok konum ölçüsü olduğu gibi birçok değişkenlik (veya yayılım, saçılım, dağılım) ölçüsü de vardır.\n"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 53,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "25"
|
||
},
|
||
"execution_count": 53,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Açıklık (Range)\n",
|
||
"tmpData = np.array(veriSeti)\n",
|
||
"tmpData.max() - tmpData.min()"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 54,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "8.0"
|
||
},
|
||
"execution_count": 54,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# IQR : Interquartile Range\n",
|
||
"# IQR, verilerin orta %50’sindeki yayılımı ölçer\n",
|
||
"qs = statistics.quantiles(veriSeti)\n",
|
||
"qs[2] - qs[0]"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 56,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "20.3"
|
||
},
|
||
"execution_count": 56,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Varyans\n",
|
||
"# s2 ile gösterilen örneklem varyansı, her gözlem ile örneklem ortalaması (x̄) arasındaki\n",
|
||
"# farkların karelerinin toplamının örneklem büyüklüğünün bir eksiğine bölünmesiyle elde\n",
|
||
"# n-1 --> degree of freedom\n",
|
||
"\n",
|
||
"statistics.variance([3,0,-2,-1,5,10])"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 58,
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": "4.51"
|
||
},
|
||
"execution_count": 58,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"# Standart sapma\n",
|
||
"# s = sqrt(s2)\n",
|
||
"\n",
|
||
"round(statistics.stdev([3,0,-2,-1,5,10]), 2)"
|
||
],
|
||
"metadata": {
|
||
"collapsed": false
|
||
}
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "Python 3",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"language_info": {
|
||
"codemirror_mode": {
|
||
"name": "ipython",
|
||
"version": 2
|
||
},
|
||
"file_extension": ".py",
|
||
"mimetype": "text/x-python",
|
||
"name": "python",
|
||
"nbconvert_exporter": "python",
|
||
"pygments_lexer": "ipython2",
|
||
"version": "2.7.6"
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 0
|
||
}
|