{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Find the max of a list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "liste = [4,5,6,7,82,3,1,5,28,2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[4, 5, 6, 7, 82, 3, 1, 5, 28, 2]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "liste"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "82"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "max(liste)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "maxi = liste[0]\n",
    "for i in range(len(liste)):\n",
    "    if(maxi < liste[i]):\n",
    "        maxi = liste[i]\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "82"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "maxi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def maximum(liste):\n",
    "    maxi = liste[0]\n",
    "    for i in range(len(liste)):\n",
    "        if(maxi < liste[i]):\n",
    "            maxi = liste[i]\n",
    "    return maxi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "maximum([3,4,2,5,9])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradyan Inise hazirlik"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Hata(w):\n",
    "    return ((w-3)**2)+5\n",
    "\n",
    "def turev(w):\n",
    "    return 2*(w-3)\n",
    "\n",
    "def gradyan_inis(w = 0, alpha = 0.05, dongu = 20):\n",
    "    W = np.zeros(dongu)\n",
    "    \n",
    "    for i in range(dongu):\n",
    "        W[i] = w\n",
    "        w = w - alpha * turev(w)\n",
    "\n",
    "    return W"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 40.        , -30.3       ,  32.97      , -23.973     ,\n",
       "        27.2757    , -18.84813   ,  22.663317  , -14.6969853 ,\n",
       "        18.92728677, -11.33455809,  15.90110228,  -8.61099206,\n",
       "        13.44989285,  -6.40490356,  11.46441321,  -4.61797189,\n",
       "         9.8561747 ,  -3.17055723,   8.55350151,  -1.99815136])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = 0.95\n",
    "W = gradyan_inis(w = 40, alpha = 0.95)\n",
    "W"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1080db7b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "t = np.arange(-50,50,0.5)\n",
    "\n",
    "plt.plot(W,Hata(W),'k')\n",
    "plt.scatter(W,Hata(W), color = 'red')\n",
    "plt.plot(t,Hata(t))\n",
    "plt.xlabel('w'); plt.ylabel('Hata(w)'); plt.title(\"alpha = \" + str(a))\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1082d6278>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = 0.05\n",
    "W = gradyan_inis(w = 40, alpha = a)\n",
    "t = np.arange(-50,50,0.5)\n",
    "\n",
    "plt.plot(W,Hata(W),'k')\n",
    "plt.scatter(W,Hata(W), color = 'red')\n",
    "plt.plot(t,Hata(t))\n",
    "plt.xlabel('w'); plt.ylabel('Hata(w)'); plt.title(\"alpha = \" + str(a))\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7.9981513553900685"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "W[-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f990d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = 0.05\n",
    "W = gradyan_inis(w = 40, alpha = a, dongu=100000)\n",
    "t = np.arange(-50,50,0.5)\n",
    "\n",
    "plt.plot(W,Hata(W),'k')\n",
    "plt.scatter(W,Hata(W), color = 'red')\n",
    "plt.plot(t,Hata(t))\n",
    "plt.xlabel('w'); plt.ylabel('Hata(w)'); plt.title(\"alpha = \" + str(a))\n",
    "plt.grid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.0000000000000018"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "W[-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.arange(10, 100, step= 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = 2 * x + 5 + 20 * np.random.randn(18)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'y')"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a14b14198>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(x,y)\n",
    "plt.xlabel(\"x\")\n",
    "plt.ylabel(\"y\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Built-in Linear Regression from scikit learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = LinearRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = x.reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(x,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[10],\n",
       "       [15],\n",
       "       [20],\n",
       "       [25],\n",
       "       [30],\n",
       "       [35],\n",
       "       [40],\n",
       "       [45],\n",
       "       [50],\n",
       "       [55],\n",
       "       [60],\n",
       "       [65],\n",
       "       [70],\n",
       "       [75],\n",
       "       [80],\n",
       "       [85],\n",
       "       [90],\n",
       "       [95]])"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_test = np.arange(100)\n",
    "x_test = x_test.reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[10],\n",
       "       [15],\n",
       "       [20],\n",
       "       [25],\n",
       "       [30],\n",
       "       [35],\n",
       "       [40],\n",
       "       [45],\n",
       "       [50],\n",
       "       [55],\n",
       "       [60],\n",
       "       [65],\n",
       "       [70],\n",
       "       [75],\n",
       "       [80],\n",
       "       [85],\n",
       "       [90],\n",
       "       [95]])"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = model.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'x')"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1549c2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.plot(x_test, y_pred, 'r')\n",
    "plt.scatter(x,y)\n",
    "plt.xlabel(\"x\")\n",
    "plt.ylabel(\"x\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## All in one place"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline \n",
    "from sklearn.linear_model import LinearRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Collect data\n",
    "x = np.arange(10, 100, step= 5)\n",
    "y = 2 * x + 5 + 20 * np.random.randn(18)\n",
    "\n",
    "x = x.reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train test split\n",
    "x_test = np.arange(100)\n",
    "x_test = x_test.reshape(-1, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Apply Machine learning algorithm\n",
    "model = LinearRegression()\n",
    "# learn parameters with fit\n",
    "model.fit(x,y)\n",
    "# prediction\n",
    "y_pred = model.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'x')"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x110ab3390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(x_test, y_pred, 'r')\n",
    "plt.scatter(x,y)\n",
    "plt.xlabel(\"x\")\n",
    "plt.ylabel(\"x\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8970703805274145"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Evaluation\n",
    "# print the R-squared value for the model\n",
    "model.score(x, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "\n",
    "%matplotlib inline \n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 284,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.332974</td>\n",
       "      <td>0.563395</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.360866</td>\n",
       "      <td>0.673493</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.704928</td>\n",
       "      <td>2.181285</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         x0        x1    y\n",
       "0  1.332974  0.563395  0.0\n",
       "1  2.360866  0.673493  0.0\n",
       "2  1.704928  2.181285  0.0"
      ]
     },
     "execution_count": 284,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Collect data\n",
    "x0 = np.random.randn(100) + 2\n",
    "x1 = np.random.randn(100) + 2\n",
    "\n",
    "x0_ = np.random.randn(100) + 3\n",
    "x1_ = np.random.randn(100) + 3\n",
    "\n",
    "xx0 = np.concatenate((x0,x0_))\n",
    "xx1 = np.concatenate((x1,x1_))\n",
    "y = np.concatenate((np.zeros(100) ,np.ones(100)))\n",
    "\n",
    "d = {'x0': xx0, \"x1\":xx1, \"y\":y}\n",
    "data = pd.DataFrame(d)\n",
    "\n",
    "data.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {},
   "outputs": [],
   "source": [
    "c=['b','r']\n",
    "mycolors = [c[0] if i==0 else c[1] for i in y]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1c414780>"
      ]
     },
     "execution_count": 286,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1c430630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.plot.scatter('x0', 'x1' ,c=mycolors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train test split\n",
    "from sklearn.cross_validation import train_test_split\n",
    "X = data.iloc[:, :-1].values\n",
    "y = data.iloc[:, -1].values\n",
    "\n",
    "# Split dataset into train ab=nd test sets\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size = 1/4, random_state = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(150, 50)"
      ]
     },
     "execution_count": 265,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train), len(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Apply Machine learning algorithm\n",
    "model = LogisticRegression()\n",
    "# learn parameters with fit\n",
    "model.fit(X_train, y_train)\n",
    "# prediction\n",
    "y_pred = model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.88"
      ]
     },
     "execution_count": 267,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Evaluation\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "accuracy_score(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[20,  3],\n",
       "       [ 3, 24]])"
      ]
     },
     "execution_count": 268,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "confusion_matrix(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}