day52-types-of-gradient-descent/stochastic-gradient-descent-from-scratch.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "60e43fd7",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "\n",
    "import numpy as np\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a90f5266",
   "metadata": {},
   "outputs": [],
   "source": [
    "X,y = load_diabetes(return_X_y=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "51ea6596",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(442, 10)\n",
      "(442,)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7077f503",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "eb4d3f4e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LinearRegression()"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg = LinearRegression()\n",
    "reg.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8af3b5af",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  -9.16088483 -205.46225988  516.68462383  340.62734108 -895.54360867\n",
      "  561.21453306  153.88478595  126.73431596  861.12139955   52.41982836]\n",
      "151.88334520854633\n"
     ]
    }
   ],
   "source": [
    "print(reg.coef_)\n",
    "print(reg.intercept_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "33459e4e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4399387660024645"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = reg.predict(X_test)\n",
    "r2_score(y_test,y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ed75f3c1",
   "metadata": {},
   "outputs": [],
   "source": [
    "class SGDRegressor:\n",
    "    \n",
    "    def __init__(self,learning_rate=0.01,epochs=100):\n",
    "        \n",
    "        self.coef_ = None\n",
    "        self.intercept_ = None\n",
    "        self.lr = learning_rate\n",
    "        self.epochs = epochs\n",
    "        \n",
    "    def fit(self,X_train,y_train):\n",
    "        # init your coefs\n",
    "        self.intercept_ = 0\n",
    "        self.coef_ = np.ones(X_train.shape[1])\n",
    "        \n",
    "        for i in range(self.epochs):\n",
    "            for j in range(X_train.shape[0]):\n",
    "                idx = np.random.randint(0,X_train.shape[0])\n",
    "                \n",
    "                y_hat = np.dot(X_train[idx],self.coef_) + self.intercept_\n",
    "                \n",
    "                intercept_der = -2 * (y_train[idx] - y_hat)\n",
    "                self.intercept_ = self.intercept_ - (self.lr * intercept_der)\n",
    "                \n",
    "                coef_der = -2 * np.dot((y_train[idx] - y_hat),X_train[idx])\n",
    "                self.coef_ = self.coef_ - (self.lr * coef_der)\n",
    "        \n",
    "        print(self.intercept_,self.coef_)\n",
    "    \n",
    "    def predict(self,X_test):\n",
    "        return np.dot(X_test,self.coef_) + self.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "028e7219",
   "metadata": {},
   "outputs": [],
   "source": [
    "sgd = SGDRegressor(learning_rate=0.01,epochs=40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "aafb36fa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "157.2641767094034 [  58.34618831  -41.65635944  314.02875897  219.47021721   30.2641263\n",
      "   -8.98591625 -165.98040412  132.42543708  294.42536821  122.71528461]\n",
      "The time taken is 0.17459607124328613\n"
     ]
    }
   ],
   "source": [
    "start = time.time()\n",
    "sgd.fit(X_train,y_train)\n",
    "print(\"The time taken is\",time.time() - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "829e9938",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = sgd.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "128832b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4205755931585481"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(y_test,y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "457600ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import SGDRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "fdd03179",
   "metadata": {},
   "outputs": [],
   "source": [
    "reg = SGDRegressor(max_iter=100,learning_rate='constant',eta0=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "37fd9602",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\91842\\anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_stochastic_gradient.py:1220: ConvergenceWarning: Maximum number of iteration reached before convergence. Consider increasing max_iter to improve the fit.\n",
      "  warnings.warn(\"Maximum number of iteration reached before \"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "SGDRegressor(learning_rate='constant', max_iter=100)"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "6e28beda",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = reg.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "3d3d24d4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.43059547063734904"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "r2_score(y_test,y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eea1ab2a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4284bb71",
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}