diff --git a/machine_learning/dbscan/dbscan.ipynb b/machine_learning/dbscan/dbscan.ipynb deleted file mode 100644 index 603a4cd40..000000000 --- a/machine_learning/dbscan/dbscan.ipynb +++ /dev/null @@ -1,376 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## DBSCAN\n", - "This implementation and notebook is inspired from the original DBSCAN algorithm and article as given in \n", - "[DBSCAN Wikipedia](https://en.wikipedia.org/wiki/DBSCAN).\n", - "\n", - "Stands for __Density-based spatial clustering of applications with noise__ . \n", - "\n", - "DBSCAN is clustering algorithm that tries to captures the intuition that if two points belong to the same cluster they should be close to one another. It does so by finding regions that are densely packed together, i.e, the points that have many close neighbours.\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### When to use ?\n", - "\n", - "1. You need a robust clustering algorithm.\n", - "2. You don't know how many clusters there are in the dataset\n", - "3. You find it difficult to guess the number of clusters there are just by eyeballing the dataset.\n", - "4. The clusters are of arbitrary shapes.\n", - "5. You want to detect outliers/noise." - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Why DBSCAN ? \n", - "\n", - "This algorithm is way better than other clustering algorithms such as [k-means](https://en.wikipedia.org/wiki/K-means_clustering) whose only job is to find circular blobs. It is smart enough to figure out the number of clusters in the dataset on its own, unlike k-means where you need to specify 'k'. It can also find clusters of arbitrary shapes, not just circular blobs. Its too robust to be affected by outliers (the noise points) and isn't fooled by them, unlike k-means where the entire centroid get pulled thanks to pesky outliers. Plus, you can fine-tune its parameters depending on what you are clustering.\n", - "\n", - "#### Have a look at these [neat animations](https://www.naftaliharris.com/blog/visualizing-dbscan-clustering/) of DBSCAN to see for yourself." - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [], - "source": [ - "import matplotlib.pyplot as plt\n", - "import numpy as np\n", - "\n", - "%matplotlib inline" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## First lets grab a dataset\n", - "We will take the moons dataset which is pretty good at showing the power of DBSCAN. \n", - "\n", - "Lets generate 200 random points in the shape of two moons" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "from sklearn.datasets import make_moons\n", - "\n", - "x, label = make_moons(n_samples=200, noise=0.1, random_state=19)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Visualize the dataset using matplotlib\n", - "You will observe that the points are in the shape of two crescent moons. \n", - "\n", - "The challenge here is to cluster the two moons. " - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": { - "scrolled": false - }, - "outputs": [ - { - "data": { - "text/plain": [ - "[]" - ] - }, - "execution_count": 3, - "metadata": {}, - "output_type": "execute_result" - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "plt.plot(x[:,0], x[:,1],'ro')" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Abstract of the Algorithm\n", - "The DBSCAN algorithm can be abstracted into the following steps:\n", - "\n", - "- Find the points in the $ε$ (eps) neighborhood of every point, and identify the core points with more than min_pts neighbors.\n", - "- Find the connected components of core points on the neighbor graph, ignoring all non-core points.\n", - "- Assign each non-core point to a nearby cluster if the cluster is an $ε$ (eps) neighbor, otherwise assign it to noise.\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Preparing the points\n", - "Initially we label all the points in the dataset as __undefined__ .\n", - "\n", - "__points__ is our database of all points in the dataset." - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "points = { (point[0],point[1]):{'label':'undefined'} for point in x }" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Helper functions" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "def euclidean_distance(q, p):\n", - " \"\"\"\n", - " Calculates the Euclidean distance\n", - " between points P and Q\n", - " \"\"\"\n", - " a = pow((q[0] - p[0]), 2)\n", - " b = pow((q[1] - p[1]), 2)\n", - " return pow((a + b), 0.5)" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "def find_neighbors(db, q, eps):\n", - " \"\"\"\n", - " Finds all points in the DB that\n", - " are within a distance of eps from Q\n", - " \"\"\"\n", - " return [p for p in db if euclidean_distance(q, p) <= eps]" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [], - "source": [ - "def plot_cluster(db, clusters):\n", - " \"\"\"\n", - " Extracts all the points in the DB and puts them together\n", - " as seperate clusters and finally plots them\n", - " \"\"\"\n", - " temp = []\n", - " noise = []\n", - " for i in clusters:\n", - " stack = []\n", - " for k, v in db.items():\n", - " if v[\"label\"] == i:\n", - " stack.append(k)\n", - " elif v[\"label\"] == \"noise\":\n", - " noise.append(k)\n", - " temp.append(stack)\n", - "\n", - " color = iter(plt.cm.rainbow(np.linspace(0, 1, len(clusters))))\n", - " for i in range(0, len(temp)):\n", - " c = next(color)\n", - " x = [l[0] for l in temp[i]]\n", - " y = [l[1] for l in temp[i]]\n", - " plt.plot(x, y, \"ro\", c=c)\n", - "\n", - " x = [l[0] for l in noise]\n", - " y = [l[1] for l in noise]\n", - " plt.plot(x, y, \"ro\", c=\"0\")" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "# Implementation of DBSCAN\n", - "\n", - "Initialize an empty list, clusters = $[ ]$ and cluster identifier, c = 0\n", - "\n", - "1. For each point p in our database/dict db :\n", - "\n", - " 1.1 Check if p is already labelled. If it's already labelled (means it already been associated to a cluster), continue to the next point,i.e, go to step 1\n", - " \n", - " 1.2. Find the list of neighbors of p , i.e, points that are within a distance of eps from p\n", - " \n", - " 1.3. If p does not have atleast min_pts neighbours, we label it as noise and go back to step 1\n", - " \n", - " 1.4. Initialize the cluster, by incrementing c by 1\n", - " \n", - " 1.5. Append the cluster identifier c to clusters\n", - " \n", - " 1.6. Label p with the cluster identifier c\n", - " \n", - " 1.7 Remove p from the list of neighbors (p will be detected as its own neighbor because it is within eps of itself)\n", - " \n", - " 1.8. Initialize the seed_set as a copy of neighbors\n", - " \n", - " 1.9. While the seed_set is not empty:\n", - " 1.9.1. Removing the 1st point from seed_set and initialise it as q\n", - " 1.9.2. If it's label is noise, label it with c\n", - " 1.9.3. If it's not unlabelled, go back to step 1.9\n", - " 1.9.4. Label q with c\n", - " 1.9.5. Find the neighbours of q \n", - " 1.9.6. If there are atleast min_pts neighbors, append them to the seed_set" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [], - "source": [ - "def dbscan(db,eps,min_pts):\n", - " '''\n", - " Implementation of the DBSCAN algorithm\n", - " '''\n", - " clusters = []\n", - " c = 0\n", - " for p in db:\n", - " if db[p][\"label\"] != \"undefined\":\n", - " continue\n", - " neighbors = find_neighbors(db, p, eps)\n", - " if len(neighbors) < min_pts:\n", - " db[p][\"label\"] = \"noise\"\n", - " continue\n", - " c += 1\n", - " clusters.append(c)\n", - " db[p][\"label\"] = c\n", - " neighbors.remove(p)\n", - " seed_set = neighbors.copy()\n", - " while seed_set != []:\n", - " q = seed_set.pop(0)\n", - " if db[q][\"label\"] == \"noise\":\n", - " db[q][\"label\"] = c\n", - " if db[q][\"label\"] != \"undefined\":\n", - " continue\n", - " db[q][\"label\"] = c\n", - " neighbors_n = find_neighbors(db, q, eps)\n", - " if len(neighbors_n) >= min_pts:\n", - " seed_set = seed_set + neighbors_n\n", - " return db, clusters\n" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Lets run it!" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAD6CAYAAACs/ECRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO2de7Ac1X3nPz9dSVhabIyuiEMAXeGEPJwQv7Q4cVIpHGwHUxtkJ94KzjWRHbvuIsW7SaV2N6RUlcRkVYtx1RpSsSAqwhpbd40fix05MSEY2+s/EhxECpCBxcgEYalILCQvsRbWQuK3f3QP6ju3n9Pv6e+nampmTp/uPtPTfX7n/F7H3B0hhBDDZUXbDRBCCNEuEgRCCDFwJAiEEGLgSBAIIcTAkSAQQoiBI0EghBADpxJBYGa3mNl3zOwbCdvnzexBM9tnZn9rZq+ObHsiLL/fzPZW0R4hhBD5sSriCMzsF4BjwMfd/aditr8ReMTdv2tmbwP+yN3fEG57Atjk7k/nPd/69et948aNpdsthBBD4r777nva3c8aL19ZxcHd/WtmtjFl+99Gvt4DnFvmfBs3bmTvXk0ehBCiCGZ2IK68DRvB+4A7It8d+Bszu8/MFlpojxBCDJpKZgR5MbM3EQiCn48U/7y7HzKzHwDuMrP/7e5fi9l3AVgA2LBhQyPtFUKIIdDYjMDMfhq4Gdjs7kdG5e5+KHz/DvA54KK4/d19l7tvcvdNZ521TMUlhBBiQhoRBGa2AbgduNLdvxkp/1dm9tLRZ+CtQKznkRBCiHqoRDVkZp8ELgbWm9lB4A+BVQDufhPwB8AssNPMAE64+ybgFcDnwrKVwP9w97+uok1CCCHyUZXX0Lsytr8feH9M+ePAq5fvIUT17FuEu7fDM0/CGRvgkh1w4XzbrRKifRo1FgvRFvsW4QsL8PyzwfdnDgTfQcJACKWYEIPg7u2nhMCI558NyoUYOhIEYhA882SxciGGhASBqJ19i3D9RvjgiuB932LzbTgjIfQkqVyIISFBIGplpJt/5gDgp3TzRYVBWWFyyQ5YtXZp2aq1QbkQQ0eCQNRKFbr5KoTJhfPwy7vgjDnAgvdf3iVDsRAgryFRM1Xo5tOESZGO/MJ5dfxCxKEZgaiVKnTzMvQKUS8SBKJWqtDNy9ArRL1IEIhaqUI3L0OvEPUiG4EA6k2/UFY3P9q3S+khlK5CTBOVLFXZNJs2bXKtUFYd4+kXIBhxd9Wrpu1OuG/XS4gRZnZfmPBzCVINiV6lX4hzJb393XDd+noC1eLiF/p0vYTIg1RDoldeOXGdMMBzR6pPIpeUqC7u/NDN6yVEHjQjEL3yyknrbKselSeN/G0mvn4Xr5cQeZAgEJV65dSdVyirs61yVJ50LD8pLyYxXUgQiMrSL0yaCqKI8IgTWlHWrCvW5jQSZ0pzSlchpgt5DYnSvOjFcyB++xlz8DtPJO9b1ANn3yJ87jfAX1i+bc0s/Oenc7Q1h8eRvIPEtFGr15CZ3WJm3zGz2IXnLeBPzGy/mT1oZq+LbNtiZo+Fry1VtEc0x5JZQAJp6ppJPHAunIek8ctzR3O2NceMRYnqxFCoymvoY8CfAh9P2P424ILw9QbgRuANZraOYKH7TYAD95nZHnf/bkXtEjWT5MUTJU2vn+axlDZ6P2NDvPBJO9ckyeuUqE4MgUpmBO7+NSBlLMZm4OMecA/wcjM7G/gl4C53Pxp2/ncBl1bRJtEMWcbZLCNqUse9Zl366H0SA3ef3GSFaJKmjMXnAN+OfD8YliWVi56QNgLPo0pJ6tAhXWU0idomsa3e3sppQnSB3gSUmdkCsACwYYMctrvCJTvKGVST8gjdfmV8/ejovajaJq6tLx73QPUBaUL0haYEwSHgvMj3c8OyQ8DFY+VfjTuAu+8CdkHgNVRHI0VxqkgIF9ehJ3khlQnaWtLWmGM//2wggG5/d/B9zSy87QYJBjH9NKUa2gP8Rug99DPAM+7+FHAn8FYzO9PMzgTeGpYNki4s8j4JF84H7qF/+ELwXkXHWVfq6VFbsYQKkSHGc0fgL36zP/+DEJNSyYzAzD5JMLJfb2YHCTyBVgG4+03AF4HLgP3As8B7w21HzeyPgXvDQ13j7mlG56klKa8NDHNEWnfq6SSvo3FOHi++JKYQfUMBZR3h+o0JqpCUYKy+0Xb66PG2pCWQW4IFsx0h+k5SQFlvjMXTzrS7NnZlxhMVRmvWwco1QRCarQhyCMWhZHJi2lGuoY7Qpwygk9CFHP7jkcXPHYETz8Gmq+AlL4/fZ2Z1MbtEX+08YthIEHSEJtbljeuk0jquKju1Lsx4koTR3hsDoTDOmlnYfEv+GcukSfeEaBuphjpCUeNoUX17nGrm8+8Fs8AgOiobqWugWlXOJCkhqqaI0JnENjNJCgshuoAEQYfIGyA1ib49rpN64fnl9aLqmio7taTAsyZz+Of1FILJZipdmPUIMQlSDfWQSfTtRTqjZ56svlPrQibPIkJnkpnKtNt5xPQiQdBDJumki3RGZ2xI79QmtR3UEXhWhAvnA71/FpPOVJqw8whRBxIEPWSSkWdcJ7ViVeAVE2XUcSV1ahdc1m+D6NtuWP67ZlaHAqLkTKULsx4hJkE2gh4yib49yRgdVxbtuMa39d0gWnfEstYvEH1EgqCHTNqZjXdScZ5HEEY5Pxm//u80GETVWQuxFKWYGChxKRZmVgdLQMZ5E0Ew61i5Jt7nfppSYdRNl1JtiGFR65rFIh9dijqNU/GcPJ4sBOBUfRlEJ2eSoLMu3TdiOpEgaIgmo07zdByTqnKeOzosg2jV0dhFXX//aluwRkJfjfOiH0g11BBNZReNU/nErRiW1J4shqQCyqs+G11fyL72H1zBkjUPXiQmw+m+xXCltpj6Q/ofRHVINdQyTRlZk0acn9uydBQZ5x46szpwKU1iaCqgvOqz0Yg+z2i/iOvv3duJFxr0yzgvuo8EQUM0FXWa1EH4yaUqhTif9823wNv/+6myNbPV+NdH6ZO+u45o7CJBZ1UFCAqRhdxHG6KpXDtp+XTG/f2T3ChHZSPvlucqWjOuK2sS5KVIbqJRx5yVWK+I62/a+Y8fC4SpvI5EFWhG0BBNRZ3GjTij5B3l1mHc7sKaBEXIqz7LisaeVNgn/ZcrVoYuvDIei4qoas3iS4EbgBngZne/dmz7R4A3hV/XAj/g7i8Pt50E9oXbnnT3y6toUxdpIpBpdPzPbYlfcSuvSqGOCOK+BaNVFY09HsSXNCuK2/eXdy0tO35seRxHnyK7RTcp7TVkZjPAN4G3AAcJFqJ/l7s/nFD/3wOvdfffDL8fc/fTi5yzj15DVZInICmv91ASRbxb8jKEdZmzSLoGa2aD1dKy/q86/hcxHOr0GroI2O/uj7v7ceA2YHNK/XcBn6zgvJ2nDsNoXpVNWVVUHcZtZedMnv08dySf2kyprkUdVCEIzgG+Hfl+MCxbhpnNAecDX44Uv8TM9prZPWb29qSTmNlCWG/v4cOHK2h2vdQVQFZEz14m7XMdnXbT2Tm76KFUtMMeFxwSpqIOmvYaugL4rPsS7fWcux8ys1cCXzazfe7+rfEd3X0XsAsC1VAzzS1GVGVjK5br6KvQ5TalZ68rS2dTCd+66qGU5D2WmMNpTHDUnT1VDJMqBMEh4LzI93PDsjiuAH4rWuDuh8L3x83sq8BrgWWCoOuMdzxxhloo32E3ufZvn7N0djVddpoBOq97cZ//F9FNqhAE9wIXmNn5BALgCuDXxyuZ2Y8DZwJ/Fyk7E3jW3b9vZuuBnwOuq6BNjRPX8cRRtsPuwtq/faDLHkppHblG+qINSgsCdz9hZh8A7iRwH73F3R8ys2uAve6+J6x6BXCbL3VT+gngz8zsBQJ7xbVJ3kZdJ08HU0WHLdVAPpqcOcWRN9V0FSmpldZalEVJ5yoiyS3QZsBfSHfzrOshHnIHUdZ9tolzV9HGNn+n6B9KOlczSd4c77g12Wsny7OojNdLk2mvu0ib6wfn9eyqItI6zzG66D0luoVyDVXEJCqbrIe4jNdLV42lTdKWUTWvfaIKO0bWMbrqPSW6hQRBhRTteNIe4rIdeZeNpdNOXvtEFXaMrGNoQCDyINVQS+xbDGIN4jhjQ/mOXBGo7ZE36KuK4LCsY2hAIPIgQVABRXWwo+l6XKzB6CEu25ErArU98tonqrBjZB1DAwKRB3kNlWQSr400D6N33BrsV5VHyVC9hkSAvIpElCSvIdkISlJUB7tvMXmxEX9h6aIxo+NP2pErAnX6yRL2ijsReZAgKEkRHexodJZEXF4ZPbAiibweQbqPRBayEZSk6GLkSWkopL8XRclyP1b8gMiLBEFJqlqMXDpbEUdaZ542Gx16QKEohgRBSYp4fiTOHuYkBF5kcRE2boQVK4L3xeH2XFmdedpstG/rQ4t2kSCogLwLwMilM4PFRVhYgAMHwD14X1gYrDDI6sxjF7c3uOAyxQ+IYkgQNEib+W96wfbt8OxYz/fss0H5AMnqzC+ch1dvASyy0eGBW2HNuvh9i8QPyMbQMC3OhuU11DDy4EjhyYSeL6l8ysmTguKxL7JsMfvnnw1WPFu1dvJ1K5SjqGFGs+HRQGg0GwaYr/+Ca0ZQExpN5WB8BLQuYRi7YZhhsHlUiUmzhueOlpt9ysbQMBmz4W3btrFy5UrMjJUrV7Jt27ZKTy9BUAPy2MhBnD3ge9+DVauW1lu7FnbsWL7vAAzKeVSJaQbjvLarOGRjaJiU2fC2bdu48cYbOXkyyElz8uRJbrzxxkqFgVJMlCQusvPu7QlT+rnggRQEHfiBmIs0Owunnx48GBs2BEIgOjUen0JDICx27WpkCt016kohkZQGRfdwTSQ9D3NzrDx48EUhEGVmZoYTJ04UOk2tC9OY2aVm9qiZ7Tezq2O2v8fMDpvZ/eHr/ZFtW8zssfC1pYr2NEXSyD8phYRGUxGSRkBHj8ITT8ALLwTv4527DMpLqMsBQR5uDbNjRzCgiRLOhuOEAJBYPgmljcVmNgN8FHgLcBC418z2xKw9/Cl3/8DYvuuAPwQ2EZi87gv3/W7ZdjVBkh7VZuIziyrjY4QNG+JHQFn2ABmUl1HWASEtX5FyFDXEaMCzffuy2fDMli2JM4KqqGJGcBGw390fd/fjwG3A5pz7/hJwl7sfDTv/u4BLK2hTIySN8P2kRlOZpIyAUkkSFAM1KJclzZ5VxMYg54gKmJ+PnQ0vLMQnKEsqn4QqBME5wLcj3w+GZeP8qpk9aGafNbPzCu7bSdIihRUvkMH8fKDXn5sDs+A9j55/UgEiYqnCO0jOEfWyc+dOtm7d+uIMYGZmhq1bt7Jz587KztGU19AXgI3u/tMEo/5bix7AzBbMbK+Z7T18+HDlDRxRZGSTpkcdjaZ+5RNB+e1XaqS0jIQRUOY+kwgQEUsV3kFyNa2fnTt3cuLECdydEydOVCoEoBpBcAg4L/L93LDsRdz9iLt/P/x6M/D6vPtGjrHL3Te5+6azzjqrgmYvp+jIJstQp5FSTUwiQEQsVaxgJlfT/lOFILgXuMDMzjez1cAVwJ5oBTM7O/L1cuCR8POdwFvN7EwzOxN4a1jWClWPbIoeb7B61oHEBXSRKryDtBxmTjp8n5f2GnL3E2b2AYIOfAa4xd0fMrNrgL3uvgf4D2Z2OXACOAq8J9z3qJn9MYEwAbjG3Y+WbdOkFB3ZZIXhT7JozeBC+lsOre87ZZcjrcI76JId8bEMco6I0PH7XAFlEYoG0WTVL3K8wQbwpATS8MQTTbemV3RpPWKtj51BR+7zWgPKpoWi0+SsEX8Vi9ZMvZ5VcQET0yUjbZl0FoOg4/e5BEGEolGaWal+K1m0Ztr1rIoLmJjBDh76SMfvc6WhHiNvlOa+RTj+veXlK1YtHfHnPd5g9aw7dsTnDlJcQCZ50lSLjtDx+1wzggySPHnu3g4njy+vf9rLJpsWD3bRGsUFTIzyAfWIjt/nMhankGaMu/1Kli0IAoAFelIhmkBGWlEEGYsnIM0YN1idvugUZY20k8SuKN6le3EAZZEgSCHNGKdpueg7k0S+DzZaPm4hpYWFqREGEgQpZK3+NEidfleZ4tFaXUziftoll9VGmfJ1MCQIUsga9ct3ugYm6dCnfLRWBXHqnEncTwfrslo2DqDjAxUJghQ06m+YSTv0KR+tleWvtgXODVF1zu1XLh/kjEizcw3WNjZpHMDiIqxfD+9+d6cHKhIEGWjU3yCTdugdj9psk32LsPcmlnu4OTz/f4O4lyhZdq64WTLA8WNTbifIsw7G+Kh/27agwz9yZPnxOjZQkSAQ3SFvhz7+wK1LCPHuSNRmm9y9nXg355DTXlZsxjuaJa+ZXVr+3JEpNxpnxQHEzWZvumn5wCZKhwYqiiMQzbG4GLsm64skJeY6/XSYnQ32W7cO/uVf4PnnT21ftSp4OI9HIvzWru1UwE5bfHAFqYJg0riXwSZJTCLp3k2jhcSKiiOYkMH6TFdNHv3/jh2wevXyfY8dO7XfkSNLhQAE31/60s5GbbZJlu5+Ut3+YI3GSRQd3XcovQRIEKQyWJ/pOsij/5+fDzr0STh6VKuWxZCk04dycS+DNRonkaSGNFteNjvbuYGKBEEKg/WZroO8+v+jE65LJHtALEs83wAL1j9/0R4Ak814FVA5RpIx+aqrls5Ud++Gp5/ulBAAZR9NRdPfCtmwIV6HOt6BJ9VLo2PT7K6RlAG3zKp4VaxsNlWMOvY0G1iHkbE4BRnEKmR8qT6IN+jG1Rtn9epAhXT0aO8euC6h+3t41GosNrNLzexRM9tvZlfHbP9dM3vYzB40s7vNbC6y7aSZ3R++9ozv2yaa/lZI3jS8cfW2bl36/ZZbgum17AGl0Ix3QjoeJTwJpWcEZjYDfBN4C3CQYCH6d7n7w5E6bwK+7u7PmtlW4GJ3/7Vw2zF3P73IOZt0H1WaXzGtaEYwAXlnth2lzhnBRcB+d3/c3Y8DtwGboxXc/SvuPrpy9wDnVnDeRlBksZhWNOOdgKLR7z2ZPVQhCM4Bvh35fjAsS+J9wB2R7y8xs71mdo+Zvb2C9nQSxSOIrqFcWhNQJJ1Jj5IhNuo+ambvBjYBH44Uz4VTlV8HrjezH07YdyEUGHsPHz5cWxvr6LAVj1AxPRll9QHNeAuwuBjcc3HEuS/3KBliFYLgEHBe5Pu5YdkSzOzNwHbgcnf//qjc3Q+F748DXwVeG3cSd9/l7pvcfdNZZ51VQbOXE9dhf/69cN36coJB8QgVkmeUJUEhqmZ03508uXzbyH15/L5LcoPuUI6hEVUIgnuBC8zsfDNbDVwBLPH+MbPXAn9GIAS+Eyk/08xOCz+vB34OeJiWiOuwX3g+SKhVZiQv74wKyRpl9Wg63hZSU05A3H0HMDMTGIph+X0XF1UMnQx+LC0I3P0E8AHgTuAR4NPu/pCZXWNml4fVPgycDnxmzE30J4C9ZvYA8BXg2qi3UdPk6ZjjRvJZD5bC8SskS0fbo+l4G0hNOSFJ990LLwTeQnH3XZxH5urVnQx+rMRG4O5fdPcfdfcfdvcdYdkfuPue8POb3f0V7v6a8HV5WP637n6hu786fP/zKtozKXk75qjAyPNgyTujQrIWCEl6YA8c0KwAqSknZtL7bpyOBvAq11CEtARdUaICI8+DJe+MCslaICRt2i0VkdSUk7C4GGTAHSfvfRfl+ec7OTuVIIgw3mGvmYWZmKzI0dWY8j5Y8s6oiKwI5ThBMUIqIqkpizKyOY2vMjaeQTTtvhung8Zi5RpKYd8i3PHbobF4jFVrA6Fx93ZFZ3aOxcVgjdg4zAK97gCIi4qHpYnm4NS9rMFJDEneP3GLyowvvHTsWPwylS0sSDNCC9MUZKT7jxMCcEr9I/1/B5mfDx62ODrosVEHSbYrkJqyEEUCyObnl66JccMN2escdwQJggTidP/jjGYCerA6SJ7FxqeYJNvV57bA7VcG33/lE1JTZpJlJE4jb6LFDqD1CBLIazz7wkLQ8UsN1DF6nh++LEn3r4fxUEXWHhg0O3bEJ5nLO6CYn+/FPacZQQJ5jWdyvesw41P1HjyQVZHn/tW9m4MejerLIEGQQF5XUpDrnegeee9f3bs5GMCAQoIggTjf/zWz8XXXrGu0aUJkMrp/R2sUJyG3UQESBKmM+/6/7QZYsWp5vePfU4i+6B4XzoOneMrKu02MkCAowIXzcNrLlpefPC5dq+gmSSN+m5F3mziFBEFBnjsaXy5dq+giSXEu77hVQkCcQoKgIArR7yhagyAW5blqmW3bYOXKwONo5crgewdRiomCjCI2FaLfIeIWFIcgH8wNN0yll4foAdu2wY03Li/fuhV27my+PSSnmJAgmIC4HC4SAi2SthrU2rVT6fddB7qvK2blyvgVzWZm4MSJ5tuDBIFu8mlmxYr0PO8tJvnqC5rp1kDSCmXQ2roEg046p1WZppysvC8dTPvbNbRgTQ3MJARxJJW3yCAEgW7yKSDNGJyVC34gGUfLoAVramBhoVh5i1QiCMzsUjN71Mz2m9nVMdtPM7NPhdu/bmYbI9t+Pyx/1Mx+qYr2jKObvOdkLUg/ygczGxP6PaCMo2WQN1wN7NwZGIZHM4CZmVYNxWmUFgRmNgN8FHgb8CrgXWb2qrFq7wO+6+4/AnwE+FC476uAK4CfBC4FdobHqxTd5D0nz4L08/Pw9NOwe/fUJwirA62rURM7dwaGYffgvYNCAKqZEVwE7Hf3x939OHAbsHmszmbg1vDzZ4FLzMzC8tvc/fvu/o/A/vB4laKbvOeUWRxEQiAXReIN9i3C9RvhgyuCd9naIvQ0nqWK9QjOAb4d+X4QeENSHXc/YWbPALNh+T1j+54TdxIzWwAWADYU1PmObmZ5DfWUDRvi3UOl+6+UC+ezn4lx7yKtaxBhPJ5lpMKEzg9IemMsdvdd7r7J3TedddZZhffX4vE9ZtLVxno6OquSqkfvcrwg+b7Ko8LsKFXMCA4B50W+nxuWxdU5aGYrgTOAIzn3bQ3FHnSESVYb6/HorCrqGL0P3vEi7b4qosLsGFXMCO4FLjCz881sNYHxd89YnT3AlvDzO4EvexDJtge4IvQqOh+4APj7CtpUGsUedIyo7n/HjkAopI30ezw6q4o6Ru+DcbyYZNRfZn3jliktCNz9BPAB4E7gEeDT7v6QmV1jZpeH1f4cmDWz/cDvAleH+z4EfBp4GPhr4LfcPSYmu3k0Be4oWa6kI3o8OquKOkbvg3C8SLvH0u6rSVWYHWAwKSaK8sEVQNylscDOIFoiKa/QeBqJvPWmmOs3hjPaMc6YC+xkkzLVKtPFRdiyJT5H0Nxc8J52Xy0uFlNhNsygU0xMwmCmwH0j70i/x6Ozqqhr9D61jhejmUCcEIB8o/6q3JcbdnSQIEhgEFPgPpJXDzuKNh5wcJnWIihInP4/yoYN2fdVFR14XvVnlbh7716vf/3rvQke3O3+kTn3P7Lg/cHdjZxWpLF7t/vate7BIxK81q4NytP2mZtzNwve0+qK4WK29L6Kvkbb0u6fSe7NOObm4tswN1fyB7oDez2mT229U5/k1ZQgEB1l1LGD+8xM+gNa1cMppp+kDnj8lXT/VNWBJwkks9I/MUkQSDUk+sf8/Cld7UifmzR9LuJGqgC0YTL63w8cWL6GQNyaAkn3T1Weai24oUoQiH6S1sFHO/SklcvGH8429LKifaL/OwT//ajzn5tLXkAmrnMv2oEnDTzacHSImyZ0/SXV0MCI0/Gn6XPHVUF5pus16mVFh8n634vcF0XUkFl1a7JrIRuB6CVJD8zsbPwDOrIZFNXx1qiXFR0m638vamPK24G3NPBIEgRSDYluk6QCgvjpc5IPOKS7kfY4PYAoQdb/XtQNOW8cQcci3yUIRLdJejCOHo1/QEfRn+PMzaU/nApAGyZ5/vc61rjo2MBDgkB0m7QHJu4BnbRDn58PUgtElxXcsmVQAWiDJM+Ivw5vsq4NPOL0RV1/yUYwIJoKIFO8gYijzvuihUBHZCwWvaWJB0ZeQ92iK9HgU3ZfJAkCqYZE92liHeI0450CzZqlSzEdHTPq1oUEgRCQbItYt647ndJQaGNRoSRh3zGjbl1IEAgBycY7GPxKZ43T9Cg8bQbSNaNuTUgQCAHJ3iNHj8bX76tqoG41VxXHb3oUnjYDGUo68zjDQd4XsA64C3gsfD8zps5rgL8DHgIeBH4tsu1jwD8C94ev1+Q5r4zFIpOqjI3TZCys2zOqquM37cE1oKhy6vAaAq4Drg4/Xw18KKbOjwIXhJ9/CHgKeLmfEgTvLHpeCQKRSpUdyTS5ldYt1Ko8fpNeQ9Mk7DOoSxA8Cpwdfj4beDTHPg9EBIMEgaieqh/srrgylqXukW9fR9Z15RPqIEmCoKyN4BXu/lT4+Z+AV6RVNrOLgNXAtyLFO8zsQTP7iJmdVrI9QlRvbGzCfbUJ6ta999XDpogdoEuurRWSKQjM7Etm9o2Y1+ZovVDaeMpxzgY+AbzX3V8Ii38f+HHgXxPYG34vZf8FM9trZnsPHz6c/cvEcGmyQ+pTjEHdHjB99rDJK+zbcG1tgrhpQt4XOVVDwMuAfyBFDQRcDPxlnvNKNSRSaUqv30f7Qd1qjR6rTXLRV/VXCDXZCD7MUmPxdTF1VgN3A78Ts20kRAy4Hrg2z3klCEQmbaalGNkjpq0TbJKuCpSeG5brEgSzYSf/GPAlYF1Yvgm4Ofz8buB5TrmIvugmCnwZ2Ad8A9gNnJ7nvBIEohOkrZLWh9lBV+nyTKvLbctBLYKgrZcEgchNnSPLtBlBnSPFro6Wq6Lro+4eX38JAjE82gigqlt33PMRaS6a0sP3uEOflCRBYMG2frFp0ybfu3dv280QXWfjxsC9b5y5ucAzpAoWF4MFbJKWyKzyXNDMb2qbpv63hYWlHkBr105n+ogIZnafu28aL1euITG9NJG8bH4+cDlMomrXySGkRa7CDTXLrXda3UAnRIJATC9NxRMkHW92drLRZVon1tegrSKUTfSWJ+hrCAK1CHH6oq6/ZCMQuehaPEEenXTWsabFRtCGET9qbO66QbomkLFYDJKmDIJZ52kH/WcAAArwSURBVMnbgefpoNo2cpY9fx3CLNqmPIb7aRGoBZEgEKJN8o5A0zqyLni4VNGB1pEUMMt7K+744wJt69ZuXOMaSRIE8hoSoglWrAi6o3HMlhqbkzxmzJbu35aHSxUePXmvRdk2Rcm6XgPxIpLXkBDjNJkwLsmYu2LF0vPGecyMCwFoz8MlyZia1RFHqdrgnWbgzWtsHrgXkQSBGCZ5PEuqFBRxHTwE8QfR88Z5zCTN2tvwcEnqrM3Sr0/0Wh47BqtWLd1eJktpUpvm5pZmE037P4fuRRSnL+r6SzYCUZosPXUZXXiSMXX3bveZmeL68S55uOzenWzHSGpP3LVcvdp9drYafXye/yqrTpeucY0gY7EQEbLSGEzaMWR1OJOkT2jbw2VcsKV55cQJwSY62TRPpjwCuO1r3BASBEJEyeqcJs13k3XcMgKmiEdLVS6mcR1k0rWZnY3vTNMER91keRSNu5QO1Guo9U59kpcEgShNXaqCLAHSxMizynMkXYfx37l2bSAI8tStU+0y3pkntanONnQYCQIhxslSJ0zSmXYhIKwqVczu3dmdaPQ3ZK3PEH2tWtVMhHfaawpVP1lIEAhRlEk67C7omqtQxWR1qnFCJc/6DFE1UtUUOf/MzOCEgHuyIJD7qBBJ5F3QfHyfMgnTyrK4GJw3jiJ++nF+9SOSXD2TXGTjOHo0f1vyktfVc+1auPXWqQoUK4sEgRBVM4kASaNIPMP27cGYdxyzYn76aZ3qmjVw5ZXL2xInBGdn449RR7bUtCywbQnmvhA3Tcj7AtYBdxGsWXwXcGZCvZOcWq94T6T8fODrwH7gU8DqPOeVakgMhqKqpjQ9/eh4edRdRYzEaSqWumMIss41QDtAGtS0eP11wNXh56uBDyXUO5ZQ/mngivDzTcDWPOeVIBC9ooxxuKjhN61+kY6yqNto3t8/OxsYiuvqrAfgAlqGugTBo8DZ4eezgUcT6i0TBIABTwMrw+8/C9yZ57wSBKI3lB2lFo1nSDtfUaGSN5AM8v+egUTwdpUkQVAq+6iZ/R93f3n42YDvjr6P1TsRqoVOANe6++fNbD1wj7v/SFjnPOAOd/+prPMq+6joDWWzdU6y/+JiYCt48slAb75jR6ATL5v1My3LZ97fU3XmUVGIibOPmtmXzOwbMa/N0XqhtEmSKnPhyX8duN7MfniCH7BgZnvNbO/hw4eL7i5EO5RNZlZ0/d4kIQDls36mGZvz/p4hLLXZR+KmCXlf5FQNje3zMeCdSDUkhkAVqpBxHXuSobWJZS6TInXz/p6mIqtlJ4iFmmwEH2apsfi6mDpnAqeFn9cTeBi9Kvz+GZYai7flOa8EgegNVXZ8VaTF6MIyk1F7xSgZXFUdtjyHUqlLEMwCd4ed+5eAdWH5JuDm8PMbgX3AA+H7+yL7vxL4ewL30c+MBEbWS4JA9IoinW9a3boS5dX5e9KOUUeHLWN0KkmCQEtVCtEVtm2Dm25aakyNLpeYZWitYhnJpqirrTJGp6KlKoXoMouLy4UALF0uMcvQWtSw3CZ1rQgmY/RESBAI0QWSUkPAqc4xq6NvO89REerqsPskDDuEBIEQXSBtJDzqHPN09FXnOaqLujrsPgnDDiFBIEQXSFsUPto59qWjjxKXNK/ODruP16hlVrbdACEEQWe/sLA09bMZXHVVvzuyxcWlv+vAgeA7BL+rz79titCMQIguEDdC/sQnYOfOtltWjrh1DaIGcNEJ5D4qhKgPuXN2CrmPCiGaR+6cvUCCQAhRH3Ln7AUSBEKI+pA7Zy+QIBCiCoqsKzw05M7ZeeQ+KkRZslwkheg4mhEIURa5SIqeI0EgRFnqSqAmRENIEAhRFrlIip4jQSBEWeQiKXqOBIEQZZGLpOg58hoSogqUQE30mFIzAjNbZ2Z3mdlj4fuZMXXeZGb3R17/z8zeHm77mJn9Y2Tba8q0RwghRHHKqoauBu529wsIFrG/eryCu3/F3V/j7q8BfhF4FvibSJX/NNru7veXbI8QQoiClBUEm4Fbw8+3Am/PqP9O4A53fzajnhBCiIYoKwhe4e5PhZ//CXhFRv0rgE+Ole0wswfN7CNmdlrJ9gghhChIprHYzL4E/GDMpiVhk+7uZpa4uIGZnQ1cCNwZKf59AgGyGtgF/B5wTcL+C8ACwAb5ZwshRGWUWpjGzB4FLnb3p8KO/qvu/mMJdX8b+El3X0jYfjHwH9393+Q472HgwMQNL8d64OmWzl2GvrYb+tt2tbt5+tr2pto95+5njReWdR/dA2wBrg3f/yKl7rsIZgAvYmZnh0LECOwL38hz0rgf0hRmtjduhZ+u09d2Q3/brnY3T1/b3na7y9oIrgXeYmaPAW8Ov2Nmm8zs5lElM9sInAf8r7H9F81sH7CPQCL+l5LtEUIIUZBSMwJ3PwJcElO+F3h/5PsTwDkx9X6xzPmFEEKURykmirOr7QZMSF/bDf1tu9rdPH1te6vtLmUsFkII0X80IxBCiIEjQZCBmf1bM3vIzF4ws0SrvpldamaPmtl+M1uWaqNp8uSBCuudjOR62tN0O8faknoNzew0M/tUuP3roRNC6+Ro93vM7HDkOr8/7jhNY2a3mNl3zCzWW88C/iT8XQ+a2euabmMcOdp9sZk9E7nef9B0G+Mws/PM7Ctm9nDYp/x2TJ12rrm765XyAn4C+DHgq8CmhDozwLeAVxIExz0AvKrldl8HXB1+vhr4UEK9Y21f47zXENgG3BR+vgL4VE/a/R7gT9tua0zbfwF4HfCNhO2XAXcABvwM8PW225yz3RcDf9l2O2PadTbwuvDzS4FvxtwrrVxzzQgycPdH3P3RjGoXAfvd/XF3Pw7cRpCHqU2K5oFqmzzXMPqbPgtcEsagtEkX//tcuPvXgKMpVTYDH/eAe4CXh4GjrZKj3Z3E3Z9y938IP38PeITl3pStXHMJgmo4B/h25PtBYtxlGyZvHqiXmNleM7tnlB68JfJcwxfruPsJ4BlgtpHWJZP3v//VcKr/WTM7r5mmlaaL93VeftbMHjCzO8zsJ9tuzDihWvO1wNfHNrVyzbUwDen5lNw9LVq6VSrKAzXn7ofM7JXAl81sn7t/q+q2DpwvAJ909++b2b8jmNUohqY+/oHgvj5mZpcBnwcuaLlNL2JmpwP/E/gdd/+XttsDEgQAuPubSx7iEEHk9Ihzw7JaSWu3mf1zJIXH2cB3Eo5xKHx/3My+SjBKaUMQ5LmGozoHzWwlcAZwpJnmJZLZbg8CL0fcTGC/6QOt3NdliXau7v5FM9tpZuvdvfUcRGa2ikAILLr77TFVWrnmUg1Vw73ABWZ2vpmtJjBktuqBw6k8UJCQB8rMzhyl/jaz9cDPAQ831sKl5LmG0d/0TuDLHlrYWiSz3WM63ssJdMN9YA/wG6Eny88Az0TUjZ3FzH5wZDsys4sI+rm2BwyEbfpz4BF3/28J1dq55m1b0rv+At5BoKf7PvDPwJ1h+Q8BX4zUu4zAC+BbBCqltts9S7Bq3GPAl4B1Yfkm4Obw8xsJ8jw9EL6/r+U2L7uGBGnJLw8/vwT4DLAf+HvglW1f55zt/q/AQ+F1/grw4223OWzXJ4GngOfDe/x9wFXAVeF2Az4a/q59JHjNdbDdH4hc73uAN7bd5rBdPw848CBwf/i6rAvXXJHFQggxcKQaEkKIgSNBIIQQA0eCQAghBo4EgRBCDBwJAiGEGDgSBEIIMXAkCIQQYuBIEAghxMD5/1SENh4utwCVAAAAAElFTkSuQmCC\n", - "text/plain": [ - "
" - ] - }, - "metadata": { - "needs_background": "light" - }, - "output_type": "display_data" - } - ], - "source": [ - "eps = 0.25\n", - "min_pts = 12\n", - "\n", - "db,clusters = dbscan(points,eps,min_pts)\n", - "\n", - "plot_cluster(db,clusters)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "I encourage you to try with different datasets and playing with the values of eps and min_pts.\n", - "\n", - "Also, try kmeans on this dataset and see how it compares to dbscan. " - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "I hope by now you are convinced about about how cool dbscan is. But it has its pitfalls.\n", - "### When NOT to use ?\n", - "\n", - "1. You have a high dimentional dataset. Euclidean distance will fail thanks to '[curse of dimentionality](https://en.wikipedia.org/wiki/Curse_of_dimensionality#Distance_functions)'.\n", - "2. We have used a dict to store the points. So we can't do anything about the order in which the points will be processed. So it's not entirely deterministic.\n", - "3. Won't work well if there are large differences in density. Finding the min_pts and $ε$ combination will be difficult.\n", - "4. Choosing the $ε$ without understanding the data and its scale, might result is poor clustering performance." - ] - }, - { - "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": "3.7.0" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -} diff --git a/machine_learning/dbscan/dbscan.py b/machine_learning/dbscan/dbscan.py deleted file mode 100644 index 04fb5f018..000000000 --- a/machine_learning/dbscan/dbscan.py +++ /dev/null @@ -1,271 +0,0 @@ -import matplotlib.pyplot as plt -import numpy as np -from sklearn.datasets import make_moons -import warnings - - -def euclidean_distance(q, p): - """ - Calculates the Euclidean distance - between points q and p - - Distance can only be calculated between numeric values - >>> euclidean_distance([1,'a'],[1,2]) - Traceback (most recent call last): - ... - ValueError: Non-numeric input detected - - The dimentions of both the points must be the same - >>> euclidean_distance([1,1,1],[1,2]) - Traceback (most recent call last): - ... - ValueError: expected dimensions to be 2-d, instead got p:3 and q:2 - - Supports only two dimentional points - >>> euclidean_distance([1,1,1],[1,2]) - Traceback (most recent call last): - ... - ValueError: expected dimensions to be 2-d, instead got p:3 and q:2 - - Input should be in the format [x,y] or (x,y) - >>> euclidean_distance(1,2) - Traceback (most recent call last): - ... - TypeError: inputs must be iterable, either list [x,y] or tuple (x,y) - """ - if not hasattr(q, "__iter__") or not hasattr(p, "__iter__"): - raise TypeError("inputs must be iterable, either list [x,y] or tuple (x,y)") - - if isinstance(q, str) or isinstance(p, str): - raise TypeError("inputs cannot be str") - - if len(q) != 2 or len(p) != 2: - raise ValueError( - "expected dimensions to be 2-d, instead got p:{} and q:{}".format( - len(q), len(p) - ) - ) - - for num in q + p: - try: - num = int(num) - except: - raise ValueError("Non-numeric input detected") - - a = pow((q[0] - p[0]), 2) - b = pow((q[1] - p[1]), 2) - return pow((a + b), 0.5) - - -def find_neighbors(db, q, eps): - """ - Finds all points in the db that - are within a distance of eps from Q - - eps value should be a number - >>> find_neighbors({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}}, (2,5),'a') - Traceback (most recent call last): - ... - ValueError: eps should be either int or float - - Q must be a 2-d point as list or tuple - >>> find_neighbors({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}}, 2, 0.5) - Traceback (most recent call last): - ... - TypeError: Q must a 2-dimentional point in the format (x,y) or [x,y] - - Points must be in correct format - >>> find_neighbors([], (2,2) ,0.4) - Traceback (most recent call last): - ... - TypeError: db must be a dict of points in the format {(x,y):{'label':'boolean/undefined'}} - """ - - if not isinstance(eps, (int, float)): - raise ValueError("eps should be either int or float") - - if not hasattr(q, "__iter__"): - raise TypeError("Q must a 2-dimentional point in the format (x,y) or [x,y]") - - if not isinstance(db, dict): - raise TypeError( - "db must be a dict of points in the format {(x,y):{'label':'boolean/undefined'}}" - ) - - return [p for p in db if euclidean_distance(q, p) <= eps] - - -def plot_cluster(db, clusters, ax): - """ - Extracts all the points in the db and puts them together - as seperate clusters and finally plots them - - db cannot be empty - >>> fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(7, 5)) - >>> plot_cluster({},[1,2], axes[1] ) - Traceback (most recent call last): - ... - Exception: db is empty. No points to cluster - - clusters cannot be empty - >>> fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(7, 5)) - >>> plot_cluster({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}},[],axes[1] ) - Traceback (most recent call last): - ... - Exception: nothing to cluster. Empty clusters - - clusters cannot be empty - >>> fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(7, 5)) - >>> plot_cluster({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}},[],axes[1] ) - Traceback (most recent call last): - ... - Exception: nothing to cluster. Empty clusters - - ax must be a plotable - >>> plot_cluster({ (1,2):{'label':'1'}, (2,3):{'label':'2'}},[1,2], [] ) - Traceback (most recent call last): - ... - TypeError: ax must be an slot in a matplotlib figure - """ - if len(db) == 0: - raise Exception("db is empty. No points to cluster") - - if len(clusters) == 0: - raise Exception("nothing to cluster. Empty clusters") - - if not hasattr(ax, "plot"): - raise TypeError("ax must be an slot in a matplotlib figure") - - temp = [] - noise = [] - for i in clusters: - stack = [] - for k, v in db.items(): - if v["label"] == i: - stack.append(k) - elif v["label"] == "noise": - noise.append(k) - temp.append(stack) - - color = iter(plt.cm.rainbow(np.linspace(0, 1, len(clusters)))) - for i in range(0, len(temp)): - c = next(color) - x = [l[0] for l in temp[i]] - y = [l[1] for l in temp[i]] - ax.plot(x, y, "ro", c=c) - - x = [l[0] for l in noise] - y = [l[1] for l in noise] - ax.plot(x, y, "ro", c="0") - - -def dbscan(db, eps, min_pts): - """ - Implementation of the DBSCAN algorithm - - Points must be in correct format - >>> dbscan([], (2,2) ,0.4) - Traceback (most recent call last): - ... - TypeError: db must be a dict of points in the format {(x,y):{'label':'boolean/undefined'}} - - eps value should be a number - >>> dbscan({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}},'a',20 ) - Traceback (most recent call last): - ... - ValueError: eps should be either int or float - - min_pts value should be an integer - >>> dbscan({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}},0.4,20.0 ) - Traceback (most recent call last): - ... - ValueError: min_pts should be int - - db cannot be empty - >>> dbscan({},0.4,20.0 ) - Traceback (most recent call last): - ... - Exception: db is empty, nothing to cluster - - min_pts cannot be negative - >>> dbscan({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}}, 0.4, -20) - Traceback (most recent call last): - ... - ValueError: min_pts or eps cannot be negative - - eps cannot be negative - >>> dbscan({ (1,2):{'label':'undefined'}, (2,3):{'label':'undefined'}},-0.4, 20) - Traceback (most recent call last): - ... - ValueError: min_pts or eps cannot be negative - - """ - if not isinstance(db, dict): - raise TypeError( - "db must be a dict of points in the format {(x,y):{'label':'boolean/undefined'}}" - ) - - if len(db) == 0: - raise Exception("db is empty, nothing to cluster") - - if not isinstance(eps, (int, float)): - raise ValueError("eps should be either int or float") - - if not isinstance(min_pts, int): - raise ValueError("min_pts should be int") - - if min_pts < 0 or eps < 0: - raise ValueError("min_pts or eps cannot be negative") - - if min_pts == 0: - warnings.warn("min_pts is 0. Are you sure you want this ?") - - if eps == 0: - warnings.warn("eps is 0. Are you sure you want this ?") - - clusters = [] - c = 0 - for p in db: - if db[p]["label"] != "undefined": - continue - neighbors = find_neighbors(db, p, eps) - if len(neighbors) < min_pts: - db[p]["label"] = "noise" - continue - c += 1 - clusters.append(c) - db[p]["label"] = c - neighbors.remove(p) - seed_set = neighbors.copy() - while seed_set != []: - q = seed_set.pop(0) - if db[q]["label"] == "noise": - db[q]["label"] = c - if db[q]["label"] != "undefined": - continue - db[q]["label"] = c - neighbors_n = find_neighbors(db, q, eps) - if len(neighbors_n) >= min_pts: - seed_set = seed_set + neighbors_n - return db, clusters - - -if __name__ == "__main__": - - fig, axes = plt.subplots(nrows=1, ncols=2, figsize=(7, 5)) - - x, label = make_moons(n_samples=200, noise=0.1, random_state=19) - - axes[0].plot(x[:, 0], x[:, 1], "ro") - - points = {(point[0], point[1]): {"label": "undefined"} for point in x} - - eps = 0.25 - - min_pts = 12 - - db, clusters = dbscan(points, eps, min_pts) - - plot_cluster(db, clusters, axes[1]) - - plt.show() diff --git a/machine_learning/naive_bayes.ipynb b/machine_learning/naive_bayes.ipynb deleted file mode 100644 index 5a427c5cb..000000000 --- a/machine_learning/naive_bayes.ipynb +++ /dev/null @@ -1,1659 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 1, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "from sklearn import datasets\n", - "import pandas as pd" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "iris = datasets.load_iris()" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "df = pd.DataFrame(iris.data)\n", - "df.columns = [\"sl\", \"sw\", 'pl', 'pw']" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "def abc(k, *val):\n", - " if k < val[0]:\n", - " return 0\n", - " else:\n", - " return 1" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "0 1\n", - "1 0\n", - "2 0\n", - "3 0\n", - "4 1\n", - "5 1\n", - "6 0\n", - "7 1\n", - "8 0\n", - "9 0\n", - "10 1\n", - "11 0\n", - "12 0\n", - "13 0\n", - "14 1\n", - "15 1\n", - "16 1\n", - "17 1\n", - "18 1\n", - "19 1\n", - "20 1\n", - "21 1\n", - "22 0\n", - "23 1\n", - "24 0\n", - "25 1\n", - "26 1\n", - "27 1\n", - "28 1\n", - "29 0\n", - " ..\n", - "120 1\n", - "121 1\n", - "122 1\n", - "123 1\n", - "124 1\n", - "125 1\n", - "126 1\n", - "127 1\n", - "128 1\n", - "129 1\n", - "130 1\n", - "131 1\n", - "132 1\n", - "133 1\n", - "134 1\n", - "135 1\n", - "136 1\n", - "137 1\n", - "138 1\n", - "139 1\n", - "140 1\n", - "141 1\n", - "142 1\n", - "143 1\n", - "144 1\n", - "145 1\n", - "146 1\n", - "147 1\n", - "148 1\n", - "149 1\n", - "Name: sl, dtype: int64" - ] - }, - "execution_count": 5, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df.sl.apply(abc, args=(5,))" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "def label(val, *boundaries):\n", - " if (val < boundaries[0]):\n", - " return 'a'\n", - " elif (val < boundaries[1]):\n", - " return 'b'\n", - " elif (val < boundaries[2]):\n", - " return 'c'\n", - " else:\n", - " return 'd'\n", - "\n", - "def toLabel(df, old_feature_name):\n", - " second = df[old_feature_name].mean()\n", - " minimum = df[old_feature_name].min()\n", - " first = (minimum + second)/2\n", - " maximum = df[old_feature_name].max()\n", - " third = (maximum + second)/2\n", - " return df[old_feature_name].apply(label, args= (first, second, third))" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
slswplpwsl_labeledsw_labeledpl_labeledpw_labeled
05.13.51.40.2bcaa
14.93.01.40.2abaa
24.73.21.30.2acaa
34.63.11.50.2acaa
45.03.61.40.2acaa
55.43.91.70.4bdaa
64.63.41.40.3acaa
75.03.41.50.2acaa
84.42.91.40.2abaa
94.93.11.50.1acaa
105.43.71.50.2bcaa
114.83.41.60.2acaa
124.83.01.40.1abaa
134.33.01.10.1abaa
145.84.01.20.2bdaa
155.74.41.50.4bdaa
165.43.91.30.4bdaa
175.13.51.40.3bcaa
185.73.81.70.3bdaa
195.13.81.50.3bdaa
205.43.41.70.2bcaa
215.13.71.50.4bcaa
224.63.61.00.2acaa
235.13.31.70.5bcaa
244.83.41.90.2acaa
255.03.01.60.2abaa
265.03.41.60.4acaa
275.23.51.50.2bcaa
285.23.41.40.2bcaa
294.73.21.60.2acaa
...........................
1206.93.25.72.3dcdd
1215.62.84.92.0bbcd
1227.72.86.72.0dbdd
1236.32.74.91.8cbcc
1246.73.35.72.1ccdd
1257.23.26.01.8dcdc
1266.22.84.81.8cbcc
1276.13.04.91.8cbcc
1286.42.85.62.1cbdd
1297.23.05.81.6dbdc
1307.42.86.11.9dbdd
1317.93.86.42.0dddd
1326.42.85.62.2cbdd
1336.32.85.11.5cbcc
1346.12.65.61.4cbdc
1357.73.06.12.3dbdd
1366.33.45.62.4ccdd
1376.43.15.51.8ccdc
1386.03.04.81.8cbcc
1396.93.15.42.1dcdd
1406.73.15.62.4ccdd
1416.93.15.12.3dccd
1425.82.75.11.9bbcd
1436.83.25.92.3ccdd
1446.73.35.72.5ccdd
1456.73.05.22.3cbcd
1466.32.55.01.9cacd
1476.53.05.22.0cbcd
1486.23.45.42.3ccdd
1495.93.05.11.8cbcc
\n", - "

150 rows × 8 columns

\n", - "
" - ], - "text/plain": [ - " sl sw pl pw sl_labeled sw_labeled pl_labeled pw_labeled\n", - "0 5.1 3.5 1.4 0.2 b c a a\n", - "1 4.9 3.0 1.4 0.2 a b a a\n", - "2 4.7 3.2 1.3 0.2 a c a a\n", - "3 4.6 3.1 1.5 0.2 a c a a\n", - "4 5.0 3.6 1.4 0.2 a c a a\n", - "5 5.4 3.9 1.7 0.4 b d a a\n", - "6 4.6 3.4 1.4 0.3 a c a a\n", - "7 5.0 3.4 1.5 0.2 a c a a\n", - "8 4.4 2.9 1.4 0.2 a b a a\n", - "9 4.9 3.1 1.5 0.1 a c a a\n", - "10 5.4 3.7 1.5 0.2 b c a a\n", - "11 4.8 3.4 1.6 0.2 a c a a\n", - "12 4.8 3.0 1.4 0.1 a b a a\n", - "13 4.3 3.0 1.1 0.1 a b a a\n", - "14 5.8 4.0 1.2 0.2 b d a a\n", - "15 5.7 4.4 1.5 0.4 b d a a\n", - "16 5.4 3.9 1.3 0.4 b d a a\n", - "17 5.1 3.5 1.4 0.3 b c a a\n", - "18 5.7 3.8 1.7 0.3 b d a a\n", - "19 5.1 3.8 1.5 0.3 b d a a\n", - "20 5.4 3.4 1.7 0.2 b c a a\n", - "21 5.1 3.7 1.5 0.4 b c a a\n", - "22 4.6 3.6 1.0 0.2 a c a a\n", - "23 5.1 3.3 1.7 0.5 b c a a\n", - "24 4.8 3.4 1.9 0.2 a c a a\n", - "25 5.0 3.0 1.6 0.2 a b a a\n", - "26 5.0 3.4 1.6 0.4 a c a a\n", - "27 5.2 3.5 1.5 0.2 b c a a\n", - "28 5.2 3.4 1.4 0.2 b c a a\n", - "29 4.7 3.2 1.6 0.2 a c a a\n", - ".. ... ... ... ... ... ... ... ...\n", - "120 6.9 3.2 5.7 2.3 d c d d\n", - "121 5.6 2.8 4.9 2.0 b b c d\n", - "122 7.7 2.8 6.7 2.0 d b d d\n", - "123 6.3 2.7 4.9 1.8 c b c c\n", - "124 6.7 3.3 5.7 2.1 c c d d\n", - "125 7.2 3.2 6.0 1.8 d c d c\n", - "126 6.2 2.8 4.8 1.8 c b c c\n", - "127 6.1 3.0 4.9 1.8 c b c c\n", - "128 6.4 2.8 5.6 2.1 c b d d\n", - "129 7.2 3.0 5.8 1.6 d b d c\n", - "130 7.4 2.8 6.1 1.9 d b d d\n", - "131 7.9 3.8 6.4 2.0 d d d d\n", - "132 6.4 2.8 5.6 2.2 c b d d\n", - "133 6.3 2.8 5.1 1.5 c b c c\n", - "134 6.1 2.6 5.6 1.4 c b d c\n", - "135 7.7 3.0 6.1 2.3 d b d d\n", - "136 6.3 3.4 5.6 2.4 c c d d\n", - "137 6.4 3.1 5.5 1.8 c c d c\n", - "138 6.0 3.0 4.8 1.8 c b c c\n", - "139 6.9 3.1 5.4 2.1 d c d d\n", - "140 6.7 3.1 5.6 2.4 c c d d\n", - "141 6.9 3.1 5.1 2.3 d c c d\n", - "142 5.8 2.7 5.1 1.9 b b c d\n", - "143 6.8 3.2 5.9 2.3 c c d d\n", - "144 6.7 3.3 5.7 2.5 c c d d\n", - "145 6.7 3.0 5.2 2.3 c b c d\n", - "146 6.3 2.5 5.0 1.9 c a c d\n", - "147 6.5 3.0 5.2 2.0 c b c d\n", - "148 6.2 3.4 5.4 2.3 c c d d\n", - "149 5.9 3.0 5.1 1.8 c b c c\n", - "\n", - "[150 rows x 8 columns]" - ] - }, - "execution_count": 7, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df['sl_labeled'] = toLabel(df, 'sl')\n", - "df['sw_labeled'] = toLabel(df, 'sw')\n", - "df['pl_labeled'] = toLabel(df, 'pl')\n", - "df['pw_labeled'] = toLabel(df, 'pw')\n", - "df" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [], - "source": [ - "df.drop(['sl', 'sw', 'pl', 'pw'], axis = 1, inplace = True)" - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "{'a', 'b', 'c', 'd'}" - ] - }, - "execution_count": 9, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "set(df['sl_labeled'])" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [], - "source": [ - "df[\"output\"] = iris.target" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "data": { - "text/html": [ - "
\n", - "\n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - " \n", - "
sl_labeledsw_labeledpl_labeledpw_labeledoutput
0bcaa0
1abaa0
2acaa0
3acaa0
4acaa0
5bdaa0
6acaa0
7acaa0
8abaa0
9acaa0
10bcaa0
11acaa0
12abaa0
13abaa0
14bdaa0
15bdaa0
16bdaa0
17bcaa0
18bdaa0
19bdaa0
20bcaa0
21bcaa0
22acaa0
23bcaa0
24acaa0
25abaa0
26acaa0
27bcaa0
28bcaa0
29acaa0
..................
120dcdd2
121bbcd2
122dbdd2
123cbcc2
124ccdd2
125dcdc2
126cbcc2
127cbcc2
128cbdd2
129dbdc2
130dbdd2
131dddd2
132cbdd2
133cbcc2
134cbdc2
135dbdd2
136ccdd2
137ccdc2
138cbcc2
139dcdd2
140ccdd2
141dccd2
142bbcd2
143ccdd2
144ccdd2
145cbcd2
146cacd2
147cbcd2
148ccdd2
149cbcc2
\n", - "

150 rows × 5 columns

\n", - "
" - ], - "text/plain": [ - " sl_labeled sw_labeled pl_labeled pw_labeled output\n", - "0 b c a a 0\n", - "1 a b a a 0\n", - "2 a c a a 0\n", - "3 a c a a 0\n", - "4 a c a a 0\n", - "5 b d a a 0\n", - "6 a c a a 0\n", - "7 a c a a 0\n", - "8 a b a a 0\n", - "9 a c a a 0\n", - "10 b c a a 0\n", - "11 a c a a 0\n", - "12 a b a a 0\n", - "13 a b a a 0\n", - "14 b d a a 0\n", - "15 b d a a 0\n", - "16 b d a a 0\n", - "17 b c a a 0\n", - "18 b d a a 0\n", - "19 b d a a 0\n", - "20 b c a a 0\n", - "21 b c a a 0\n", - "22 a c a a 0\n", - "23 b c a a 0\n", - "24 a c a a 0\n", - "25 a b a a 0\n", - "26 a c a a 0\n", - "27 b c a a 0\n", - "28 b c a a 0\n", - "29 a c a a 0\n", - ".. ... ... ... ... ...\n", - "120 d c d d 2\n", - "121 b b c d 2\n", - "122 d b d d 2\n", - "123 c b c c 2\n", - "124 c c d d 2\n", - "125 d c d c 2\n", - "126 c b c c 2\n", - "127 c b c c 2\n", - "128 c b d d 2\n", - "129 d b d c 2\n", - "130 d b d d 2\n", - "131 d d d d 2\n", - "132 c b d d 2\n", - "133 c b c c 2\n", - "134 c b d c 2\n", - "135 d b d d 2\n", - "136 c c d d 2\n", - "137 c c d c 2\n", - "138 c b c c 2\n", - "139 d c d d 2\n", - "140 c c d d 2\n", - "141 d c c d 2\n", - "142 b b c d 2\n", - "143 c c d d 2\n", - "144 c c d d 2\n", - "145 c b c d 2\n", - "146 c a c d 2\n", - "147 c b c d 2\n", - "148 c c d d 2\n", - "149 c b c c 2\n", - "\n", - "[150 rows x 5 columns]" - ] - }, - "execution_count": 13, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "df" - ] - }, - { - "cell_type": "code", - "execution_count": 14, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "def fit(data):\n", - " output_name = data.columns[-1]\n", - " features = data.columns[0:-1]\n", - " counts = {}\n", - " possible_outputs = set(data[output_name])\n", - " for output in possible_outputs:\n", - " counts[output] = {}\n", - " smallData = data[data[output_name] == output]\n", - " counts[output][\"total_count\"] = len(smallData)\n", - " for f in features:\n", - " counts[output][f] = {}\n", - " possible_values = set(smallData[f])\n", - " for value in possible_values:\n", - " val_count = len(smallData[smallData[f] == value])\n", - " counts[output][f][value] = val_count\n", - " return counts" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "{0: {'pl_labeled': {'a': 50},\n", - " 'pw_labeled': {'a': 50},\n", - " 'sl_labeled': {'a': 28, 'b': 22},\n", - " 'sw_labeled': {'a': 1, 'b': 7, 'c': 32, 'd': 10},\n", - " 'total_count': 50},\n", - " 1: {'pl_labeled': {'b': 7, 'c': 43},\n", - " 'pw_labeled': {'b': 10, 'c': 40},\n", - " 'sl_labeled': {'a': 3, 'b': 21, 'c': 24, 'd': 2},\n", - " 'sw_labeled': {'a': 13, 'b': 29, 'c': 8},\n", - " 'total_count': 50},\n", - " 2: {'pl_labeled': {'c': 20, 'd': 30},\n", - " 'pw_labeled': {'c': 16, 'd': 34},\n", - " 'sl_labeled': {'a': 1, 'b': 5, 'c': 29, 'd': 15},\n", - " 'sw_labeled': {'a': 5, 'b': 28, 'c': 15, 'd': 2},\n", - " 'total_count': 50}}" - ] - }, - "execution_count": 15, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "fit(df)" - ] - } - ], - "metadata": { - "anaconda-cloud": {}, - "kernelspec": { - "display_name": "Python [default]", - "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": "3.6.5" - } - }, - "nbformat": 4, - "nbformat_minor": 1 -} diff --git a/machine_learning/random_forest_classification/Social_Network_Ads.csv b/machine_learning/random_forest_classification/Social_Network_Ads.csv deleted file mode 100644 index 4a53849c2..000000000 --- a/machine_learning/random_forest_classification/Social_Network_Ads.csv +++ /dev/null @@ -1,401 +0,0 @@ -User ID,Gender,Age,EstimatedSalary,Purchased -15624510,Male,19,19000,0 -15810944,Male,35,20000,0 -15668575,Female,26,43000,0 -15603246,Female,27,57000,0 -15804002,Male,19,76000,0 -15728773,Male,27,58000,0 -15598044,Female,27,84000,0 -15694829,Female,32,150000,1 -15600575,Male,25,33000,0 -15727311,Female,35,65000,0 -15570769,Female,26,80000,0 -15606274,Female,26,52000,0 -15746139,Male,20,86000,0 -15704987,Male,32,18000,0 -15628972,Male,18,82000,0 -15697686,Male,29,80000,0 -15733883,Male,47,25000,1 -15617482,Male,45,26000,1 -15704583,Male,46,28000,1 -15621083,Female,48,29000,1 -15649487,Male,45,22000,1 -15736760,Female,47,49000,1 -15714658,Male,48,41000,1 -15599081,Female,45,22000,1 -15705113,Male,46,23000,1 -15631159,Male,47,20000,1 -15792818,Male,49,28000,1 -15633531,Female,47,30000,1 -15744529,Male,29,43000,0 -15669656,Male,31,18000,0 -15581198,Male,31,74000,0 -15729054,Female,27,137000,1 -15573452,Female,21,16000,0 -15776733,Female,28,44000,0 -15724858,Male,27,90000,0 -15713144,Male,35,27000,0 -15690188,Female,33,28000,0 -15689425,Male,30,49000,0 -15671766,Female,26,72000,0 -15782806,Female,27,31000,0 -15764419,Female,27,17000,0 -15591915,Female,33,51000,0 -15772798,Male,35,108000,0 -15792008,Male,30,15000,0 -15715541,Female,28,84000,0 -15639277,Male,23,20000,0 -15798850,Male,25,79000,0 -15776348,Female,27,54000,0 -15727696,Male,30,135000,1 -15793813,Female,31,89000,0 -15694395,Female,24,32000,0 -15764195,Female,18,44000,0 -15744919,Female,29,83000,0 -15671655,Female,35,23000,0 -15654901,Female,27,58000,0 -15649136,Female,24,55000,0 -15775562,Female,23,48000,0 -15807481,Male,28,79000,0 -15642885,Male,22,18000,0 -15789109,Female,32,117000,0 -15814004,Male,27,20000,0 -15673619,Male,25,87000,0 -15595135,Female,23,66000,0 -15583681,Male,32,120000,1 -15605000,Female,59,83000,0 -15718071,Male,24,58000,0 -15679760,Male,24,19000,0 -15654574,Female,23,82000,0 -15577178,Female,22,63000,0 -15595324,Female,31,68000,0 -15756932,Male,25,80000,0 -15726358,Female,24,27000,0 -15595228,Female,20,23000,0 -15782530,Female,33,113000,0 -15592877,Male,32,18000,0 -15651983,Male,34,112000,1 -15746737,Male,18,52000,0 -15774179,Female,22,27000,0 -15667265,Female,28,87000,0 -15655123,Female,26,17000,0 -15595917,Male,30,80000,0 -15668385,Male,39,42000,0 -15709476,Male,20,49000,0 -15711218,Male,35,88000,0 -15798659,Female,30,62000,0 -15663939,Female,31,118000,1 -15694946,Male,24,55000,0 -15631912,Female,28,85000,0 -15768816,Male,26,81000,0 -15682268,Male,35,50000,0 -15684801,Male,22,81000,0 -15636428,Female,30,116000,0 -15809823,Male,26,15000,0 -15699284,Female,29,28000,0 -15786993,Female,29,83000,0 -15709441,Female,35,44000,0 -15710257,Female,35,25000,0 -15582492,Male,28,123000,1 -15575694,Male,35,73000,0 -15756820,Female,28,37000,0 -15766289,Male,27,88000,0 -15593014,Male,28,59000,0 -15584545,Female,32,86000,0 -15675949,Female,33,149000,1 -15672091,Female,19,21000,0 -15801658,Male,21,72000,0 -15706185,Female,26,35000,0 -15789863,Male,27,89000,0 -15720943,Male,26,86000,0 -15697997,Female,38,80000,0 -15665416,Female,39,71000,0 -15660200,Female,37,71000,0 -15619653,Male,38,61000,0 -15773447,Male,37,55000,0 -15739160,Male,42,80000,0 -15689237,Male,40,57000,0 -15679297,Male,35,75000,0 -15591433,Male,36,52000,0 -15642725,Male,40,59000,0 -15701962,Male,41,59000,0 -15811613,Female,36,75000,0 -15741049,Male,37,72000,0 -15724423,Female,40,75000,0 -15574305,Male,35,53000,0 -15678168,Female,41,51000,0 -15697020,Female,39,61000,0 -15610801,Male,42,65000,0 -15745232,Male,26,32000,0 -15722758,Male,30,17000,0 -15792102,Female,26,84000,0 -15675185,Male,31,58000,0 -15801247,Male,33,31000,0 -15725660,Male,30,87000,0 -15638963,Female,21,68000,0 -15800061,Female,28,55000,0 -15578006,Male,23,63000,0 -15668504,Female,20,82000,0 -15687491,Male,30,107000,1 -15610403,Female,28,59000,0 -15741094,Male,19,25000,0 -15807909,Male,19,85000,0 -15666141,Female,18,68000,0 -15617134,Male,35,59000,0 -15783029,Male,30,89000,0 -15622833,Female,34,25000,0 -15746422,Female,24,89000,0 -15750839,Female,27,96000,1 -15749130,Female,41,30000,0 -15779862,Male,29,61000,0 -15767871,Male,20,74000,0 -15679651,Female,26,15000,0 -15576219,Male,41,45000,0 -15699247,Male,31,76000,0 -15619087,Female,36,50000,0 -15605327,Male,40,47000,0 -15610140,Female,31,15000,0 -15791174,Male,46,59000,0 -15602373,Male,29,75000,0 -15762605,Male,26,30000,0 -15598840,Female,32,135000,1 -15744279,Male,32,100000,1 -15670619,Male,25,90000,0 -15599533,Female,37,33000,0 -15757837,Male,35,38000,0 -15697574,Female,33,69000,0 -15578738,Female,18,86000,0 -15762228,Female,22,55000,0 -15614827,Female,35,71000,0 -15789815,Male,29,148000,1 -15579781,Female,29,47000,0 -15587013,Male,21,88000,0 -15570932,Male,34,115000,0 -15794661,Female,26,118000,0 -15581654,Female,34,43000,0 -15644296,Female,34,72000,0 -15614420,Female,23,28000,0 -15609653,Female,35,47000,0 -15594577,Male,25,22000,0 -15584114,Male,24,23000,0 -15673367,Female,31,34000,0 -15685576,Male,26,16000,0 -15774727,Female,31,71000,0 -15694288,Female,32,117000,1 -15603319,Male,33,43000,0 -15759066,Female,33,60000,0 -15814816,Male,31,66000,0 -15724402,Female,20,82000,0 -15571059,Female,33,41000,0 -15674206,Male,35,72000,0 -15715160,Male,28,32000,0 -15730448,Male,24,84000,0 -15662067,Female,19,26000,0 -15779581,Male,29,43000,0 -15662901,Male,19,70000,0 -15689751,Male,28,89000,0 -15667742,Male,34,43000,0 -15738448,Female,30,79000,0 -15680243,Female,20,36000,0 -15745083,Male,26,80000,0 -15708228,Male,35,22000,0 -15628523,Male,35,39000,0 -15708196,Male,49,74000,0 -15735549,Female,39,134000,1 -15809347,Female,41,71000,0 -15660866,Female,58,101000,1 -15766609,Female,47,47000,0 -15654230,Female,55,130000,1 -15794566,Female,52,114000,0 -15800890,Female,40,142000,1 -15697424,Female,46,22000,0 -15724536,Female,48,96000,1 -15735878,Male,52,150000,1 -15707596,Female,59,42000,0 -15657163,Male,35,58000,0 -15622478,Male,47,43000,0 -15779529,Female,60,108000,1 -15636023,Male,49,65000,0 -15582066,Male,40,78000,0 -15666675,Female,46,96000,0 -15732987,Male,59,143000,1 -15789432,Female,41,80000,0 -15663161,Male,35,91000,1 -15694879,Male,37,144000,1 -15593715,Male,60,102000,1 -15575002,Female,35,60000,0 -15622171,Male,37,53000,0 -15795224,Female,36,126000,1 -15685346,Male,56,133000,1 -15691808,Female,40,72000,0 -15721007,Female,42,80000,1 -15794253,Female,35,147000,1 -15694453,Male,39,42000,0 -15813113,Male,40,107000,1 -15614187,Male,49,86000,1 -15619407,Female,38,112000,0 -15646227,Male,46,79000,1 -15660541,Male,40,57000,0 -15753874,Female,37,80000,0 -15617877,Female,46,82000,0 -15772073,Female,53,143000,1 -15701537,Male,42,149000,1 -15736228,Male,38,59000,0 -15780572,Female,50,88000,1 -15769596,Female,56,104000,1 -15586996,Female,41,72000,0 -15722061,Female,51,146000,1 -15638003,Female,35,50000,0 -15775590,Female,57,122000,1 -15730688,Male,41,52000,0 -15753102,Female,35,97000,1 -15810075,Female,44,39000,0 -15723373,Male,37,52000,0 -15795298,Female,48,134000,1 -15584320,Female,37,146000,1 -15724161,Female,50,44000,0 -15750056,Female,52,90000,1 -15609637,Female,41,72000,0 -15794493,Male,40,57000,0 -15569641,Female,58,95000,1 -15815236,Female,45,131000,1 -15811177,Female,35,77000,0 -15680587,Male,36,144000,1 -15672821,Female,55,125000,1 -15767681,Female,35,72000,0 -15600379,Male,48,90000,1 -15801336,Female,42,108000,1 -15721592,Male,40,75000,0 -15581282,Male,37,74000,0 -15746203,Female,47,144000,1 -15583137,Male,40,61000,0 -15680752,Female,43,133000,0 -15688172,Female,59,76000,1 -15791373,Male,60,42000,1 -15589449,Male,39,106000,1 -15692819,Female,57,26000,1 -15727467,Male,57,74000,1 -15734312,Male,38,71000,0 -15764604,Male,49,88000,1 -15613014,Female,52,38000,1 -15759684,Female,50,36000,1 -15609669,Female,59,88000,1 -15685536,Male,35,61000,0 -15750447,Male,37,70000,1 -15663249,Female,52,21000,1 -15638646,Male,48,141000,0 -15734161,Female,37,93000,1 -15631070,Female,37,62000,0 -15761950,Female,48,138000,1 -15649668,Male,41,79000,0 -15713912,Female,37,78000,1 -15586757,Male,39,134000,1 -15596522,Male,49,89000,1 -15625395,Male,55,39000,1 -15760570,Male,37,77000,0 -15566689,Female,35,57000,0 -15725794,Female,36,63000,0 -15673539,Male,42,73000,1 -15705298,Female,43,112000,1 -15675791,Male,45,79000,0 -15747043,Male,46,117000,1 -15736397,Female,58,38000,1 -15678201,Male,48,74000,1 -15720745,Female,37,137000,1 -15637593,Male,37,79000,1 -15598070,Female,40,60000,0 -15787550,Male,42,54000,0 -15603942,Female,51,134000,0 -15733973,Female,47,113000,1 -15596761,Male,36,125000,1 -15652400,Female,38,50000,0 -15717893,Female,42,70000,0 -15622585,Male,39,96000,1 -15733964,Female,38,50000,0 -15753861,Female,49,141000,1 -15747097,Female,39,79000,0 -15594762,Female,39,75000,1 -15667417,Female,54,104000,1 -15684861,Male,35,55000,0 -15742204,Male,45,32000,1 -15623502,Male,36,60000,0 -15774872,Female,52,138000,1 -15611191,Female,53,82000,1 -15674331,Male,41,52000,0 -15619465,Female,48,30000,1 -15575247,Female,48,131000,1 -15695679,Female,41,60000,0 -15713463,Male,41,72000,0 -15785170,Female,42,75000,0 -15796351,Male,36,118000,1 -15639576,Female,47,107000,1 -15693264,Male,38,51000,0 -15589715,Female,48,119000,1 -15769902,Male,42,65000,0 -15587177,Male,40,65000,0 -15814553,Male,57,60000,1 -15601550,Female,36,54000,0 -15664907,Male,58,144000,1 -15612465,Male,35,79000,0 -15810800,Female,38,55000,0 -15665760,Male,39,122000,1 -15588080,Female,53,104000,1 -15776844,Male,35,75000,0 -15717560,Female,38,65000,0 -15629739,Female,47,51000,1 -15729908,Male,47,105000,1 -15716781,Female,41,63000,0 -15646936,Male,53,72000,1 -15768151,Female,54,108000,1 -15579212,Male,39,77000,0 -15721835,Male,38,61000,0 -15800515,Female,38,113000,1 -15591279,Male,37,75000,0 -15587419,Female,42,90000,1 -15750335,Female,37,57000,0 -15699619,Male,36,99000,1 -15606472,Male,60,34000,1 -15778368,Male,54,70000,1 -15671387,Female,41,72000,0 -15573926,Male,40,71000,1 -15709183,Male,42,54000,0 -15577514,Male,43,129000,1 -15778830,Female,53,34000,1 -15768072,Female,47,50000,1 -15768293,Female,42,79000,0 -15654456,Male,42,104000,1 -15807525,Female,59,29000,1 -15574372,Female,58,47000,1 -15671249,Male,46,88000,1 -15779744,Male,38,71000,0 -15624755,Female,54,26000,1 -15611430,Female,60,46000,1 -15774744,Male,60,83000,1 -15629885,Female,39,73000,0 -15708791,Male,59,130000,1 -15793890,Female,37,80000,0 -15646091,Female,46,32000,1 -15596984,Female,46,74000,0 -15800215,Female,42,53000,0 -15577806,Male,41,87000,1 -15749381,Female,58,23000,1 -15683758,Male,42,64000,0 -15670615,Male,48,33000,1 -15715622,Female,44,139000,1 -15707634,Male,49,28000,1 -15806901,Female,57,33000,1 -15775335,Male,56,60000,1 -15724150,Female,49,39000,1 -15627220,Male,39,71000,0 -15672330,Male,47,34000,1 -15668521,Female,48,35000,1 -15807837,Male,48,33000,1 -15592570,Male,47,23000,1 -15748589,Female,45,45000,1 -15635893,Male,60,42000,1 -15757632,Female,39,59000,0 -15691863,Female,46,41000,1 -15706071,Male,51,23000,1 -15654296,Female,50,20000,1 -15755018,Male,36,33000,0 -15594041,Female,49,36000,1 \ No newline at end of file diff --git a/machine_learning/random_forest_classification/random_forest_classification.py b/machine_learning/random_forest_classification/random_forest_classification.py deleted file mode 100644 index 6aed4e6e6..000000000 --- a/machine_learning/random_forest_classification/random_forest_classification.py +++ /dev/null @@ -1,103 +0,0 @@ -# Random Forest Classification - -# Importing the libraries -import os -import numpy as np -import matplotlib.pyplot as plt -import pandas as pd - -# Importing the dataset -script_dir = os.path.dirname(os.path.realpath(__file__)) -dataset = pd.read_csv(os.path.join(script_dir, "Social_Network_Ads.csv")) -X = dataset.iloc[:, [2, 3]].values -y = dataset.iloc[:, 4].values - -# Splitting the dataset into the Training set and Test set -from sklearn.model_selection import train_test_split - -X_train, X_test, y_train, y_test = train_test_split( - X, y, test_size=0.25, random_state=0 -) - -# Feature Scaling -from sklearn.preprocessing import StandardScaler - -sc = StandardScaler() -X_train = sc.fit_transform(X_train) -X_test = sc.transform(X_test) - -# Fitting Random Forest Classification to the Training set -from sklearn.ensemble import RandomForestClassifier - -classifier = RandomForestClassifier( - n_estimators=10, criterion="entropy", random_state=0 -) -classifier.fit(X_train, y_train) - -# Predicting the Test set results -y_pred = classifier.predict(X_test) - -# Making the Confusion Matrix -from sklearn.metrics import confusion_matrix - -cm = confusion_matrix(y_test, y_pred) - -# Visualising the Training set results -from matplotlib.colors import ListedColormap - -X_set, y_set = X_train, y_train -X1, X2 = np.meshgrid( - np.arange(start=X_set[:, 0].min() - 1, stop=X_set[:, 0].max() + 1, step=0.01), - np.arange(start=X_set[:, 1].min() - 1, stop=X_set[:, 1].max() + 1, step=0.01), -) -plt.contourf( - X1, - X2, - classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape), - alpha=0.75, - cmap=ListedColormap(("red", "green")), -) -plt.xlim(X1.min(), X1.max()) -plt.ylim(X2.min(), X2.max()) -for i, j in enumerate(np.unique(y_set)): - plt.scatter( - X_set[y_set == j, 0], - X_set[y_set == j, 1], - c=ListedColormap(("red", "green"))(i), - label=j, - ) -plt.title("Random Forest Classification (Training set)") -plt.xlabel("Age") -plt.ylabel("Estimated Salary") -plt.legend() -plt.show() - -# Visualising the Test set results -from matplotlib.colors import ListedColormap - -X_set, y_set = X_test, y_test -X1, X2 = np.meshgrid( - np.arange(start=X_set[:, 0].min() - 1, stop=X_set[:, 0].max() + 1, step=0.01), - np.arange(start=X_set[:, 1].min() - 1, stop=X_set[:, 1].max() + 1, step=0.01), -) -plt.contourf( - X1, - X2, - classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape), - alpha=0.75, - cmap=ListedColormap(("red", "green")), -) -plt.xlim(X1.min(), X1.max()) -plt.ylim(X2.min(), X2.max()) -for i, j in enumerate(np.unique(y_set)): - plt.scatter( - X_set[y_set == j, 0], - X_set[y_set == j, 1], - c=ListedColormap(("red", "green"))(i), - label=j, - ) -plt.title("Random Forest Classification (Test set)") -plt.xlabel("Age") -plt.ylabel("Estimated Salary") -plt.legend() -plt.show() diff --git a/machine_learning/random_forest_classification/random_forest_classifier.ipynb b/machine_learning/random_forest_classification/random_forest_classifier.ipynb deleted file mode 100644 index 7ee66124c..000000000 --- a/machine_learning/random_forest_classification/random_forest_classifier.ipynb +++ /dev/null @@ -1,196 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\Satyam\\AppData\\Roaming\\Python\\Python35\\site-packages\\sklearn\\ensemble\\weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n", - " from numpy.core.umath_tests import inner1d\n" - ] - } - ], - "source": [ - "# Importing the libraries\n", - "import numpy as np\n", - "import matplotlib.pyplot as plt\n", - "import pandas as pd\n", - "from sklearn.model_selection import train_test_split\n", - "from sklearn.preprocessing import StandardScaler\n", - "from sklearn.metrics import confusion_matrix\n", - "from matplotlib.colors import ListedColormap\n", - "from sklearn.ensemble import RandomForestClassifier" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "# Importing the dataset\n", - "dataset = pd.read_csv('Social_Network_Ads.csv')\n", - "X = dataset.iloc[:, [2, 3]].values\n", - "y = dataset.iloc[:, 4].values" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "# Splitting the dataset into the Training set and Test set\n", - "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 0)" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\Satyam\\AppData\\Roaming\\Python\\Python35\\site-packages\\sklearn\\utils\\validation.py:475: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n", - " warnings.warn(msg, DataConversionWarning)\n" - ] - } - ], - "source": [ - "# Feature Scaling\n", - "sc = StandardScaler()\n", - "X_train = sc.fit_transform(X_train)\n", - "X_test = sc.transform(X_test)" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "[[63 5]\n", - " [ 3 29]]\n" - ] - } - ], - "source": [ - "# Fitting classifier to the Training set\n", - "# Create your classifier here\n", - "classifier = RandomForestClassifier(n_estimators=10,criterion='entropy',random_state=0)\n", - "classifier.fit(X_train,y_train)\n", - "# Predicting the Test set results\n", - "y_pred = classifier.predict(X_test)\n", - "\n", - "# Making the Confusion Matrix\n", - "cm = confusion_matrix(y_test, y_pred)\n", - "print(cm)" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEWCAYAAABmE+CbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAfyzKuSV3NT5AAAAAElFTkSuQmCC\n", - "text/plain": [ - "" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEWCAYAAABmE+CbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n", - "text/plain": [ - "" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Visualising the Training set results\n", - "X_set, y_set = X_train, y_train\n", - "X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),\n", - " np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.01))\n", - "plt.contourf(X1, X2, classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),\n", - " alpha = 0.75, cmap = ListedColormap(('red', 'green')))\n", - "plt.xlim(X1.min(), X1.max())\n", - "plt.ylim(X2.min(), X2.max())\n", - "for i, j in enumerate(np.unique(y_set)):\n", - " plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],\n", - " c = ListedColormap(('red', 'green'))(i), label = j)\n", - "plt.title('Random Forest Classifier (Training set)')\n", - "plt.xlabel('Age')\n", - "plt.ylabel('Estimated Salary')\n", - "plt.legend()\n", - "plt.show()\n", - "\n", - "# Visualising the Test set results\n", - "X_set, y_set = X_test, y_test\n", - "X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),\n", - " np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.01))\n", - "plt.contourf(X1, X2, classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),\n", - " alpha = 0.75, cmap = ListedColormap(('red', 'green')))\n", - "plt.xlim(X1.min(), X1.max())\n", - "plt.ylim(X2.min(), X2.max())\n", - "for i, j in enumerate(np.unique(y_set)):\n", - " plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],\n", - " c = ListedColormap(('red', 'green'))(i), label = j)\n", - "plt.title('Random Forest Classifier (Test set)')\n", - "plt.xlabel('Age')\n", - "plt.ylabel('Estimated Salary')\n", - "plt.legend()\n", - "plt.show()" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": true - }, - "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": "3.5.1" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -} diff --git a/machine_learning/random_forest_regression/Position_Salaries.csv b/machine_learning/random_forest_regression/Position_Salaries.csv deleted file mode 100644 index 0c752c72a..000000000 --- a/machine_learning/random_forest_regression/Position_Salaries.csv +++ /dev/null @@ -1,11 +0,0 @@ -Position,Level,Salary -Business Analyst,1,45000 -Junior Consultant,2,50000 -Senior Consultant,3,60000 -Manager,4,80000 -Country Manager,5,110000 -Region Manager,6,150000 -Partner,7,200000 -Senior Partner,8,300000 -C-level,9,500000 -CEO,10,1000000 \ No newline at end of file diff --git a/machine_learning/random_forest_regression/random_forest_regression.ipynb b/machine_learning/random_forest_regression/random_forest_regression.ipynb deleted file mode 100644 index 17f4d42bf..000000000 --- a/machine_learning/random_forest_regression/random_forest_regression.ipynb +++ /dev/null @@ -1,147 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 1, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "# Importing the libraries\n", - "import numpy as np\n", - "import matplotlib.pyplot as plt\n", - "import pandas as pd\n", - "from sklearn.ensemble import RandomForestRegressor" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "# Importing the dataset\n", - "dataset = pd.read_csv('Position_Salaries.csv')\n", - "X = dataset.iloc[:, 1:2].values\n", - "y = dataset.iloc[:, 2].values" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=None,\n", - " max_features='auto', max_leaf_nodes=None,\n", - " min_impurity_split=1e-07, min_samples_leaf=1,\n", - " min_samples_split=2, min_weight_fraction_leaf=0.0,\n", - " n_estimators=300, n_jobs=1, oob_score=False, random_state=0,\n", - " verbose=0, warm_start=False)" - ] - }, - "execution_count": 3, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "# Fitting Random Forest Regression to the dataset\n", - "regressor = RandomForestRegressor(n_estimators = 300, random_state = 0)\n", - "regressor.fit(X, y)" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "# Predicting a new result\n", - "y_pred = regressor.predict(6.5)" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "[ 160333.33333333]\n" - ] - } - ], - "source": [ - "print(y_pred)" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaEAAAEWCAYAAADPZygPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n", - "text/plain": [ - "" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Visualising the Random Forest Regression results (higher resolution)\n", - "X_grid = np.arange(min(X), max(X), 0.01)\n", - "X_grid = X_grid.reshape((len(X_grid), 1))\n", - "plt.scatter(X, y, color = 'red')\n", - "plt.plot(X_grid, regressor.predict(X_grid), color = 'blue')\n", - "plt.title('Truth or Bluff (Random Forest Regression)')\n", - "plt.xlabel('Position level')\n", - "plt.ylabel('Salary')\n", - "plt.show()" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": true - }, - "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": "3.5.1" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -} diff --git a/machine_learning/random_forest_regression/random_forest_regression.py b/machine_learning/random_forest_regression/random_forest_regression.py deleted file mode 100644 index 2599e97e9..000000000 --- a/machine_learning/random_forest_regression/random_forest_regression.py +++ /dev/null @@ -1,44 +0,0 @@ -# Random Forest Regression - -# Importing the libraries -import os -import numpy as np -import matplotlib.pyplot as plt -import pandas as pd - -# Importing the dataset -script_dir = os.path.dirname(os.path.realpath(__file__)) -dataset = pd.read_csv(os.path.join(script_dir, "Position_Salaries.csv")) -X = dataset.iloc[:, 1:2].values -y = dataset.iloc[:, 2].values - -# Splitting the dataset into the Training set and Test set -"""from sklearn.cross_validation import train_test_split -X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)""" - -# Feature Scaling -"""from sklearn.preprocessing import StandardScaler -sc_X = StandardScaler() -X_train = sc_X.fit_transform(X_train) -X_test = sc_X.transform(X_test) -sc_y = StandardScaler() -y_train = sc_y.fit_transform(y_train)""" - -# Fitting Random Forest Regression to the dataset -from sklearn.ensemble import RandomForestRegressor - -regressor = RandomForestRegressor(n_estimators=10, random_state=0) -regressor.fit(X, y) - -# Predicting a new result -y_pred = regressor.predict([[6.5]]) - -# Visualising the Random Forest Regression results (higher resolution) -X_grid = np.arange(min(X), max(X), 0.01) -X_grid = X_grid.reshape((len(X_grid), 1)) -plt.scatter(X, y, color="red") -plt.plot(X_grid, regressor.predict(X_grid), color="blue") -plt.title("Truth or Bluff (Random Forest Regression)") -plt.xlabel("Position level") -plt.ylabel("Salary") -plt.show() diff --git a/machine_learning/reuters_one_vs_rest_classifier.ipynb b/machine_learning/reuters_one_vs_rest_classifier.ipynb deleted file mode 100644 index 968130a60..000000000 --- a/machine_learning/reuters_one_vs_rest_classifier.ipynb +++ /dev/null @@ -1,405 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [], - "source": [ - "try:\n", - " import nltk\n", - "except ModuleNotFoundError:\n", - " !pip install nltk" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [], - "source": [ - "## This code downloads the required packages.\n", - "## You can run `nltk.download('all')` to download everything.\n", - "\n", - "nltk_packages = [\n", - " (\"reuters\", \"corpora/reuters.zip\")\n", - "]\n", - "\n", - "for pid, fid in nltk_packages:\n", - " try:\n", - " nltk.data.find(fid)\n", - " except LookupError:\n", - " nltk.download(pid)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Setting up corpus" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [], - "source": [ - "from nltk.corpus import reuters" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Setting up train/test data" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [], - "source": [ - "train_documents, train_categories = zip(*[(reuters.raw(i), reuters.categories(i)) for i in reuters.fileids() if i.startswith('training/')])\n", - "test_documents, test_categories = zip(*[(reuters.raw(i), reuters.categories(i)) for i in reuters.fileids() if i.startswith('test/')])" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "all_categories = sorted(list(set(reuters.categories())))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "The following cell defines a function **tokenize** that performs following actions:\n", - "- Receive a document as an argument to the function\n", - "- Tokenize the document using `nltk.word_tokenize()`\n", - "- Use `PorterStemmer` provided by the `nltk` to remove morphological affixes from each token\n", - "- Append stemmed token to an already defined list `stems`\n", - "- Return the list `stems`" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [], - "source": [ - "from nltk.stem.porter import PorterStemmer\n", - "def tokenize(text):\n", - " tokens = nltk.word_tokenize(text)\n", - " stems = []\n", - " for item in tokens:\n", - " stems.append(PorterStemmer().stem(item))\n", - " return stems" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "To begin, I first used TF-IDF for feature selection on both train as well as test data using `TfidfVectorizer`.\n", - "\n", - "But first, What `TfidfVectorizer` actually does?\n", - "- `TfidfVectorizer` converts a collection of raw documents to a matrix of **TF-IDF** features.\n", - "\n", - "**TF-IDF**?\n", - "- TFIDF (abbreviation of the term *frequency–inverse document frequency*) is a numerical statistic that is intended to reflect how important a word is to a document in a collection or corpus. [tf–idf](https://en.wikipedia.org/wiki/Tf%E2%80%93idf)\n", - "\n", - "**Why `TfidfVectorizer`**?\n", - "- `TfidfVectorizer` scale down the impact of tokens that occur very frequently (e.g., “a”, “the”, and “of”) in a given corpus. [Feature Extraction and Transformation](https://spark.apache.org/docs/latest/mllib-feature-extraction.html#tf-idf)\n", - "\n", - "I gave following two arguments to `TfidfVectorizer`:\n", - "- tokenizer: `tokenize` function\n", - "- stop_words\n", - "\n", - "Then I used `fit_transform` and `transform` on the train and test documents repectively.\n", - "\n", - "**Why `fit_transform` for training data while `transform` for test data**?\n", - "\n", - "To avoid data leakage during cross-validation, imputer computes the statistic on the train data during the `fit`, **stores it** and uses the same on the test data, during the `transform`. This also prevents the test data from appearing in `fit` operation." - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [], - "source": [ - "from sklearn.feature_extraction.text import TfidfVectorizer\n", - "\n", - "vectorizer = TfidfVectorizer(tokenizer = tokenize, stop_words = 'english')\n", - "\n", - "vectorised_train_documents = vectorizer.fit_transform(train_documents)\n", - "vectorised_test_documents = vectorizer.transform(test_documents)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "For the **efficient implementation** of machine learning algorithms, many machine learning algorithms **requires all input variables and output variables to be numeric**. This means that categorical data must be converted to a numerical form.\n", - "\n", - "For this purpose, I used `MultiLabelBinarizer` from `sklearn.preprocessing`." - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [], - "source": [ - "from sklearn.preprocessing import MultiLabelBinarizer\n", - "\n", - "mlb = MultiLabelBinarizer()\n", - "train_labels = mlb.fit_transform(train_categories)\n", - "test_labels = mlb.transform(test_categories)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Now, To **train** the classifier, I used `LinearSVC` in combination with the `OneVsRestClassifier` function in the scikit-learn package.\n", - "\n", - "The strategy of `OneVsRestClassifier` is of **fitting one classifier per label** and the `OneVsRestClassifier` can efficiently do this task and also outputs are easy to interpret. Since each label is represented by **one and only one classifier**, it is possible to gain knowledge about the label by inspecting its corresponding classifier. [OneVsRestClassifier](http://scikit-learn.org/stable/modules/multiclass.html#one-vs-the-rest)\n", - "\n", - "The reason I combined `LinearSVC` with `OneVsRestClassifier` is because `LinearSVC` supports **Multi-class**, while we want to perform **Multi-label** classification." - ] - }, - { - "cell_type": "code", - "execution_count": 9, - "metadata": {}, - "outputs": [], - "source": [ - "%%capture\n", - "from sklearn.multiclass import OneVsRestClassifier\n", - "from sklearn.svm import LinearSVC\n", - "\n", - "classifier = OneVsRestClassifier(LinearSVC())\n", - "classifier.fit(vectorised_train_documents, train_labels)" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "After fitting the classifier, I decided to use `cross_val_score` to **measure score** of the classifier by **cross validation** on the training data. But the only problem was, I wanted to **shuffle** data to use with `cross_val_score`, but it does not support shuffle argument.\n", - "\n", - "So, I decided to use `KFold` with `cross_val_score` as `KFold` supports shuffling the data.\n", - "\n", - "I also enabled `random_state`, because `random_state` will guarantee the same output in each run. By setting the `random_state`, it is guaranteed that the pseudorandom number generator will generate the same sequence of random integers each time, which in turn will affect the split.\n", - "\n", - "Why **42**?\n", - "- [Why '42' is the preferred number when indicating something random?](https://softwareengineering.stackexchange.com/questions/507/why-42-is-the-preferred-number-when-indicating-something-random)" - ] - }, - { - "cell_type": "code", - "execution_count": 10, - "metadata": {}, - "outputs": [], - "source": [ - "%%capture\n", - "from sklearn.model_selection import KFold, cross_val_score\n", - "\n", - "kf = KFold(n_splits=10, random_state = 42, shuffle = True)\n", - "scores = cross_val_score(classifier, vectorised_train_documents, train_labels, cv = kf)" - ] - }, - { - "cell_type": "code", - "execution_count": 11, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Cross-validation scores: [0.83655084 0.86743887 0.8043758 0.83011583 0.83655084 0.81724582\n", - " 0.82754183 0.8030888 0.80694981 0.82731959]\n", - "Cross-validation accuracy: 0.8257 (+/- 0.0368)\n" - ] - } - ], - "source": [ - "print('Cross-validation scores:', scores)\n", - "print('Cross-validation accuracy: {:.4f} (+/- {:.4f})'.format(scores.mean(), scores.std() * 2))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In the end, I used different methods (`accuracy_score`, `precision_score`, `recall_score`, `f1_score` and `confusion_matrix`) provided by scikit-learn **to evaluate** the classifier. (both *Macro-* and *Micro-averages*)" - ] - }, - { - "cell_type": "code", - "execution_count": 12, - "metadata": {}, - "outputs": [], - "source": [ - "%%capture\n", - "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, confusion_matrix\n", - "\n", - "predictions = classifier.predict(vectorised_test_documents)\n", - "\n", - "accuracy = accuracy_score(test_labels, predictions)\n", - "\n", - "macro_precision = precision_score(test_labels, predictions, average='macro')\n", - "macro_recall = recall_score(test_labels, predictions, average='macro')\n", - "macro_f1 = f1_score(test_labels, predictions, average='macro')\n", - "\n", - "micro_precision = precision_score(test_labels, predictions, average='micro')\n", - "micro_recall = recall_score(test_labels, predictions, average='micro')\n", - "micro_f1 = f1_score(test_labels, predictions, average='micro')\n", - "\n", - "cm = confusion_matrix(test_labels.argmax(axis = 1), predictions.argmax(axis = 1))" - ] - }, - { - "cell_type": "code", - "execution_count": 13, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Accuracy: 0.8099\n", - "Precision:\n", - "- Macro: 0.6076\n", - "- Micro: 0.9471\n", - "Recall:\n", - "- Macro: 0.3708\n", - "- Micro: 0.7981\n", - "F1-measure:\n", - "- Macro: 0.4410\n", - "- Micro: 0.8662\n" - ] - } - ], - "source": [ - "print(\"Accuracy: {:.4f}\\nPrecision:\\n- Macro: {:.4f}\\n- Micro: {:.4f}\\nRecall:\\n- Macro: {:.4f}\\n- Micro: {:.4f}\\nF1-measure:\\n- Macro: {:.4f}\\n- Micro: {:.4f}\".format(accuracy, macro_precision, micro_precision, macro_recall, micro_recall, macro_f1, micro_f1))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "In below cell, I used `matplotlib.pyplot` to **plot the confusion matrix** (of first *few results only* to keep the readings readable) using `heatmap` of `seaborn`." - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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\n", - "text/plain": [ - "" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "import matplotlib.pyplot as plt\n", - "import seaborn as sb\n", - "import pandas as pd\n", - "\n", - "cm_plt = pd.DataFrame(cm[:73])\n", - "\n", - "plt.figure(figsize = (25, 25))\n", - "ax = plt.axes()\n", - "\n", - "sb.heatmap(cm_plt, annot=True)\n", - "\n", - "ax.xaxis.set_ticks_position('top')\n", - "\n", - "plt.show()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Pipeline" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "Now, I took the data from [Coconut - Wikipedia](https://en.wikipedia.org/wiki/Coconut) to check if the classifier is able to **correctly** predict the label(s) or not.\n", - "\n", - "And here is the output:" - ] - }, - { - "cell_type": "code", - "execution_count": 15, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Example labels: [('coconut', 'oilseed')]\n" - ] - } - ], - "source": [ - "example_text = '''The coconut tree (Cocos nucifera) is a member of the family Arecaceae (palm family) and the only species of the genus Cocos.\n", - "The term coconut can refer to the whole coconut palm or the seed, or the fruit, which, botanically, is a drupe, not a nut.\n", - "The spelling cocoanut is an archaic form of the word.\n", - "The term is derived from the 16th-century Portuguese and Spanish word coco meaning \"head\" or \"skull\", from the three indentations on the coconut shell that resemble facial features.\n", - "Coconuts are known for their versatility ranging from food to cosmetics.\n", - "They form a regular part of the diets of many people in the tropics and subtropics.\n", - "Coconuts are distinct from other fruits for their endosperm containing a large quantity of water (also called \"milk\"), and when immature, may be harvested for the potable coconut water.\n", - "When mature, they can be used as seed nuts or processed for oil, charcoal from the hard shell, and coir from the fibrous husk.\n", - "When dried, the coconut flesh is called copra.\n", - "The oil and milk derived from it are commonly used in cooking and frying, as well as in soaps and cosmetics.\n", - "The husks and leaves can be used as material to make a variety of products for furnishing and decorating.\n", - "The coconut also has cultural and religious significance in certain societies, particularly in India, where it is used in Hindu rituals.'''\n", - "\n", - "example_preds = classifier.predict(vectorizer.transform([example_text]))\n", - "example_labels = mlb.inverse_transform(example_preds)\n", - "print(\"Example labels: {}\".format(example_labels))" - ] - } - ], - "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": "3.6.6" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -} diff --git a/neural_network/fully_connected_neural_network.ipynb b/neural_network/fully_connected_neural_network.ipynb deleted file mode 100644 index a8bcf4bee..000000000 --- a/neural_network/fully_connected_neural_network.ipynb +++ /dev/null @@ -1,327 +0,0 @@ -{ - "cells": [ - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "## Standard (Fully Connected) Neural Network" - ] - }, - { - "cell_type": "code", - "execution_count": 1, - "metadata": {}, - "outputs": [], - "source": [ - "#Use in Markup cell type\n", - "#![alt text](imagename.png \"Title\") " - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "### Implementing Fully connected Neural Net" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Loading Required packages and Data" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "Using TensorFlow backend.\n" - ] - } - ], - "source": [ - "###1. Load Data and Splot Data\n", - "from keras.datasets import mnist\n", - "from keras.models import Sequential \n", - "from keras.layers.core import Dense, Activation\n", - "from keras.utils import np_utils\n", - "(X_train, Y_train), (X_test, Y_test) = mnist.load_data()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Preprocessing" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "import matplotlib.pyplot as plt\n", - "n = 10 # how many digits we will display\n", - "plt.figure(figsize=(20, 4))\n", - "for i in range(n):\n", - " # display original\n", - " ax = plt.subplot(2, n, i + 1)\n", - " plt.imshow(X_test[i].reshape(28, 28))\n", - " plt.gray()\n", - " ax.get_xaxis().set_visible(False)\n", - " ax.get_yaxis().set_visible(False)\n", - "plt.show()\n", - "plt.close()" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Previous X_train shape: (60000, 28, 28) \n", - "Previous Y_train shape:(60000,)\n", - "New X_train shape: (60000, 784) \n", - "New Y_train shape:(60000, 10)\n" - ] - } - ], - "source": [ - "print(\"Previous X_train shape: {} \\nPrevious Y_train shape:{}\".format(X_train.shape, Y_train.shape))\n", - "X_train = X_train.reshape(60000, 784) \n", - "X_test = X_test.reshape(10000, 784)\n", - "X_train = X_train.astype('float32') \n", - "X_test = X_test.astype('float32') \n", - "X_train /= 255 \n", - "X_test /= 255\n", - "classes = 10\n", - "Y_train = np_utils.to_categorical(Y_train, classes) \n", - "Y_test = np_utils.to_categorical(Y_test, classes)\n", - "print(\"New X_train shape: {} \\nNew Y_train shape:{}\".format(X_train.shape, Y_train.shape))" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Setting up parameters" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": {}, - "outputs": [], - "source": [ - "input_size = 784\n", - "batch_size = 200 \n", - "hidden1 = 400\n", - "hidden2 = 20\n", - "epochs = 2" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Building the FCN Model" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "_________________________________________________________________\n", - "Layer (type) Output Shape Param # \n", - "=================================================================\n", - "dense_1 (Dense) (None, 400) 314000 \n", - "_________________________________________________________________\n", - "dense_2 (Dense) (None, 20) 8020 \n", - "_________________________________________________________________\n", - "dense_3 (Dense) (None, 10) 210 \n", - "=================================================================\n", - "Total params: 322,230\n", - "Trainable params: 322,230\n", - "Non-trainable params: 0\n", - "_________________________________________________________________\n" - ] - } - ], - "source": [ - "###4.Build the model\n", - "model = Sequential() \n", - "model.add(Dense(hidden1, input_dim=input_size, activation='relu'))\n", - "# output = relu (dot (W, input) + bias)\n", - "model.add(Dense(hidden2, activation='relu'))\n", - "model.add(Dense(classes, activation='softmax')) \n", - "\n", - "# Compilation\n", - "model.compile(loss='categorical_crossentropy', \n", - " metrics=['accuracy'], optimizer='sgd')\n", - "model.summary()" - ] - }, - { - "cell_type": "markdown", - "metadata": {}, - "source": [ - "#### Training The Model" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 1/10\n", - " - 12s - loss: 1.4482 - acc: 0.6251\n", - "Epoch 2/10\n", - " - 3s - loss: 0.6239 - acc: 0.8482\n", - "Epoch 3/10\n", - " - 3s - loss: 0.4582 - acc: 0.8798\n", - "Epoch 4/10\n", - " - 3s - loss: 0.3941 - acc: 0.8936\n", - "Epoch 5/10\n", - " - 3s - loss: 0.3579 - acc: 0.9011\n", - "Epoch 6/10\n", - " - 4s - loss: 0.3328 - acc: 0.9070\n", - "Epoch 7/10\n", - " - 3s - loss: 0.3138 - acc: 0.9118\n", - "Epoch 8/10\n", - " - 3s - loss: 0.2980 - acc: 0.9157\n", - "Epoch 9/10\n", - " - 3s - loss: 0.2849 - acc: 0.9191\n", - "Epoch 10/10\n", - " - 3s - loss: 0.2733 - acc: 0.9223\n" - ] - }, - { - "data": { - "text/plain": [ - "" - ] - }, - "execution_count": 7, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "# Fitting on Data\n", - "model.fit(X_train, Y_train, batch_size=batch_size, epochs=10, verbose=2)\n", - "###5.Test " - ] - }, - { - "cell_type": "markdown", - "metadata": { - "collapsed": true - }, - "source": [ - "#### Testing The Model" - ] - }, - { - "cell_type": "code", - "execution_count": 8, - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "10000/10000 [==============================] - 1s 121us/step\n", - "\n", - "Test accuracy: 0.9257\n", - "[0 6 9 0 1 5 9 7 3 4]\n" - ] - }, - { - "data": { - "image/png": "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\n", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "score = model.evaluate(X_test, Y_test, verbose=1)\n", - "print('\\n''Test accuracy:', score[1])\n", - "mask = range(10,20)\n", - "X_valid = X_test[mask]\n", - "y_pred = model.predict_classes(X_valid)\n", - "print(y_pred)\n", - "plt.figure(figsize=(20, 4))\n", - "for i in range(n):\n", - " # display original\n", - " ax = plt.subplot(2, n, i + 1)\n", - " plt.imshow(X_valid[i].reshape(28, 28))\n", - " plt.gray()\n", - " ax.get_xaxis().set_visible(False)\n", - " ax.get_yaxis().set_visible(False)\n", - "plt.show()\n", - "plt.close()" - ] - }, - { - "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": "3.6.6" - } - }, - "nbformat": 4, - "nbformat_minor": 2 -}