Skip to content

Instantly share code, notes, and snippets.

Embed
What would you like to do?
k-means clustering
{
"cells": [
{
"metadata": {},
"cell_type": "markdown",
"source": "# k-means clustering"
},
{
"metadata": {},
"cell_type": "markdown",
"source": "https://gist.github.com/Cartman0/9455bb73c3f8137880ec66b61e8c4615"
},
{
"metadata": {},
"cell_type": "markdown",
"source": "1. ランダムにクラスを割り振り"
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2019-06-09T11:27:00.531456Z",
"end_time": "2019-06-09T11:27:00.536443Z"
},
"trusted": true
},
"cell_type": "code",
"source": "a = sp.array([sp.nan, sp.nan]).mean()\nsp.argmin([(1-a)**2, (1-a)**2])",
"execution_count": 7,
"outputs": [
{
"output_type": "execute_result",
"execution_count": 7,
"data": {
"text/plain": "0"
},
"metadata": {}
}
]
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2019-06-09T12:36:20.398330Z",
"end_time": "2019-06-09T12:36:20.403396Z"
},
"trusted": true
},
"cell_type": "code",
"source": "a = sp.array([3, 1, 2])\na[sp.where([True, False, False])]",
"execution_count": 36,
"outputs": [
{
"output_type": "execute_result",
"execution_count": 36,
"data": {
"text/plain": "array([3])"
},
"metadata": {}
}
]
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2019-06-09T14:14:12.921525Z",
"end_time": "2019-06-09T14:14:12.947934Z"
},
"code_folding": [],
"trusted": true
},
"cell_type": "code",
"source": "import scipy as sp\nfrom scipy import stats\n\ndef k_means(data, cluster_n = 2, try_num = 1000):\n def calc_cluster_centroid(cluster_list, data_c):\n # update centroid\n '''\n 途中で消える場合を想定\n '''\n cluster_centroid = sp.empty((len(cluster_list), data_c.shape[1]-1))\n c_bool = []\n for c in cluster_list:\n one_cluster_bool = (data_c[:,-1]==c)\n is_cluster = sp.any(one_cluster_bool)\n c_bool.append(is_cluster)\n if not is_cluster:\n continue\n \n idx = sp.where(cluster_list == c)\n cluster_centroid[idx] = sp.mean(data_c[one_cluster_bool, :-1], axis=0)\n\n# print(\"bool\", c_bool)\n# print(\"centroid\", cluster_centroid)\n return cluster_list[c_bool], cluster_centroid[c_bool]\n \n def rename_cluster(old_name, data_c, new_name):\n if sp.all(old_name==new_name):\n return data_c\n \n for idx in sp.arange(len(old_name)):\n data_c[data_c[:, -1]==old_name[idx], -1] = new_name[idx]\n return data_c\n \n cluster_list = sp.arange(0, cluster_n)\n data = sp.array(data)\n data_n = data.shape[0]\n\n # add random label\n c_l = None\n for i in sp.arange(try_num):\n# print(\"select random i:\", i)\n random_cluster = sp.random.choice(cluster_list, data_n).reshape(data_n, 1)\n \n # check\n bool_arr = [sp.any(c==cluster_list) for c in random_cluster]\n if sp.all(bool_arr):\n c_l = random_cluster\n break\n if sp.any(c_l == None):\n print(\"can't select initial random cluster.\")\n return 0\n\n data_c = sp.hstack((data, c_l))\n old_d = data_c\n for i in sp.arange(try_num):\n new_d = old_d.copy()\n# print(\"new_d[:,:-1]: \", new_d[:,:-1])\n# print(\"new_d[:,-1]==1: \", new_d[:,-1]==1)\n \n # update centroid\n cluster_list, cluster_centroid = calc_cluster_centroid(cluster_list, new_d)\n \n # update label\n for d in new_d:\n# print(\"d[:-1]:\", d[:-1])\n# print(\"d[:-1]:-cen\", d[:-1]-centroid)\n distance = sp.sum((d[:-1] - cluster_centroid)**2, axis=1)\n# print(\"distance:\", distance)\n min_idx = sp.argmin(distance)\n# print(\"min_idx:\",min_idx)\n d[-1] = cluster_list[min_idx]\n# print(\"uplabel d\", new_d)\n \n # end check\n if sp.all(new_d[:, -1]==old_d[:, -1]):\n return len(cluster_list), sp.arange(len(cluster_list)), rename_cluster(cluster_list, new_d, sp.arange(len(cluster_list)))\n \n # update for next loop\n old_d = new_d\n return len(cluster_list), sp.arange(len(cluster_list)), rename_cluster(cluster_list, new_d, sp.arange(len(cluster_list)))\n \n\nd = stats.multivariate_normal(mean=(0,0), cov=[[1,0],[0,1]]).rvs(10)\n# print(\"d:\", d)\nmeans = k_means(d)\nprint(len(means))\nmeans",
"execution_count": 60,
"outputs": [
{
"output_type": "stream",
"text": "3\n",
"name": "stdout"
},
{
"output_type": "execute_result",
"execution_count": 60,
"data": {
"text/plain": "(2, array([0, 1]), array([[ 0.62107213, -0.62472083, 1. ],\n [ 0.23010772, 1.50609055, 0. ],\n [ 1.35138413, 0.82156175, 1. ],\n [-0.63822107, -1.02179235, 0. ],\n [ 0.67619316, 0.4423283 , 1. ],\n [ 0.58199731, 0.65678262, 1. ],\n [ 1.46517772, 0.55407695, 1. ],\n [-0.2686579 , 1.08347979, 0. ],\n [-0.05886341, 1.26236087, 0. ],\n [ 0.62881164, -0.38489833, 1. ]]))"
},
"metadata": {}
}
]
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2019-06-09T15:25:44.067942Z",
"end_time": "2019-06-09T15:25:57.008309Z"
},
"trusted": true,
"scrolled": false
},
"cell_type": "code",
"source": "%matplotlib inline\n\nimport matplotlib.pyplot as plt\nimport matplotlib.gridspec as gridspec\n\ngs = gridspec.GridSpec(2, 2)\nfig = plt.figure(figsize=(10, 10))\n\nd = stats.multivariate_normal(mean=(0,0), cov=[[1,0],[0,1]]).rvs(10000)\n\ncluster_list = [2, 3, 4, 5]\nfor idx in sp.arange(len(cluster_list)):\n n, cluster_names, c_d = k_means(d, cluster_n=cluster_list[idx])\n\n # print(\"c_d: \", c_d[c_d[:,-1]==0][:,0])\n ax = fig.add_subplot(gs[int(idx/2), int(idx%2)])\n for c in cluster_names:\n ax.scatter(c_d[c_d[:,-1]==c][:,0], c_d[c_d[:,-1]==c][:,1], label=f\"cluster{c}\")\n\n ax.set_aspect(\"equal\")\n ax.set_xlim(-6, 6)\n ax.set_ylim(-6, 6)\n ax.legend()\n\nplt.tight_layout()\nplt.savefig(\"k-means.png\")\nplt.show()",
"execution_count": 66,
"outputs": [
{
"output_type": "display_data",
"data": {
"text/plain": "<Figure size 720x720 with 4 Axes>",
"image/png": "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\n"
},
"metadata": {
"needs_background": "light"
}
}
]
},
{
"metadata": {
"ExecuteTime": {
"start_time": "2019-06-09T08:08:18.846057Z",
"end_time": "2019-06-09T08:08:18.856003Z"
},
"trusted": true,
"scrolled": true
},
"cell_type": "code",
"source": "",
"execution_count": null,
"outputs": []
},
{
"metadata": {
"trusted": true
},
"cell_type": "code",
"source": "",
"execution_count": null,
"outputs": []
}
],
"metadata": {
"kernelspec": {
"name": "python3",
"display_name": "Python 3",
"language": "python"
},
"hide_input": false,
"toc": {
"nav_menu": {},
"number_sections": true,
"sideBar": true,
"skip_h1_title": false,
"base_numbering": 1,
"title_cell": "Table of Contents",
"title_sidebar": "Contents",
"toc_cell": false,
"toc_position": {},
"toc_section_display": true,
"toc_window_display": true
},
"language_info": {
"name": "python",
"version": "3.7.0",
"mimetype": "text/x-python",
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"pygments_lexer": "ipython3",
"nbconvert_exporter": "python",
"file_extension": ".py"
},
"gist": {
"id": "9455bb73c3f8137880ec66b61e8c4615",
"data": {
"description": "k-means clustering",
"public": true
}
},
"_draft": {
"nbviewer_url": "https://gist.github.com/9455bb73c3f8137880ec66b61e8c4615"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.