AlphaWave Data Machine Learning used to build a Diversified Portfolio with K-Means Clustering example.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Machine Learning used to build a Diversified Portfolio: K-Means Clustering\n",
    "\n",
    "**Hugh Donnelly, CFA**<br> \n",
    "*AlphaWave Data*\n",
    "\n",
    "**September 2021**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "<p>\n",
    "    In this article, we will explore K-Means Clustering:<br>\n",
    "    <ul>\n",
    "        <li><a href=\"#definition\"> What is K-Means Clustering? </li>\n",
    "        <li><a href=\"#algorithm\">Algorithm</a></li>\n",
    "        <li><a href=\"#application\">K-Means Clustering Application: Building a diversified portfolio</a></li>\n",
    "    </ul>\n",
    "</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Jupyter Notebooks are available on [Google Colab](https://colab.research.google.com/drive/1m5zWDrHGxEEnJoxNKQn4VXCuRK54jmj1?usp=sharing) and [Github](https://github.com/AlphaWaveData/Jupyter-Notebooks/blob/master/AlphaWave%20Data%20Machine%20Learning%20used%20to%20build%20a%20Diversified%20Portfolio%20with%20K-Means%20Clustering%20example.ipynb).\n",
    "\n",
    "For this project, we use several Python-based scientific computing technologies listed below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/Hugh/opt/anaconda3/lib/python3.7/site-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "import kneed\n",
    "import requests\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from tqdm import tqdm\n",
    "import seaborn as sns\n",
    "import ipywidgets as widgets \n",
    "from scipy.stats import mstats\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.cluster import KMeans\n",
    "from datetime import datetime, timedelta\n",
    "from requests.adapters import HTTPAdapter\n",
    "from requests.exceptions import ConnectionError\n",
    "from requests.packages.urllib3.util.retry import Retry"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a name=\"definition\"></a>\n",
    "<h2><span style=\"color:black\">What is K-Means Clustering?  </span></h2>\n",
    "<p>\n",
    "K-Means Clustering is a form of unsupervised <a href=\"https://hdonnelly6.medium.com/list/machine-learning-for-investing-7f2690bb1826\">machine learning</a> (ML). It is considered to be one of the simplest and most popular unsupervised machine learning techniques.\n",
    "Unsupervised algorithms use vectors on data points. These data points are not labeled or classified. Our goal is to discover hidden patterns and group the data points in a sensible way based on similarity of features. Each group of data points is a cluster and each cluster will have a center."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"img/k_means.png\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a href=\"https://www.analyticsvidhya.com/blog/2021/04/k-means-clustering-simplified-in-python/\">Source : Analytics Vidhya </a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Examples"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's imagine you have two dimensional data that is not labeled as shown below and you are asked to form clusters."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Row data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"img/k_means_row_data.PNG\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below is an example of good clustering."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Good Clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"img/k_means_good_clustering.PNG\" >"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, clustering can go wrong as seen below."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Naive Clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"img/k_means_bad_clustering.PNG\" >"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Source: Oreilly.com : Clustering and Unsupervised Learning"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a name=\"algorithm\"></a>\n",
    "<h2><span style=\"color:black\">Algorithm </span></h2>\n",
    "\n",
    "* Pre-process the data (Clean it, Scale it, Standardize it)\n",
    "* Select K\n",
    "* Pick K Centers \n",
    "* Repeat until there is no change of the centroid positions: <BR>\n",
    "   1) Compute the distance between data point (vector x) and all centroids. (Generally, we use the euclidean distance) <img src=\"img/k_means_euclidean.png\" >\n",
    "   2) Assign each data point to the closest cluster (centroid) <img src=\"img/k_means_argmin.png\" >\n",
    "   3) Compute the centroids for the clusters by taking the average of all data points that belong to each cluster.\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " <img src=\"img/k_means_algorithm.PNG\"  width=\"40%\" >"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Training examples are shown as dots, and cluster centroids are shown as crosses.\n",
    "    <ul>\n",
    "        (a) Original dataset.<br>\n",
    "        (b) Random initial cluster centroids.<br>\n",
    "        (c-f) Illustration of running two iterations of k-means.<br>\n",
    "    </ul>\n",
    "In each iteration, we assign each training example to the closest cluster centroid (shown by \"painting\" the training examples the same color as the cluster centroid to which it is assigned). Then we move each cluster centroid to the mean of the points assigned to it.\n",
    "\n",
    "<a href=\"https://stanford.edu/~cpiech/cs221/handouts/kmeans.html\">Source : Stanford Edu ( K-Means) </a>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a name=\"application\"></a>\n",
    "<h2><span style=\"color:black\">K-Means Clustering Application: Building a diversified portfolio </span></h2>\n",
    "\n",
    "We are going to use K-Means Clustering to build a diversified portfolio. Two ratios will be used in order to cluster the data: <BR>\n",
    "<ul>\n",
    "    <li> <code>Revenue per share:</code> Amount of sales or revenues generated per average total shares outstanding. (Sales Revenue/Average Total Shares)</li>\n",
    "    <li><code>Return on Assets:</code> Indicator of how profitable company is relative to its assets (Total Income /Total Assets)</li>\n",
    "\n",
    "</ul>\n",
    "The idea is to create clusters with similar characteristics for the components of the S&P 500 using these two factors at the end of the 2021 Q1. From each cluster, we will take the stocks with highest risk adjusted momentum to build our portfolio. <br><br>\n",
    "   After building this portfolio, we will run it for 2021 Q2 and compare it to the return of the S&P 500.<br><br>\n",
    "    \n",
    "Please note that this analysis is done using only two factors which leads to a two dimensional problem. We are using a two dimensional problem to demonstrate the concept and understand the problem. Multiple factors can be used as well. If you want to use multiple factors, you may want to use <a href=\"https://scikit-learn.org/stable/modules/generated/sklearn.decomposition.PCA.html\">Principal Component Analysis</a> to lower the number of dimensions. <br><br>\n",
    "   \n",
    "We will proceed with the following steps: <br><b>K-Means Clustering : </b><br>\n",
    "\n",
    "    1. Get the data: Revenue per share and Return on Assets for the end of 2021 Q1 for members of the S&P 500.\n",
    "    2. Analyze the data, clean it and visualize it.\n",
    "    3. Choose K.\n",
    "    4. Analyze the clustering results.\n",
    "<b>Portfolio Construction : </b><br>\n",
    "    \n",
    "    1. From each cluster, choose the stocks with the highest risk adjusted momentum. \n",
    "    2. Run the portfolio return for 2021-Q2."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "Let's apply the steps defined above:\n",
    "## K-Means Clustering \n",
    "### <I>1. Get the data: Revenue per share and Return on Assets for the end of 2021 Q1 for members of the S&P 500</I>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scrape the S&P 500 tickers from Wikipedia\n",
    "\n",
    "def get_tickers():\n",
    "    wiki_page = requests.get('https://en.wikipedia.org/wiki/List_of_S%26P_500_companies').text\n",
    "    sp_data = pd.read_html(wiki_page)\n",
    "    ticker_df = sp_data[0]\n",
    "    ticker_options = ticker_df['Symbol']\n",
    "    return ticker_options"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "505\n"
     ]
    }
   ],
   "source": [
    "# Run the ticker scrape function\n",
    "# Let's convert the get_tickers() output to a list and \n",
    "# replace tickers that have '.' with '-' so we can use AlphaWave Data APIs\n",
    "\n",
    "stock_tickers = get_tickers()\n",
    "stock_tickers = stock_tickers.to_list()\n",
    "for ticker in range(len(stock_tickers)):\n",
    "    stock_tickers[ticker] = stock_tickers[ticker].upper().replace(\".\", \"-\")\n",
    "\n",
    "print (len(stock_tickers))\n",
    "# stock_tickers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use the [Key Statistics](https://rapidapi.com/alphawave/api/stock-analysis?endpoint=apiendpoint_dff4b882-4be4-4169-a700-04275c92bdce) endpoint from the [AlphaWave Data Stock Analysis API](https://rapidapi.com/alphawave/api/stock-analysis/endpoints) to pull in the required stock information.\n",
    "\n",
    "To call this API with Python, you can choose one of the supported Python code snippets provided in the API console. The following is an example of how to invoke the API with Python Requests. You will need to insert your own <b>x-rapidapi-host</b> and <b>x-rapidapi-key</b> information in the code block below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Retrieving AlphaWave Data Stock Info: 100%|██████████| 505/505 [37:55<00:00,  4.51s/it]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MMM</th>\n",
       "      <td>10.93%</td>\n",
       "      <td>59.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABT</th>\n",
       "      <td>7.19%</td>\n",
       "      <td>22.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABBV</th>\n",
       "      <td>7.94%</td>\n",
       "      <td>30.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABMD</th>\n",
       "      <td>12.24%</td>\n",
       "      <td>20.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ACN</th>\n",
       "      <td>11.67%</td>\n",
       "      <td>75.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>YUM</th>\n",
       "      <td>20.33%</td>\n",
       "      <td>19.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZBRA</th>\n",
       "      <td>11.41%</td>\n",
       "      <td>96.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZBH</th>\n",
       "      <td>2.39%</td>\n",
       "      <td>34.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZION</th>\n",
       "      <td>1.39%</td>\n",
       "      <td>19.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZTS</th>\n",
       "      <td>12.36%</td>\n",
       "      <td>15.59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>505 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Return on Assets Rev per share\n",
       "MMM            10.93%         59.94\n",
       "ABT             7.19%         22.65\n",
       "ABBV            7.94%         30.36\n",
       "ABMD           12.24%         20.68\n",
       "ACN            11.67%         75.47\n",
       "...               ...           ...\n",
       "YUM            20.33%         19.47\n",
       "ZBRA           11.41%         96.69\n",
       "ZBH             2.39%         34.18\n",
       "ZION            1.39%         19.58\n",
       "ZTS            12.36%         15.59\n",
       "\n",
       "[505 rows x 2 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Fetch AlphaWave Data's fundamental stock information\n",
    "key_stats_url = \"https://stock-analysis.p.rapidapi.com/api/v1/resources/key-stats\"\n",
    "\n",
    "headers = {\n",
    "    'x-rapidapi-host': \"YOUR_X-RAPIDAPI-HOST_WILL_COPY_DIRECTLY_FROM_RAPIDAPI_PYTHON_CODE_SNIPPETS\",\n",
    "    'x-rapidapi-key': \"YOUR_X-RAPIDAPI-KEY_WILL_COPY_DIRECTLY_FROM_RAPIDAPI_PYTHON_CODE_SNIPPETS\"\n",
    "    }\n",
    "\n",
    "retry_strategy = Retry(total=3, backoff_factor=10, status_forcelist=[429, 500, 502, 503, 504], method_whitelist=[\"HEAD\", \"GET\", \"PUT\", \"DELETE\", \"OPTIONS\", \"TRACE\"])\n",
    "rapid_api_adapter = HTTPAdapter(max_retries=retry_strategy)\n",
    "http = requests.Session()\n",
    "http.mount(\"https://\", rapid_api_adapter)\n",
    "\n",
    "alphawave_data = []\n",
    "\n",
    "for ticker in tqdm(stock_tickers, position=0, leave=True, desc = \"Retrieving AlphaWave Data Stock Info\"):\n",
    "    \n",
    "    querystring = {\"ticker\":ticker}\n",
    "    time.sleep(3)\n",
    "    \n",
    "    try:\n",
    "        \n",
    "        # Get Key Stats\n",
    "        key_stats_response = http.get(key_stats_url, headers=key_stats_headers, params=querystring, timeout=(5, 5))\n",
    "        key_stats_response.raise_for_status()\n",
    "        key_stats_df = pd.DataFrame.from_dict(key_stats_response.json())\n",
    "        key_stats_df = key_stats_df.transpose()\n",
    "\n",
    "        roa = key_stats_df.loc[r'Return on assets (ttm)'][0]\n",
    "        rev_per_share = key_stats_df.loc[r'Revenue per share (ttm)'][0]\n",
    "\n",
    "        # Create Dataframe\n",
    "        df = pd.DataFrame({'Return on Assets': roa,\n",
    "                           'Rev per share': rev_per_share},\n",
    "                          index=[ticker])\n",
    "\n",
    "        alphawave_data.append(df)\n",
    "\n",
    "    except requests.exceptions.RequestException as err:\n",
    "        print (\"OOps: Something Else\",err)\n",
    "    except requests.exceptions.HTTPError as errh:\n",
    "        print (\"Http Error:\",errh)\n",
    "    except requests.exceptions.ConnectionError as errc:\n",
    "        print (\"Error Connecting:\",errc)\n",
    "    except requests.exceptions.Timeout as errt:\n",
    "        print (\"Timeout Error:\",errt)\n",
    "        \n",
    "    except:\n",
    "        pass\n",
    "\n",
    "data = pd.concat(alphawave_data, ignore_index=False)\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MMM</th>\n",
       "      <td>0.1093</td>\n",
       "      <td>59.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABT</th>\n",
       "      <td>0.0719</td>\n",
       "      <td>22.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABBV</th>\n",
       "      <td>0.0794</td>\n",
       "      <td>30.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABMD</th>\n",
       "      <td>0.1224</td>\n",
       "      <td>20.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ACN</th>\n",
       "      <td>0.1167</td>\n",
       "      <td>75.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>YUM</th>\n",
       "      <td>0.2033</td>\n",
       "      <td>19.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZBRA</th>\n",
       "      <td>0.1141</td>\n",
       "      <td>96.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZBH</th>\n",
       "      <td>0.0239</td>\n",
       "      <td>34.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZION</th>\n",
       "      <td>0.0139</td>\n",
       "      <td>19.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZTS</th>\n",
       "      <td>0.1236</td>\n",
       "      <td>15.59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>505 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Return on Assets  Rev per share\n",
       "ID                                   \n",
       "MMM             0.1093          59.94\n",
       "ABT             0.0719          22.65\n",
       "ABBV            0.0794          30.36\n",
       "ABMD            0.1224          20.68\n",
       "ACN             0.1167          75.47\n",
       "...                ...            ...\n",
       "YUM             0.2033          19.47\n",
       "ZBRA            0.1141          96.69\n",
       "ZBH             0.0239          34.18\n",
       "ZION            0.0139          19.58\n",
       "ZTS             0.1236          15.59\n",
       "\n",
       "[505 rows x 2 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Remove any % characters, change string values to numeric values\n",
    "data[[\"Return on Assets\"]] = data[[\"Return on Assets\"]].apply(lambda x: x.str.replace('[%]','', regex=True))\n",
    "data[[\"Return on Assets\", \n",
    "      \"Rev per share\"]] = data[[\"Return on Assets\", \n",
    "                                   \"Rev per share\"]].apply(pd.to_numeric)\n",
    "data[[\"Return on Assets\"]] = data[[\"Return on Assets\"]].apply(lambda x: x/100)\n",
    "data.index.name = 'ID'\n",
    "\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <I>2. Analyze the data, clean it and visualize it.</I>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(505, 2)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>502.000000</td>\n",
       "      <td>505.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.062860</td>\n",
       "      <td>400.912376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.055882</td>\n",
       "      <td>7499.921615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.129500</td>\n",
       "      <td>0.070000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.026150</td>\n",
       "      <td>18.890000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.052000</td>\n",
       "      <td>32.020000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.087475</td>\n",
       "      <td>64.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.296500</td>\n",
       "      <td>168571.660000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Return on Assets  Rev per share\n",
       "count        502.000000     505.000000\n",
       "mean           0.062860     400.912376\n",
       "std            0.055882    7499.921615\n",
       "min           -0.129500       0.070000\n",
       "25%            0.026150      18.890000\n",
       "50%            0.052000      32.020000\n",
       "75%            0.087475      64.800000\n",
       "max            0.296500  168571.660000"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make a copy of the original data before starting our data preprocessing\n",
    "original_data=data.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>GL</th>\n",
       "      <td>NaN</td>\n",
       "      <td>47.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OGN</th>\n",
       "      <td>NaN</td>\n",
       "      <td>31.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WRB</th>\n",
       "      <td>NaN</td>\n",
       "      <td>47.48</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Return on Assets  Rev per share\n",
       "ID                                  \n",
       "GL                NaN          47.56\n",
       "OGN               NaN          31.13\n",
       "WRB               NaN          47.48"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Check Na Values\n",
    "data[data['Return on Assets'].isna() | data['Rev per share'].isna()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MMM</th>\n",
       "      <td>0.1093</td>\n",
       "      <td>59.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABT</th>\n",
       "      <td>0.0719</td>\n",
       "      <td>22.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABBV</th>\n",
       "      <td>0.0794</td>\n",
       "      <td>30.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABMD</th>\n",
       "      <td>0.1224</td>\n",
       "      <td>20.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ACN</th>\n",
       "      <td>0.1167</td>\n",
       "      <td>75.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>YUM</th>\n",
       "      <td>0.2033</td>\n",
       "      <td>19.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZBRA</th>\n",
       "      <td>0.1141</td>\n",
       "      <td>96.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZBH</th>\n",
       "      <td>0.0239</td>\n",
       "      <td>34.18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZION</th>\n",
       "      <td>0.0139</td>\n",
       "      <td>19.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ZTS</th>\n",
       "      <td>0.1236</td>\n",
       "      <td>15.59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>502 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Return on Assets  Rev per share\n",
       "ID                                   \n",
       "MMM             0.1093          59.94\n",
       "ABT             0.0719          22.65\n",
       "ABBV            0.0794          30.36\n",
       "ABMD            0.1224          20.68\n",
       "ACN             0.1167          75.47\n",
       "...                ...            ...\n",
       "YUM             0.2033          19.47\n",
       "ZBRA            0.1141          96.69\n",
       "ZBH             0.0239          34.18\n",
       "ZION            0.0139          19.58\n",
       "ZTS             0.1236          15.59\n",
       "\n",
       "[502 rows x 2 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Dropna value \n",
    "data=data.dropna()\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1c28d9a910>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize scatterplot\n",
    "plt.style.use(\"dark_background\")\n",
    "g = sns.scatterplot(x='Return on Assets', y='Rev per share', data=data)\n",
    "plt.ylim([0,200])\n",
    "plt.title(\"Original Data\")\n",
    "\n",
    "# Some random point we want to classify\n",
    "plt.scatter(0.05, 50, marker='o', s=80, color='red')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MMM</th>\n",
       "      <td>0.1093</td>\n",
       "      <td>59.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABT</th>\n",
       "      <td>0.0719</td>\n",
       "      <td>22.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABBV</th>\n",
       "      <td>0.0794</td>\n",
       "      <td>30.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ABMD</th>\n",
       "      <td>0.1224</td>\n",
       "      <td>20.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ACN</th>\n",
       "      <td>0.1167</td>\n",
       "      <td>75.47</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Return on Assets  Rev per share\n",
       "ID                                   \n",
       "MMM             0.1093          59.94\n",
       "ABT             0.0719          22.65\n",
       "ABBV            0.0794          30.36\n",
       "ABMD            0.1224          20.68\n",
       "ACN             0.1167          75.47"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Both Revenue per share and Return on Assets are ratios. They are already scaled to the company size.\n",
    "# We can use Winsorization to transforms data by limiting extreme values, typically by setting all outliers to a specified percentile of data\n",
    "X =np.asarray([np.asarray(data['Return on Assets']),np.asarray(data['Rev per share'])])\n",
    "X = mstats.winsorize(X, limits = [0.05, 0.05])\n",
    "data=pd.DataFrame(X, index=['Return on Assets','Rev per share'], columns=data.index).T\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize scatterplot\n",
    "plt.style.use(\"dark_background\")\n",
    "g = sns.scatterplot(x='Return on Assets', y='Rev per share', data=data)\n",
    "plt.title(\"Winsorized Data\")\n",
    "\n",
    "# Some random point we want to classify\n",
    "plt.scatter(0.05, 50, marker='o', s=80, color='red')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <I>3. Choose K</I>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The two most common methods to choose K ( the appropriate number of clusters) are :\n",
    "    <ul>\n",
    "        <li>The silhouette Coefficient</li>\n",
    "        <li>The Elbow Method </li>\n",
    "    </ul>\n",
    "\n",
    "The silhouette coefficient is a value that ranges between -1 and 1. It quantifies how well a data point fits into its assigned cluster based on two factors:\n",
    "1. How close the data point is to other points in the cluster\n",
    "2. How far away the data point is from points in other clusters\n",
    "\n",
    "Larger numbers for Silhouette coefficient indicate that samples are closer to their clusters than they are to other clusters.\n",
    "\n",
    "The elbow method is used by running several k-means, increment k with each iteration, and record the SSE ( Sum Of Squared Error) <br><br>\n",
    "$$SSE= Sum  \\; Of  \\; Euclidean  \\; Squared  \\; Distances  \\; of  \\; each  \\; point \\; to \\; its  \\; closest \\; centroid $$<br>\n",
    "After that , we plot SSE as a function of the number of clusters. SSE continues to decrease as you increase k. As more centroids are added, the distance from each point to its closest centroid will decrease.\n",
    "There’s a sweet spot where the SSE curve starts to bend known as the elbow point. The x-value of this point is thought to be a reasonable trade-off between error and number of clusters. <br>\n",
    "\n",
    "<a href=\"https://realpython.com/k-means-clustering-python/#choosing-the-appropriate-number-of-clusters\"> (Source)</a>\n",
    "\n",
    "In this example, we will use the Elbow Method to determine K:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "distorsions = []\n",
    "clusters_iterations=range(2, 20)\n",
    "for k in clusters_iterations:\n",
    "    k_means = KMeans(n_clusters=k)\n",
    "    k_means.fit(data)\n",
    "    distorsions.append(k_means.inertia_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SSE</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cluster</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>168519.799163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>67633.824231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>37234.562646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>23070.887920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>15687.451817</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   SSE\n",
       "Cluster               \n",
       "2        168519.799163\n",
       "3         67633.824231\n",
       "4         37234.562646\n",
       "5         23070.887920\n",
       "6         15687.451817"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "elbow_curve_data=pd.DataFrame(zip(clusters_iterations,distorsions),columns=['Cluster','SSE']).set_index('Cluster')\n",
    "elbow_curve_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize plot\n",
    "plt.figure(figsize=(11,7))\n",
    "plt.style.use(\"dark_background\")\n",
    "plt.plot(elbow_curve_data['SSE'])\n",
    "plt.title(\"Elbow Curve\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Elbow = 5\n"
     ]
    }
   ],
   "source": [
    "# get elbow programmatically\n",
    "from kneed import KneeLocator \n",
    "kl = KneeLocator(\n",
    "clusters_iterations, distorsions, curve=\"convex\", direction=\"decreasing\")\n",
    "elbow=kl.elbow\n",
    "\n",
    "print('Elbow = {}'.format(elbow))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <I>4. Analyze the clustering results</I>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We apply KMeans for the Elbow's value  ( in this case = 5)\n",
    "kmeans = KMeans(n_clusters=elbow)\n",
    "kmeans.fit(data)\n",
    "y_kmeans = kmeans.predict(data)\n",
    "df_kmeans = data.copy()\n",
    "df_kmeans['cluster']=y_kmeans.astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize the results\n",
    "plt.style.use(\"dark_background\")\n",
    "g = sns.scatterplot(x='Return on Assets', y='Rev per share', hue=df_kmeans['cluster'].astype(int), \n",
    "                    palette=['blue','green','yellow','orange','red'], data=df_kmeans)\n",
    "plt.title(\"K-Means Clustering\")\n",
    "\n",
    "# Some random point we want to classify\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.057726</td>\n",
       "      <td>63.871077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.076280</td>\n",
       "      <td>133.355625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.060268</td>\n",
       "      <td>15.883474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.076032</td>\n",
       "      <td>97.086129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.070734</td>\n",
       "      <td>36.405891</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Return on Assets  Rev per share\n",
       "0          0.057726      63.871077\n",
       "1          0.076280     133.355625\n",
       "2          0.060268      15.883474\n",
       "3          0.076032      97.086129\n",
       "4          0.070734      36.405891"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# see the centers \n",
    "clusters_centers_df=pd.DataFrame(kmeans.cluster_centers_,columns=['Return on Assets','Rev per share'])\n",
    "clusters_centers_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Company</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cluster</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MMM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ABT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ABBV</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ABMD</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ACN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Company\n",
       "Cluster        \n",
       "0           MMM\n",
       "2           ABT\n",
       "4          ABBV\n",
       "2          ABMD\n",
       "0           ACN"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# See the clustering by Company \n",
    "clustering_result=pd.DataFrame(zip(y_kmeans,data.index),columns=['Cluster','Company'])\n",
    "clustering_result.set_index('Cluster').head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        Company\n",
      "Cluster        \n",
      "0           MMM\n",
      "0           ACN\n",
      "0           AIG\n",
      "0           AON\n",
      "0          APTV\n",
      "        Company\n",
      "Cluster        \n",
      "2           ABT\n",
      "2          ABMD\n",
      "2          ATVI\n",
      "2           AMD\n",
      "2           AES\n",
      "        Company\n",
      "Cluster        \n",
      "4          ABBV\n",
      "4          ADBE\n",
      "4           AFL\n",
      "4           APD\n",
      "4           ALK\n",
      "        Company\n",
      "Cluster        \n",
      "1           AAP\n",
      "1           ALL\n",
      "1         GOOGL\n",
      "1          GOOG\n",
      "1          AMZN\n",
      "        Company\n",
      "Cluster        \n",
      "3           AMP\n",
      "3           AVY\n",
      "3           BIO\n",
      "3            BA\n",
      "3           CAT\n"
     ]
    }
   ],
   "source": [
    "for cluster_num in list(clustering_result.set_index('Cluster').index.unique()):\n",
    "    print (clustering_result.set_index('Cluster').loc[cluster_num].head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Rev per share</th>\n",
       "      <th>Count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.057726</td>\n",
       "      <td>63.871077</td>\n",
       "      <td>65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.076280</td>\n",
       "      <td>133.355625</td>\n",
       "      <td>64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.060268</td>\n",
       "      <td>15.883474</td>\n",
       "      <td>213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.076032</td>\n",
       "      <td>97.086129</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.070734</td>\n",
       "      <td>36.405891</td>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Return on Assets  Rev per share  Count\n",
       "0          0.057726      63.871077     65\n",
       "1          0.076280     133.355625     64\n",
       "2          0.060268      15.883474    213\n",
       "3          0.076032      97.086129     31\n",
       "4          0.070734      36.405891    129"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Enrich Centers Df with the number of elements by Cluster\n",
    "clusters_centers_df['Count']=clustering_result['Cluster'].value_counts().to_frame().rename(columns={'Cluster':'Count'})['Count']\n",
    "clusters_centers_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 792x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize Count of Elements by Cluster \n",
    "plt.figure(figsize=(11,7))\n",
    "plt.style.use(\"dark_background\")\n",
    "plt.bar(clusters_centers_df.index.values,clusters_centers_df['Count'])\n",
    "plt.title(\"Count of Elements by Cluster\")\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Portfolio Construction\n",
    "### <I>1. From each cluster, choose the stocks with the highest Risk Adjusted Momentum </I>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can use the [2 Year Historical Daily Prices](https://rapidapi.com/alphawave/api/stock-prices2?endpoint=apiendpoint_33fa1878-1727-4775-beeb-f6b0da5314fd) endpoint from the [AlphaWave Data Stock Prices API](https://rapidapi.com/alphawave/api/stock-prices2/endpoints) to pull in the two year historical prices.\n",
    "\n",
    "To call this API with Python, you can choose one of the supported Python code snippets provided in the API console. The following is an example of how to invoke the API with Python Requests. You will need to insert your own <b>x-rapidapi-host</b> and <b>x-rapidapi-key</b> information in the code block below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Retrieving AlphaWave Data Stock Info: 100%|██████████| 505/505 [10:35<00:00,  1.26s/it]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>MMM</th>\n",
       "      <th>ABT</th>\n",
       "      <th>ABBV</th>\n",
       "      <th>ABMD</th>\n",
       "      <th>ACN</th>\n",
       "      <th>ATVI</th>\n",
       "      <th>ADBE</th>\n",
       "      <th>AMD</th>\n",
       "      <th>AAP</th>\n",
       "      <th>AES</th>\n",
       "      <th>...</th>\n",
       "      <th>WLTW</th>\n",
       "      <th>WYNN</th>\n",
       "      <th>XEL</th>\n",
       "      <th>XLNX</th>\n",
       "      <th>XYL</th>\n",
       "      <th>YUM</th>\n",
       "      <th>ZBRA</th>\n",
       "      <th>ZBH</th>\n",
       "      <th>ZION</th>\n",
       "      <th>ZTS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2019-09-09</th>\n",
       "      <td>154.981232</td>\n",
       "      <td>81.63871</td>\n",
       "      <td>59.288097</td>\n",
       "      <td>186.809998</td>\n",
       "      <td>189.160187</td>\n",
       "      <td>54.565197</td>\n",
       "      <td>279.299988</td>\n",
       "      <td>30.5</td>\n",
       "      <td>144.989563</td>\n",
       "      <td>14.154182</td>\n",
       "      <td>...</td>\n",
       "      <td>190.402863</td>\n",
       "      <td>112.754280</td>\n",
       "      <td>60.126400</td>\n",
       "      <td>103.516769</td>\n",
       "      <td>75.715034</td>\n",
       "      <td>113.642380</td>\n",
       "      <td>196.809998</td>\n",
       "      <td>135.137802</td>\n",
       "      <td>39.775383</td>\n",
       "      <td>125.237442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-09-10</th>\n",
       "      <td>157.708389</td>\n",
       "      <td>80.65966</td>\n",
       "      <td>60.603638</td>\n",
       "      <td>191.820007</td>\n",
       "      <td>184.14064</td>\n",
       "      <td>55.395687</td>\n",
       "      <td>276.140015</td>\n",
       "      <td>30.23</td>\n",
       "      <td>152.139206</td>\n",
       "      <td>14.295442</td>\n",
       "      <td>...</td>\n",
       "      <td>184.716019</td>\n",
       "      <td>114.957245</td>\n",
       "      <td>59.746990</td>\n",
       "      <td>104.703911</td>\n",
       "      <td>76.545448</td>\n",
       "      <td>111.192238</td>\n",
       "      <td>192.639999</td>\n",
       "      <td>136.557404</td>\n",
       "      <td>40.986385</td>\n",
       "      <td>121.648193</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-09-11</th>\n",
       "      <td>156.979919</td>\n",
       "      <td>81.929497</td>\n",
       "      <td>61.76807</td>\n",
       "      <td>191.070007</td>\n",
       "      <td>185.800858</td>\n",
       "      <td>54.288364</td>\n",
       "      <td>277.779999</td>\n",
       "      <td>29.76</td>\n",
       "      <td>156.230331</td>\n",
       "      <td>14.643881</td>\n",
       "      <td>...</td>\n",
       "      <td>186.712265</td>\n",
       "      <td>112.429741</td>\n",
       "      <td>59.993603</td>\n",
       "      <td>103.065460</td>\n",
       "      <td>77.463799</td>\n",
       "      <td>109.774239</td>\n",
       "      <td>197.419998</td>\n",
       "      <td>138.302322</td>\n",
       "      <td>41.446384</td>\n",
       "      <td>119.344376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-09-12</th>\n",
       "      <td>159.314804</td>\n",
       "      <td>82.598351</td>\n",
       "      <td>62.959167</td>\n",
       "      <td>188.020004</td>\n",
       "      <td>189.830093</td>\n",
       "      <td>54.822258</td>\n",
       "      <td>278.019989</td>\n",
       "      <td>30.209999</td>\n",
       "      <td>152.630905</td>\n",
       "      <td>14.756888</td>\n",
       "      <td>...</td>\n",
       "      <td>189.584900</td>\n",
       "      <td>115.144096</td>\n",
       "      <td>60.208393</td>\n",
       "      <td>103.045822</td>\n",
       "      <td>76.271904</td>\n",
       "      <td>109.262985</td>\n",
       "      <td>197.279999</td>\n",
       "      <td>140.589462</td>\n",
       "      <td>41.896992</td>\n",
       "      <td>120.323257</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019-09-13</th>\n",
       "      <td>160.118011</td>\n",
       "      <td>81.687164</td>\n",
       "      <td>62.674721</td>\n",
       "      <td>186.410004</td>\n",
       "      <td>188.956284</td>\n",
       "      <td>55.316593</td>\n",
       "      <td>278.350006</td>\n",
       "      <td>30.690001</td>\n",
       "      <td>154.401108</td>\n",
       "      <td>14.869897</td>\n",
       "      <td>...</td>\n",
       "      <td>189.594635</td>\n",
       "      <td>115.517815</td>\n",
       "      <td>59.769260</td>\n",
       "      <td>103.251862</td>\n",
       "      <td>76.858070</td>\n",
       "      <td>107.420563</td>\n",
       "      <td>205.160004</td>\n",
       "      <td>136.823578</td>\n",
       "      <td>42.028416</td>\n",
       "      <td>119.660774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-09-01</th>\n",
       "      <td>193.440002</td>\n",
       "      <td>127.400002</td>\n",
       "      <td>112.269997</td>\n",
       "      <td>363.269989</td>\n",
       "      <td>337.899994</td>\n",
       "      <td>82.300003</td>\n",
       "      <td>665.890015</td>\n",
       "      <td>109.989998</td>\n",
       "      <td>201.880005</td>\n",
       "      <td>24.120001</td>\n",
       "      <td>...</td>\n",
       "      <td>223.699997</td>\n",
       "      <td>100.980003</td>\n",
       "      <td>69.750000</td>\n",
       "      <td>154.470001</td>\n",
       "      <td>136.550003</td>\n",
       "      <td>131.550003</td>\n",
       "      <td>582.780029</td>\n",
       "      <td>150.869995</td>\n",
       "      <td>57.169998</td>\n",
       "      <td>206.440002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-09-02</th>\n",
       "      <td>194.889999</td>\n",
       "      <td>128.490005</td>\n",
       "      <td>112.019997</td>\n",
       "      <td>369.829987</td>\n",
       "      <td>341.0</td>\n",
       "      <td>81.309998</td>\n",
       "      <td>664.640015</td>\n",
       "      <td>109.199997</td>\n",
       "      <td>200.809998</td>\n",
       "      <td>24.530001</td>\n",
       "      <td>...</td>\n",
       "      <td>230.059998</td>\n",
       "      <td>102.000000</td>\n",
       "      <td>70.320000</td>\n",
       "      <td>154.139999</td>\n",
       "      <td>138.029999</td>\n",
       "      <td>131.270004</td>\n",
       "      <td>588.250000</td>\n",
       "      <td>149.089996</td>\n",
       "      <td>57.150002</td>\n",
       "      <td>209.220001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-09-03</th>\n",
       "      <td>194.389999</td>\n",
       "      <td>128.940002</td>\n",
       "      <td>111.620003</td>\n",
       "      <td>372.690002</td>\n",
       "      <td>343.329987</td>\n",
       "      <td>81.18</td>\n",
       "      <td>666.590027</td>\n",
       "      <td>109.919998</td>\n",
       "      <td>199.440002</td>\n",
       "      <td>24.639999</td>\n",
       "      <td>...</td>\n",
       "      <td>228.520004</td>\n",
       "      <td>100.300003</td>\n",
       "      <td>69.800003</td>\n",
       "      <td>155.960007</td>\n",
       "      <td>137.250000</td>\n",
       "      <td>130.919998</td>\n",
       "      <td>590.599976</td>\n",
       "      <td>146.740005</td>\n",
       "      <td>57.119999</td>\n",
       "      <td>209.690002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-09-07</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>...</td>\n",
       "      <td>227.100006</td>\n",
       "      <td>105.930000</td>\n",
       "      <td>67.900002</td>\n",
       "      <td>154.750000</td>\n",
       "      <td>135.649994</td>\n",
       "      <td>128.830002</td>\n",
       "      <td>581.140015</td>\n",
       "      <td>146.619995</td>\n",
       "      <td>57.500000</td>\n",
       "      <td>207.809998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-09-08</th>\n",
       "      <td>187.889999</td>\n",
       "      <td>129.059998</td>\n",
       "      <td>110.360001</td>\n",
       "      <td>374.790009</td>\n",
       "      <td>340.320007</td>\n",
       "      <td>80.290001</td>\n",
       "      <td>663.219971</td>\n",
       "      <td>106.169998</td>\n",
       "      <td>200.179993</td>\n",
       "      <td>24.15</td>\n",
       "      <td>...</td>\n",
       "      <td>225.949997</td>\n",
       "      <td>104.620003</td>\n",
       "      <td>69.309998</td>\n",
       "      <td>153.309998</td>\n",
       "      <td>135.110001</td>\n",
       "      <td>130.190002</td>\n",
       "      <td>579.750000</td>\n",
       "      <td>147.479996</td>\n",
       "      <td>56.500000</td>\n",
       "      <td>209.330002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>505 rows × 505 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   MMM         ABT        ABBV        ABMD         ACN  \\\n",
       "2019-09-09  154.981232    81.63871   59.288097  186.809998  189.160187   \n",
       "2019-09-10  157.708389    80.65966   60.603638  191.820007   184.14064   \n",
       "2019-09-11  156.979919   81.929497    61.76807  191.070007  185.800858   \n",
       "2019-09-12  159.314804   82.598351   62.959167  188.020004  189.830093   \n",
       "2019-09-13  160.118011   81.687164   62.674721  186.410004  188.956284   \n",
       "...                ...         ...         ...         ...         ...   \n",
       "2021-09-01  193.440002  127.400002  112.269997  363.269989  337.899994   \n",
       "2021-09-02  194.889999  128.490005  112.019997  369.829987       341.0   \n",
       "2021-09-03  194.389999  128.940002  111.620003  372.690002  343.329987   \n",
       "2021-09-07                                                               \n",
       "2021-09-08  187.889999  129.059998  110.360001  374.790009  340.320007   \n",
       "\n",
       "                 ATVI        ADBE         AMD         AAP        AES  ...  \\\n",
       "2019-09-09  54.565197  279.299988        30.5  144.989563  14.154182  ...   \n",
       "2019-09-10  55.395687  276.140015       30.23  152.139206  14.295442  ...   \n",
       "2019-09-11  54.288364  277.779999       29.76  156.230331  14.643881  ...   \n",
       "2019-09-12  54.822258  278.019989   30.209999  152.630905  14.756888  ...   \n",
       "2019-09-13  55.316593  278.350006   30.690001  154.401108  14.869897  ...   \n",
       "...               ...         ...         ...         ...        ...  ...   \n",
       "2021-09-01  82.300003  665.890015  109.989998  201.880005  24.120001  ...   \n",
       "2021-09-02  81.309998  664.640015  109.199997  200.809998  24.530001  ...   \n",
       "2021-09-03      81.18  666.590027  109.919998  199.440002  24.639999  ...   \n",
       "2021-09-07                                                            ...   \n",
       "2021-09-08  80.290001  663.219971  106.169998  200.179993      24.15  ...   \n",
       "\n",
       "                  WLTW        WYNN        XEL        XLNX         XYL  \\\n",
       "2019-09-09  190.402863  112.754280  60.126400  103.516769   75.715034   \n",
       "2019-09-10  184.716019  114.957245  59.746990  104.703911   76.545448   \n",
       "2019-09-11  186.712265  112.429741  59.993603  103.065460   77.463799   \n",
       "2019-09-12  189.584900  115.144096  60.208393  103.045822   76.271904   \n",
       "2019-09-13  189.594635  115.517815  59.769260  103.251862   76.858070   \n",
       "...                ...         ...        ...         ...         ...   \n",
       "2021-09-01  223.699997  100.980003  69.750000  154.470001  136.550003   \n",
       "2021-09-02  230.059998  102.000000  70.320000  154.139999  138.029999   \n",
       "2021-09-03  228.520004  100.300003  69.800003  155.960007  137.250000   \n",
       "2021-09-07  227.100006  105.930000  67.900002  154.750000  135.649994   \n",
       "2021-09-08  225.949997  104.620003  69.309998  153.309998  135.110001   \n",
       "\n",
       "                   YUM        ZBRA         ZBH       ZION         ZTS  \n",
       "2019-09-09  113.642380  196.809998  135.137802  39.775383  125.237442  \n",
       "2019-09-10  111.192238  192.639999  136.557404  40.986385  121.648193  \n",
       "2019-09-11  109.774239  197.419998  138.302322  41.446384  119.344376  \n",
       "2019-09-12  109.262985  197.279999  140.589462  41.896992  120.323257  \n",
       "2019-09-13  107.420563  205.160004  136.823578  42.028416  119.660774  \n",
       "...                ...         ...         ...        ...         ...  \n",
       "2021-09-01  131.550003  582.780029  150.869995  57.169998  206.440002  \n",
       "2021-09-02  131.270004  588.250000  149.089996  57.150002  209.220001  \n",
       "2021-09-03  130.919998  590.599976  146.740005  57.119999  209.690002  \n",
       "2021-09-07  128.830002  581.140015  146.619995  57.500000  207.809998  \n",
       "2021-09-08  130.190002  579.750000  147.479996  56.500000  209.330002  \n",
       "\n",
       "[505 rows x 505 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#fetch 2 year daily return data\n",
    "\n",
    "url = \"https://stock-prices2.p.rapidapi.com/api/v1/resources/stock-prices/2y\"\n",
    "\n",
    "headers = {\n",
    "    'x-rapidapi-host': \"YOUR_X-RAPIDAPI-HOST_WILL_COPY_DIRECTLY_FROM_RAPIDAPI_PYTHON_CODE_SNIPPETS\",\n",
    "    'x-rapidapi-key': \"YOUR_X-RAPIDAPI-KEY_WILL_COPY_DIRECTLY_FROM_RAPIDAPI_PYTHON_CODE_SNIPPETS\"\n",
    "    }\n",
    "\n",
    "stock_frames = []\n",
    "\n",
    "# for ticker in stock_tickers:\n",
    "for ticker in tqdm(stock_tickers, position=0, leave=True, desc = \"Retrieving AlphaWave Data Stock Info\"):\n",
    "    \n",
    "    querystring = {\"ticker\":ticker}\n",
    "    stock_daily_price_response = requests.request(\"GET\", url, headers=headers, params=querystring)\n",
    "\n",
    "    # Create Stock Prices DataFrame\n",
    "    stock_daily_price_df = pd.DataFrame.from_dict(stock_daily_price_response.json())\n",
    "    stock_daily_price_df = stock_daily_price_df.transpose()\n",
    "    stock_daily_price_df = stock_daily_price_df.rename(columns={'Close':ticker})\n",
    "    stock_daily_price_df = stock_daily_price_df[{ticker}]\n",
    "    stock_frames.append(stock_daily_price_df)\n",
    "\n",
    "combined_stock_price_df = pd.concat(stock_frames, axis=1, sort=True)\n",
    "combined_stock_price_df = combined_stock_price_df.dropna(how='all')\n",
    "combined_stock_price_df = combined_stock_price_df.fillna(\"\")\n",
    "combined_stock_price_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Company</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cluster</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>13.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Company\n",
       "Cluster         \n",
       "0            6.0\n",
       "1            6.0\n",
       "2           21.0\n",
       "3            3.0\n",
       "4           13.0"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Build of Portfolio of 50 stocks\n",
    "number_of_stocks=50\n",
    "\n",
    "# From each Cluster, we will pick the stocks with the highest risk adjusted momentum. The number of stocks from each cluster is proportional to its size\n",
    "# Let's start by calculate the number of stocks to pick from each cluster\n",
    "number_of_stocks_by_cluster=pd.DataFrame(round(number_of_stocks*clustering_result.groupby(by='Cluster').count()['Company']/clustering_result.count()['Company'],0))\n",
    "number_of_stocks_by_cluster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['PCAR',\n",
       " 'PXD',\n",
       " 'PPG',\n",
       " 'RJF',\n",
       " 'ROK',\n",
       " 'TT',\n",
       " 'POOL',\n",
       " 'PVH',\n",
       " 'TGT',\n",
       " 'ULTA',\n",
       " 'URI',\n",
       " 'WHR',\n",
       " 'MPWR',\n",
       " 'NWSA',\n",
       " 'NWS',\n",
       " 'NVDA',\n",
       " 'OKE',\n",
       " 'PYPL',\n",
       " 'PNR',\n",
       " 'PTC',\n",
       " 'RF',\n",
       " 'RCL',\n",
       " 'SLB',\n",
       " 'SYF',\n",
       " 'TPR',\n",
       " 'TER',\n",
       " 'TXN',\n",
       " 'TRMB',\n",
       " 'TWTR',\n",
       " 'UAA',\n",
       " 'UA',\n",
       " 'WY',\n",
       " 'WMB',\n",
       " 'PWR',\n",
       " 'SIVB',\n",
       " 'ZBRA',\n",
       " 'MS',\n",
       " 'MOS',\n",
       " 'NXPI',\n",
       " 'PENN',\n",
       " 'QRVO',\n",
       " 'QCOM',\n",
       " 'RHI',\n",
       " 'SNPS',\n",
       " 'TEL',\n",
       " 'TSLA',\n",
       " 'VIAC',\n",
       " 'DIS',\n",
       " 'WST']"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# From each Cluster, pick the stocks with the highest risk adjusted momentum.\n",
    "as_of_date='2021-03-30'\n",
    "\n",
    "portfolio_stocks=[]\n",
    "for cluster_num in list(number_of_stocks_by_cluster.index):\n",
    "    # for each cluster,get all the companies within this cluster\n",
    "    list_stocks=list(clustering_result.set_index('Cluster').loc[cluster_num]['Company'])\n",
    "    #get the number of stocks that we will pick for our portfolio     \n",
    "    number_stocks=number_of_stocks_by_cluster.loc[cluster_num]['Company']\n",
    "    if number_stocks>0:\n",
    "        # Compute the risk adjusted momentum for the past year\n",
    "        last_year_date=pd.to_datetime(as_of_date)+ pd.offsets.DateOffset(years=-1)\n",
    "        last_month_date=pd.to_datetime(as_of_date)+ pd.tseries.offsets.BusinessDay(offset = timedelta(days = -30))\n",
    "        stock_price_last_year_date = last_year_date.strftime('%Y-%m-%d')\n",
    "        stock_price_last_month_date = last_month_date.strftime('%Y-%m-%d')\n",
    "        \n",
    "        risk_adjusted_mom_frames = []\n",
    "        for ticker in list_stocks:\n",
    "\n",
    "            try:\n",
    "                momentum = (combined_stock_price_df.loc[stock_price_last_month_date,][ticker] - \\\n",
    "                            combined_stock_price_df.loc[stock_price_last_year_date,][ticker]) / \\\n",
    "                            combined_stock_price_df.loc[stock_price_last_year_date,][ticker]\n",
    "\n",
    "                annualized_volatility = np.log(combined_stock_price_df.loc[stock_price_last_year_date:as_of_date,][ticker] / \\\n",
    "                                               combined_stock_price_df.loc[stock_price_last_year_date:as_of_date,][ticker].shift(1)).dropna().std()*252**.5\n",
    "\n",
    "                risk_adjusted_momentum = momentum / annualized_volatility\n",
    "\n",
    "                # Create Dataframe\n",
    "                df = pd.DataFrame({'Risk Adj MoM': risk_adjusted_momentum},\n",
    "                                  index=[ticker])\n",
    "\n",
    "                risk_adjusted_mom_frames.append(df)\n",
    "\n",
    "            except:\n",
    "                pass\n",
    "\n",
    "        risk_adjusted_mom_df = pd.concat(risk_adjusted_mom_frames, ignore_index=False)\n",
    "        risk_adjusted_mom_df[\"Rank\"] = risk_adjusted_mom_df[\"Risk Adj MoM\"].rank(ascending=False)\n",
    "        risk_adjusted_mom_df[[\"Risk Adj MoM\", \n",
    "                              \"Rank\"]] = risk_adjusted_mom_df[[\"Risk Adj MoM\", \n",
    "                                                               \"Rank\"]].apply(pd.to_numeric)\n",
    "        filtered_risk_adjusted_mom_df = risk_adjusted_mom_df[risk_adjusted_mom_df['Rank'] <= number_stocks]\n",
    "        portfolio_stocks=portfolio_stocks+list(filtered_risk_adjusted_mom_df.index)\n",
    "\n",
    "portfolio_stocks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <I> 2. Compute Portfolio's Performance for 2021-Q2 </I>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cluster_portfolio_return</th>\n",
       "      <th>spx_index_return</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DATE</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-03-31</th>\n",
       "      <td>1.009275</td>\n",
       "      <td>0.999547</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-04-01</th>\n",
       "      <td>1.014843</td>\n",
       "      <td>1.010877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-04-05</th>\n",
       "      <td>1.013042</td>\n",
       "      <td>1.009535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-04-06</th>\n",
       "      <td>0.999495</td>\n",
       "      <td>1.001167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-04-07</th>\n",
       "      <td>0.994658</td>\n",
       "      <td>0.996562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-06-24</th>\n",
       "      <td>1.012652</td>\n",
       "      <td>1.006553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-06-25</th>\n",
       "      <td>1.003557</td>\n",
       "      <td>1.007312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-06-28</th>\n",
       "      <td>0.997640</td>\n",
       "      <td>0.995732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-06-29</th>\n",
       "      <td>1.002404</td>\n",
       "      <td>0.999105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-06-30</th>\n",
       "      <td>1.005130</td>\n",
       "      <td>1.002494</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>64 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            cluster_portfolio_return  spx_index_return\n",
       "DATE                                                  \n",
       "2021-03-31                  1.009275          0.999547\n",
       "2021-04-01                  1.014843          1.010877\n",
       "2021-04-05                  1.013042          1.009535\n",
       "2021-04-06                  0.999495          1.001167\n",
       "2021-04-07                  0.994658          0.996562\n",
       "...                              ...               ...\n",
       "2021-06-24                  1.012652          1.006553\n",
       "2021-06-25                  1.003557          1.007312\n",
       "2021-06-28                  0.997640          0.995732\n",
       "2021-06-29                  1.002404          0.999105\n",
       "2021-06-30                  1.005130          1.002494\n",
       "\n",
       "[64 rows x 2 columns]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Since we chose our portfolio stocks by end the of 2021-Q1, we will run it for 2021-Q2 \n",
    "end_date='2021-06-30'\n",
    "\n",
    "# Compute the portfolio return. We will use equal weights for all the stocks\n",
    "combined_stock_price_df = combined_stock_price_df.apply(pd.to_numeric)\n",
    "s_p_500_daily_return = (combined_stock_price_df.loc[as_of_date:end_date,].pct_change().sum(axis=1).dropna()/len(combined_stock_price_df.columns)) + 1\n",
    "cluster_portfolio_return=0\n",
    "for stock in portfolio_stocks:\n",
    "    daily_return = combined_stock_price_df.loc[as_of_date:end_date,][stock].pct_change().dropna() + 1\n",
    "    cluster_portfolio_return=cluster_portfolio_return+(daily_return/len(portfolio_stocks))\n",
    "\n",
    "# Create Dataframe\n",
    "df = pd.DataFrame({'cluster_portfolio_return':cluster_portfolio_return,\n",
    "                   'spx_index_return':s_p_500_daily_return},)\n",
    "\n",
    "df.index.name = 'DATE'\n",
    "return_ptf_index = df.dropna()\n",
    "return_ptf_index = return_ptf_index.apply(pd.to_numeric)\n",
    "\n",
    "return_ptf_index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Portfolio Annual Excess Return : 40.91%\n",
      "Portfolio Annual Volatility    : 17.11% \n",
      "Portfolio Sharpe Ratio         : 2.39\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute the annual volatility, sharpe ratio and annual excess return and plot the cumulative return\n",
    "from math import sqrt\n",
    "\n",
    "# compute the timeline for annualization\n",
    "T = (pd.to_datetime(return_ptf_index['cluster_portfolio_return'].index[-1]) - pd.to_datetime(return_ptf_index['cluster_portfolio_return'].index[0])) / np.timedelta64(1, 'Y')\n",
    "\n",
    "#portfolio Excess Return\n",
    "portfolio_excess_return=round(100*(return_ptf_index['cluster_portfolio_return'].cumprod().iloc[-1]**(1/T) - 1),2)\n",
    "\n",
    "#Portfolio Annual Volatility\n",
    "portfolio_annual_volatility=round(100*return_ptf_index['cluster_portfolio_return'].std()*sqrt(252),2)\n",
    "\n",
    "#Portfolio Sharpe Ratio\n",
    "portfolio_sharpe_ratio=round((portfolio_excess_return)/portfolio_annual_volatility,2)\n",
    "\n",
    "# Plot Results\n",
    "print (\"Portfolio Annual Excess Return : {}%\".format(portfolio_excess_return))\n",
    "print (\"Portfolio Annual Volatility    : {}% \".format(portfolio_annual_volatility))\n",
    "print (\"Portfolio Sharpe Ratio         : {}\".format(portfolio_sharpe_ratio)) \n",
    "\n",
    "plt.figure(figsize = (18,8))\n",
    "ax = plt.gca()\n",
    "plt.title(\"Portfolio Performance\")\n",
    "return_ptf_index['cluster_portfolio_return'].cumprod().plot(ax=ax,color=sns.color_palette()[1],linewidth=2)\n",
    "return_ptf_index['spx_index_return'].cumprod().plot(ax=ax,color=sns.color_palette()[2],linewidth=2)\n",
    "plt.ylabel(\"Cumulative Return %\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can repeat this analysis in order to build a portfolio that rebalances every end of Quarter.  Be sure the <code>as_of_date</code> and <code>end_date</code> variables are updated to reflect the most recent Quarter end and that these dates fit within the [2 Year Historical Daily Prices](https://rapidapi.com/alphawave/api/stock-prices2?endpoint=apiendpoint_33fa1878-1727-4775-beeb-f6b0da5314fd) endpoint from the [AlphaWave Data Stock Prices API](https://rapidapi.com/alphawave/api/stock-prices2/endpoints)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## References and Additional Resources\n",
    "<ul>\n",
    "  <li><a href =\"https://hdonnelly6.medium.com/list/machine-learning-for-investing-7f2690bb1826\"> Machine Learning for Investing </a></li>\n",
    "  <li><a href =\"https://www.cs.princeton.edu/sites/default/files/uploads/karina_marvin.pdf\"> Princeton University: Creating Diversified Portfolios Using Cluster Analysis </a></li>\n",
    " <li> <a href =\"https://scholarship.claremont.edu/cgi/viewcontent.cgi?article=3517&context=cmc_theses\"> Scholarship @ Claremont :K-Means Stock Clustering Analysis Based on Historical Price Movements and Financial Ratios  </a> </li>\n",
    " <li> <a href =\"https://realpython.com/k-means-clustering-python/\"> Real Python: K-Means Clustering in Python: A Practical Guide  </a> </li>\n",
    "</ul>"
   ]
  },
  {
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}