1. data_collection_graph_analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "36ee7fa9",
   "metadata": {},
   "source": [
    "# Collecting the data, creating and analysing the graph (network) representations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "67336f96",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pandas_datareader as pdr \n",
    "import yfinance as yf\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import scipy.stats as st\n",
    "import seaborn as sns\n",
    "import networkx as nx\n",
    "import datetime\n",
    "from scipy.stats import pearsonr\n",
    "from trading_functions import *"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c26a08fd",
   "metadata": {},
   "source": [
    "# Data collection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "59d84367",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_data(tickers, start_date, end_date):\n",
    "    \"\"\"\n",
    "    Downloads and processes stock data for a given list of tickers within a specified date range.\n",
    "\n",
    "    This function fetches the Adjusted Close prices and calculates the daily returns for each ticker\n",
    "    provided. It ensures that the data covers the exact date range requested and combines the results into two\n",
    "    dataframes: one for prices and one for daily returns, with missing values cleaned.\n",
    "\n",
    "    Parameters:\n",
    "    tickers (list): A list of stock ticker symbols.\n",
    "    start_date (str): The start date for the data retrieval, in YYYY-MM-DD format.\n",
    "    end_date (str): The end date for the data retrieval, in YYYY-MM-DD format.\n",
    "\n",
    "    Returns:\n",
    "    tuple: Contains two pandas DataFrames. The first DataFrame contains the Adjusted Close prices for each ticker,\n",
    "    and the second contains the daily returns. If no data is available, empty lists are returned.\n",
    "    \"\"\"\n",
    "    # Initialize variables to hold the number of tickers, and lists for price and daily returns dataframes\n",
    "    number_of_tickers = len(tickers)\n",
    "    price_list_df = []\n",
    "    daily_returns_list_df = []\n",
    "    \n",
    "    # Convert start and end dates from strings to pandas Timestamps for comparison\n",
    "    \n",
    "    start_date_datetime = pd.Timestamp(start_date)\n",
    "    end_date_datetime = pd.Timestamp(end_date)\n",
    "    \n",
    "    # Iterate over each ticker in the list\n",
    "    for index, ticker in enumerate(tickers):\n",
    "        print(index, len(tickers))  # Debug print: current index and total tickers\n",
    "        sp = yf.download(ticker, start=start_date, end=end_date)  # Download stock data\n",
    "        \n",
    "        # Check if the download returned data\n",
    "        if len(sp) > 0:\n",
    "            sp_mod = sp[[\"Adj Close\"]].copy()  # Copy the Adjusted Close prices\n",
    "            # Calculate daily returns and add as a new column\n",
    "            sp_mod.loc[:, \"Daily Return\"] = sp_mod[\"Adj Close\"] / sp_mod[\"Adj Close\"].shift(1) - 1\n",
    "            sp_mod[ticker] = sp_mod['Adj Close']  # Duplicate 'Adj Close' column for easier access\n",
    "            \n",
    "            # Determine the first and last date of the downloaded data\n",
    "            min_date = sp_mod[ticker].index[0]\n",
    "            max_date = sp_mod[ticker].index[-1]\n",
    "            print(min_date, start_date_datetime)  # Debug print: first date of data vs. requested start\n",
    "            print(max_date, end_date_datetime)  # Debug print: last date of data vs. requested end\n",
    "            \n",
    "            # Check if the data covers the desired date range exactly\n",
    "            if (min_date == start_date_datetime + datetime.timedelta(days=1)) and (max_date == end_date_datetime - datetime.timedelta(days=1)):\n",
    "                price_list_df.append(sp_mod[[ticker]].copy())  # Add price data to list\n",
    "                sp_mod[ticker] = sp_mod['Daily Return']  # Replace price data with daily returns\n",
    "                daily_returns_list_df.append(sp_mod[ticker])  # Add daily returns data to list\n",
    "    \n",
    "    # Check if we have collected any data\n",
    "    if len(price_list_df) > 0:\n",
    "        # Concatenate all price dataframes and drop missing values\n",
    "        price_df = pd.concat(price_list_df, axis=1).dropna()\n",
    "        # Concatenate all daily returns dataframes and drop missing values\n",
    "        daily_returns_df = pd.concat(daily_returns_list_df, axis=1).dropna()\n",
    "    else:\n",
    "        price_df = []\n",
    "        daily_returns_df = []\n",
    "    \n",
    "    return price_df, daily_returns_df\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9bb50be4",
   "metadata": {},
   "outputs": [],
   "source": [
    "format_string = \"%Y-%m-%d\"\n",
    "training_start_date = \"2010-01-03\"\n",
    "training_end_date = \"2021-01-01\"\n",
    "testing_start_date = '2021-01-03'\n",
    "testing_end_date = '2023-04-01'\n",
    "\n",
    "ticker_df = pd.read_csv('nasdaq_screener_1686671592339.csv')\n",
    "tickers = list(ticker_df['Symbol'].values)\n",
    "tickers = tickers[0:500]\n",
    "price_df_train, daily_returns_df_train = get_data(tickers,training_start_date,training_end_date)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "3aa14937",
   "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>AACG</th>\n",
       "      <th>AAL</th>\n",
       "      <th>AAME</th>\n",
       "      <th>AAON</th>\n",
       "      <th>AAPL</th>\n",
       "      <th>ABCB</th>\n",
       "      <th>ABEO</th>\n",
       "      <th>ABIO</th>\n",
       "      <th>ABUS</th>\n",
       "      <th>ABVC</th>\n",
       "      <th>...</th>\n",
       "      <th>AZPN</th>\n",
       "      <th>AZTA</th>\n",
       "      <th>BANF</th>\n",
       "      <th>BANFP</th>\n",
       "      <th>BANR</th>\n",
       "      <th>BBGI</th>\n",
       "      <th>BBSI</th>\n",
       "      <th>BCAL</th>\n",
       "      <th>BCBP</th>\n",
       "      <th>BCDA</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2010-01-04</th>\n",
       "      <td>0.377391</td>\n",
       "      <td>4.496877</td>\n",
       "      <td>1.193414</td>\n",
       "      <td>3.492817</td>\n",
       "      <td>6.470741</td>\n",
       "      <td>6.419696</td>\n",
       "      <td>4000.0</td>\n",
       "      <td>2479.679932</td>\n",
       "      <td>4.60</td>\n",
       "      <td>343949.03125</td>\n",
       "      <td>...</td>\n",
       "      <td>9.90</td>\n",
       "      <td>7.001814</td>\n",
       "      <td>13.661978</td>\n",
       "      <td>10.091043</td>\n",
       "      <td>13.484695</td>\n",
       "      <td>2.683620</td>\n",
       "      <td>9.686828</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.683233</td>\n",
       "      <td>91.800003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-05</th>\n",
       "      <td>0.373376</td>\n",
       "      <td>5.005959</td>\n",
       "      <td>1.265742</td>\n",
       "      <td>3.391475</td>\n",
       "      <td>6.481929</td>\n",
       "      <td>6.489953</td>\n",
       "      <td>4062.5</td>\n",
       "      <td>2449.439941</td>\n",
       "      <td>4.55</td>\n",
       "      <td>343949.03125</td>\n",
       "      <td>...</td>\n",
       "      <td>9.76</td>\n",
       "      <td>7.245425</td>\n",
       "      <td>13.230927</td>\n",
       "      <td>10.091043</td>\n",
       "      <td>13.239519</td>\n",
       "      <td>2.606946</td>\n",
       "      <td>9.648111</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.657359</td>\n",
       "      <td>91.800003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-06</th>\n",
       "      <td>0.362135</td>\n",
       "      <td>4.798553</td>\n",
       "      <td>1.265742</td>\n",
       "      <td>3.256933</td>\n",
       "      <td>6.378824</td>\n",
       "      <td>6.621684</td>\n",
       "      <td>3975.0</td>\n",
       "      <td>2449.439941</td>\n",
       "      <td>4.60</td>\n",
       "      <td>343949.03125</td>\n",
       "      <td>...</td>\n",
       "      <td>9.70</td>\n",
       "      <td>7.913384</td>\n",
       "      <td>13.311294</td>\n",
       "      <td>10.091043</td>\n",
       "      <td>13.141451</td>\n",
       "      <td>2.384588</td>\n",
       "      <td>9.679087</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.657359</td>\n",
       "      <td>91.800003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-07</th>\n",
       "      <td>0.362938</td>\n",
       "      <td>4.939964</td>\n",
       "      <td>1.193414</td>\n",
       "      <td>3.377497</td>\n",
       "      <td>6.367032</td>\n",
       "      <td>6.841237</td>\n",
       "      <td>4062.5</td>\n",
       "      <td>2487.239990</td>\n",
       "      <td>4.60</td>\n",
       "      <td>315286.62500</td>\n",
       "      <td>...</td>\n",
       "      <td>9.51</td>\n",
       "      <td>8.345594</td>\n",
       "      <td>13.406274</td>\n",
       "      <td>10.091043</td>\n",
       "      <td>13.876980</td>\n",
       "      <td>2.606946</td>\n",
       "      <td>10.577306</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.657359</td>\n",
       "      <td>91.800003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010-01-08</th>\n",
       "      <td>0.381406</td>\n",
       "      <td>4.845691</td>\n",
       "      <td>1.256701</td>\n",
       "      <td>3.417684</td>\n",
       "      <td>6.409363</td>\n",
       "      <td>6.867583</td>\n",
       "      <td>4112.5</td>\n",
       "      <td>2494.800049</td>\n",
       "      <td>3.95</td>\n",
       "      <td>300955.40625</td>\n",
       "      <td>...</td>\n",
       "      <td>9.41</td>\n",
       "      <td>8.400604</td>\n",
       "      <td>13.424540</td>\n",
       "      <td>10.091043</td>\n",
       "      <td>14.661549</td>\n",
       "      <td>2.683620</td>\n",
       "      <td>10.708936</td>\n",
       "      <td>9.0</td>\n",
       "      <td>4.657359</td>\n",
       "      <td>91.800003</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 131 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                AACG       AAL      AAME      AAON      AAPL      ABCB  \\\n",
       "Date                                                                     \n",
       "2010-01-04  0.377391  4.496877  1.193414  3.492817  6.470741  6.419696   \n",
       "2010-01-05  0.373376  5.005959  1.265742  3.391475  6.481929  6.489953   \n",
       "2010-01-06  0.362135  4.798553  1.265742  3.256933  6.378824  6.621684   \n",
       "2010-01-07  0.362938  4.939964  1.193414  3.377497  6.367032  6.841237   \n",
       "2010-01-08  0.381406  4.845691  1.256701  3.417684  6.409363  6.867583   \n",
       "\n",
       "              ABEO         ABIO  ABUS          ABVC  ...  AZPN      AZTA  \\\n",
       "Date                                                 ...                   \n",
       "2010-01-04  4000.0  2479.679932  4.60  343949.03125  ...  9.90  7.001814   \n",
       "2010-01-05  4062.5  2449.439941  4.55  343949.03125  ...  9.76  7.245425   \n",
       "2010-01-06  3975.0  2449.439941  4.60  343949.03125  ...  9.70  7.913384   \n",
       "2010-01-07  4062.5  2487.239990  4.60  315286.62500  ...  9.51  8.345594   \n",
       "2010-01-08  4112.5  2494.800049  3.95  300955.40625  ...  9.41  8.400604   \n",
       "\n",
       "                 BANF      BANFP       BANR      BBGI       BBSI  BCAL  \\\n",
       "Date                                                                     \n",
       "2010-01-04  13.661978  10.091043  13.484695  2.683620   9.686828   9.0   \n",
       "2010-01-05  13.230927  10.091043  13.239519  2.606946   9.648111   9.0   \n",
       "2010-01-06  13.311294  10.091043  13.141451  2.384588   9.679087   9.0   \n",
       "2010-01-07  13.406274  10.091043  13.876980  2.606946  10.577306   9.0   \n",
       "2010-01-08  13.424540  10.091043  14.661549  2.683620  10.708936   9.0   \n",
       "\n",
       "                BCBP       BCDA  \n",
       "Date                             \n",
       "2010-01-04  4.683233  91.800003  \n",
       "2010-01-05  4.657359  91.800003  \n",
       "2010-01-06  4.657359  91.800003  \n",
       "2010-01-07  4.657359  91.800003  \n",
       "2010-01-08  4.657359  91.800003  \n",
       "\n",
       "[5 rows x 131 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "price_df_train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c0a8094",
   "metadata": {},
   "source": [
    "# 2 Network creation\n",
    "I have to determine where to split the dataset and create a network for that period (time window), which ised to check for cross-correlations in that period."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a2c85c8",
   "metadata": {},
   "source": [
    "# 2.1. Equal time splits\n",
    "I approach the network creation with do equal split between different periods (time windows)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "80994370",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def split_time_windows(tot_days, n_time_windows, chosen_time_window_size, size_boolean):\n",
    "    \"\"\"\n",
    "    Splits the total number of days into time windows either based on the number of windows or a specified window size.\n",
    "\n",
    "    This function divides the total days into equal-sized time windows. If size_boolean is True, it instead uses a specified\n",
    "    window size to determine the splits. The function rounds window boundaries to the nearest integer day and adjusts the\n",
    "    window sizes accordingly.\n",
    "\n",
    "    Parameters:\n",
    "    tot_days (int): Total number of days to be split.\n",
    "    n_time_windows (int): The desired number of time windows (ignored if size_boolean is True).\n",
    "    chosen_time_window_size (int): The size of each time window if size_boolean is True.\n",
    "    size_boolean (bool): Determines whether the time windows are based on a fixed number (False) or a specified size (True).\n",
    "\n",
    "    Returns:\n",
    "    tuple: A list of time window splits and the size of a single time window.\n",
    "    \"\"\"\n",
    "    \n",
    "    # Generate evenly spaced values over the specified interval\n",
    "    time_window_splits_total = np.linspace(0, tot_days, n_time_windows)\n",
    "    # Round the values to ensure they correspond to whole days\n",
    "    time_window_splits = [int(x) for x in time_window_splits_total]\n",
    "    # Calculate the size of the time window based on the first interval\n",
    "    time_window_size = time_window_splits[1] - time_window_splits[0]\n",
    "    \n",
    "    # If size_boolean is True, calculate the time window splits based on the chosen time window size\n",
    "    if size_boolean:\n",
    "        # Generate a range of values with the specified step size\n",
    "        time_window_splits_total = np.arange(0, tot_days, chosen_time_window_size)\n",
    "        # Adjust the calculation to account for the specific window size, reversing the order\n",
    "        time_window_splits = sorted([tot_days + 1 - int(x) for x in time_window_splits_total])\n",
    "        # The size of the time window is set to the chosen window size\n",
    "        time_window_size = chosen_time_window_size\n",
    "        # Recalculate the actual size of the first window to ensure accuracy\n",
    "        time_window_size = time_window_splits[1] - time_window_splits[0]\n",
    "    \n",
    "    return time_window_splits, time_window_size\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b46bd8fc",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The total number of days is 2769\n",
      "Total number of time windows is 93\n",
      "The number of days for each time window is 30\n"
     ]
    }
   ],
   "source": [
    "# Define the number of time windows desired\n",
    "n_time_windows = 100\n",
    "# Set to True to indicate that the split should be based on a specific window size, rather than an equal division of days\n",
    "size_boolean = True\n",
    "# Specify the size of each time window when 'size_boolean' is True\n",
    "chosen_time_window_size = 30\n",
    "# Calculate the total number of days based on the length of the price dataframe for training\n",
    "tot_days = len(price_df_train)\n",
    "# Use the split_time_windows function to calculate the splits and size of the time windows based on the above parameters\n",
    "time_window_splits, time_window_size = split_time_windows(tot_days, n_time_windows, chosen_time_window_size, size_boolean)\n",
    "\n",
    "# Output the total number of days covered by the price dataframe\n",
    "print('The total number of days is', tot_days)\n",
    "# Output the total number of time windows generated\n",
    "print('Total number of time windows is', len(time_window_splits))\n",
    "# Output the number of days that each time window covers\n",
    "print('The number of days for each time window is', time_window_size)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "bb7b854e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def find_correlation_groups(time_window_splits, price_df, link_threshold, frequency_threshold, draw):\n",
    "    \"\"\"\n",
    "    Identifies groups of assets with strong correlations within specified time windows.\n",
    "    \n",
    "    Parameters:\n",
    "    - time_window_splits: Indices to split the price DataFrame into time windows.\n",
    "    - price_df: DataFrame containing price data for different assets.\n",
    "    - link_threshold: Correlation coefficient threshold for linking two assets.\n",
    "    - frequency_threshold: Threshold for the frequency of a link appearing across time windows to be considered significant.\n",
    "    - draw: Boolean flag indicating whether to draw the network graph.\n",
    "    \n",
    "    Returns:\n",
    "    - List of sets, each set contains tickers of assets that are strongly correlated within the time windows.\n",
    "    \"\"\"\n",
    "\n",
    "    # Calculate correlation matrices for each time window\n",
    "    corr_df_dict = {}\n",
    "    for i in range(1, len(time_window_splits)):\n",
    "        # Compute correlation matrix for each time window and store it\n",
    "        corr_df_dict[i] = price_df.iloc[time_window_splits[i-1]:time_window_splits[i], :].corr()\n",
    "\n",
    "    # Prepare asset ticker list from the columns of the first correlation matrix\n",
    "    asset_ticker_list = list(corr_df_dict[1].columns)\n",
    "    \n",
    "    # Initialize dictionary to hold correlation pairs for each time window\n",
    "    correlation_pairs_dict = {}\n",
    "    for time_window_index in corr_df_dict.keys():\n",
    "        correlation_pairs_dict[time_window_index] = {}\n",
    "        # Iterate over all possible asset pairs\n",
    "        for ticker_1 in asset_ticker_list:\n",
    "            for ticker_2 in asset_ticker_list:\n",
    "                if ticker_1 != ticker_2:  # Avoid self-correlation\n",
    "                    # Store the correlation coefficient between each pair\n",
    "                    correlation_pairs_dict[time_window_index][(ticker_1, ticker_2)] = corr_df_dict[time_window_index].loc[ticker_1, ticker_2]\n",
    "\n",
    "    # Identify strongly correlated pairs and build a network for each time window\n",
    "    network_dict = {}\n",
    "    for time_window_index in correlation_pairs_dict.keys():\n",
    "        G = nx.Graph()  # Initialize a new graph for the time window\n",
    "        for pair, correlation in correlation_pairs_dict[time_window_index].items():\n",
    "            if correlation >= link_threshold:  # Check if the correlation exceeds the threshold\n",
    "                G.add_edge(pair[0], pair[1])  # Add an edge for strongly correlated pairs\n",
    "        network_dict[time_window_index] = G  # Store the graph\n",
    "\n",
    "    # Count the frequency of each link across all time windows\n",
    "    weighted_edgelist_dict = {}\n",
    "    number_of_windows = len(network_dict.keys())\n",
    "    for G in network_dict.values():\n",
    "        for edge in G.edges:\n",
    "            if edge in weighted_edgelist_dict:\n",
    "                weighted_edgelist_dict[edge] += 1 / number_of_windows\n",
    "            else:\n",
    "                weighted_edgelist_dict[edge] = 1 / number_of_windows\n",
    "\n",
    "    # Filter links by frequency threshold\n",
    "    thresholded_edgelist_dict = {}\n",
    "    for edge, freq in weighted_edgelist_dict.items():\n",
    "        if freq >= frequency_threshold:\n",
    "            thresholded_edgelist_dict[edge] = np.round(freq, 2)\n",
    "\n",
    "    # Create a final graph with edges that meet the frequency threshold\n",
    "    G_thresh = nx.Graph()\n",
    "    for edge, weight in thresholded_edgelist_dict.items():\n",
    "        G_thresh.add_edge(edge[0], edge[1], weight=weight)\n",
    "    \n",
    "    # Identify separate groups of connected components\n",
    "    separate_groups = list(nx.connected_components(G_thresh))\n",
    "\n",
    "    # Optionally draw the graph\n",
    "    if draw:\n",
    "        pos = nx.spring_layout(G_thresh)  # Position nodes using Spring layout\n",
    "        plt.figure()\n",
    "        nx.draw(G_thresh, pos, edge_color='black', node_size=1050, with_labels=True, node_color='black', font_color='white')\n",
    "        nx.draw_networkx_edge_labels(G_thresh, pos, edge_labels=thresholded_edgelist_dict, font_color='red')\n",
    "\n",
    "    # Filter and return groups larger than 1 asset\n",
    "    separate_groups_list = [group for group in separate_groups if len(group) > 1]\n",
    "    return separate_groups_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "caeedb75",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Each link has a correlation threshold link_threshold= 0.6 , meaning that it needs to have at least this much of correlation to be considered at all.\n",
      "Each link has to appear in at least  frequency_threshold 0.5  of all time windows.\n",
      "This gives us robust groups, as it is shown in the drawing below:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# As an example, set the parameters \n",
    "link_threshold = 0.6\n",
    "frequency_threshold = 0.5\n",
    "draw = True\n",
    "separate_groups  = find_correlation_groups(time_window_splits,price_df_train,link_threshold,frequency_threshold ,draw)\n",
    "print('Each link has a correlation threshold','link_threshold=',link_threshold,', meaning that it needs to have at least this much of correlation to be considered at all.')\n",
    "print('Each link has to appear in at least ','frequency_threshold',frequency_threshold,' of all time windows.')\n",
    "print(\"This gives us robust groups, as it is shown in the drawing below:\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec7dbf93",
   "metadata": {},
   "source": [
    "# 2.2 Parameter analysis\n",
    "In this section, I analyze the interplay between the link and frequency threshold."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "48ac26e1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set step size for threshold increments and draw flag for visualization\n",
    "step = 0.1\n",
    "draw = False\n",
    "\n",
    "# Initialize dictionary to store results for different combinations of thresholds\n",
    "link_frequency_dictionary = {'frequency_threshold': [], 'link_threshold': [], 'number_of_groups': [], 'mean_group_size': [], 'std_group_size': []}\n",
    "\n",
    "# Loop through possible link_threshold values from 0 to 1 by step\n",
    "for link_threshold in np.arange(0, 1 + step, step):\n",
    "    # Loop through possible frequency_threshold values from 0 to 1 by step\n",
    "    for frequency_threshold in np.arange(0, 1 + step, step):\n",
    "        print(np.round(link_threshold,2),np.round(frequency_threshold,2))\n",
    "        \n",
    "        # Find separate groups based on current thresholds\n",
    "        separate_groups = find_correlation_groups(time_window_splits, price_df_train, link_threshold, frequency_threshold, draw)\n",
    "        # Calculate the number of groups found\n",
    "        nr_groups = len(separate_groups)\n",
    "        # Calculate sizes of each group as a fraction of total tickers\n",
    "        group_sizes = [len(group) / len(tickers) for group in separate_groups]\n",
    "        \n",
    "        # Calculate mean and standard deviation of group sizes if any groups are found\n",
    "        if len(group_sizes) > 0:\n",
    "            mean_group_size = np.mean(group_sizes)\n",
    "            std_group_size = np.std(group_sizes)\n",
    "        else:  # Default to 0 if no groups are found\n",
    "            mean_group_size = 0\n",
    "            std_group_size = 0\n",
    "        \n",
    "        # Append results for current threshold combination to the dictionary\n",
    "        link_frequency_dictionary['frequency_threshold'].append(np.round(frequency_threshold, 1))\n",
    "        link_frequency_dictionary['link_threshold'].append(np.round(link_threshold, 1))\n",
    "        link_frequency_dictionary['mean_group_size'].append(np.round(mean_group_size, 3))\n",
    "        link_frequency_dictionary['std_group_size'].append(np.round(std_group_size, 3))\n",
    "        link_frequency_dictionary['number_of_groups'].append(nr_groups)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "25bcd035",
   "metadata": {},
   "outputs": [],
   "source": [
    "link_frequency_df = pd.DataFrame(link_frequency_dictionary)\n",
    "number_of_groups_df = link_frequency_df.pivot(index='frequency_threshold', columns='link_threshold', values='number_of_groups')\n",
    "mean_group_size_df = link_frequency_df.pivot(index='frequency_threshold', columns='link_threshold', values='mean_group_size')\n",
    "std_group_size_df = link_frequency_df.pivot(index='frequency_threshold', columns='link_threshold', values='std_group_size')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9c29918d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (8,6))\n",
    "ax = sns.heatmap(number_of_groups_df, annot=True, fmt=\"g\", cmap='viridis')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Number of groups');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "c8390edf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (8,6))\n",
    "ax = sns.heatmap(mean_group_size_df, annot=True, fmt=\"g\", cmap='viridis')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Mean group size');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "394a52f2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (8,6))\n",
    "ax = sns.heatmap(std_group_size_df, annot=True, fmt=\"g\", cmap='viridis')\n",
    "ax.invert_yaxis()\n",
    "plt.title('Standard deviation of group size');"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}