05_Data_Hubs.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SimFin Tutorial 05 - Data Hubs\n",
    "\n",
    "[Original repository on GitHub](https://github.com/simfin/simfin-tutorials)\n",
    "\n",
    "This tutorial was originally written by [Hvass Labs](https://github.com/Hvass-Labs)\n",
    "\n",
    "----\n",
    "\n",
    "\"I just have to type Y instead of Y-E-S ... Hey Miss-didn't-find-me-attractive-sexually-anymore, I just tripled my productivity!\" &ndash; [The Simpsons](https://www.youtube.com/watch?v=R_rF4kcqLkI)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introduction\n",
    "\n",
    "In the previous tutorials we saw how to download and load financial data from the SimFin server using the `sf.load` functions, how to resample and reindex the data using the `sf.resample` and `sf.reindex` functions, how to calculate growth-rates and stock-returns using the `sf.rel_change` and `sf.mean_log_change` functions, and finally how to combine all of this to calculate signals using functions like `sf.price_signals` and `sf.val_signals`.\n",
    "\n",
    "Because we have to go through all these basic steps every time we want to work on the financial data, and this mostly consists of copy-pasting our previous code, the simfin package provides what we call *Data Hubs* to do most of the work for us.\n",
    "\n",
    "The data-hub also uses disk-caching for its slow computations, so the results are automatically saved to disk and will be reloaded very quickly the next time the slow functions are called. The data-hub also ensures the disk-cache is updated whenever new datasets are downloaded from the SimFin server. The data-hub's functions are also RAM-cached, so the second time you call them, they return the results nearly instantly.\n",
    "\n",
    "Altogether, the data-hub makes it much easier and faster to work with the financial data.\n",
    "\n",
    "It is not strictly necessary for you to be familiar with the previous tutorials on the basics of simfin, but it is helpful because you would have a better understanding of what exactly the data-hub is doing."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "\n",
    "# Import the main functionality from the SimFin Python API.\n",
    "import simfin as sf\n",
    "\n",
    "# Import names used for easy access to SimFin's data-columns.\n",
    "from simfin.names import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0.3.0'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Version of the SimFin Python API.\n",
    "sf.__version__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Config"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SimFin data-directory.\n",
    "sf.set_data_dir('~/simfin_data/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SimFin load API key or use free data.\n",
    "sf.load_api_key(path='~/simfin_api_key.txt', default_key='free')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Seaborn set plotting style.\n",
    "sns.set_style(\"whitegrid\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Hub\n",
    "\n",
    "In these examples, we are interested in US stock-market data for the tickers AAPL, AMZN and MSFT. We want to add a 60-day offset to the dates of the financial reports, to simulate the delay between the date of a financial report and when it was actually made available to the public. And finally we want to download the datasets from the SimFin server every 30 days, except for the share-prices which should be downloaded every 10 days."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We are interested in the US stock-market.\n",
    "market = 'us'\n",
    "\n",
    "# List of tickers we want. If None then all tickers are used.\n",
    "tickers = ['AAPL', 'AMZN', 'MSFT']\n",
    "\n",
    "# Add this date-offset to the fundamental data such as\n",
    "# Income Statements etc., because the REPORT_DATE is not\n",
    "# when it was actually made available to the public,\n",
    "# which can be 1, 2 or even 3 months after the Report Date.\n",
    "offset = pd.DateOffset(days=60)\n",
    "\n",
    "# Refresh the fundamental datasets (Income Statements etc.)\n",
    "# every 30 days.\n",
    "refresh_days = 30\n",
    "\n",
    "# Refresh the dataset with shareprices every 10 days.\n",
    "refresh_days_shareprices = 10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using these settings, we can then create the data-hub object. This basically just creates and configures the data-hub. It does not load any data or perform any calculations at this point, so it returns nearly instantly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 20 µs, sys: 1 µs, total: 21 µs\n",
      "Wall time: 23.4 µs\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "hub = sf.StockHub(market=market, tickers=tickers, offset=offset,\n",
    "                  refresh_days=refresh_days,\n",
    "                  refresh_days_shareprices=refresh_days_shareprices)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Datasets\n",
    "\n",
    "We can now use the data-hub to load financial datasets. This is very similar to the `sf.load` functions, except that we don't have to set the market-argument because that is already known by the data-hub. This is how we load the dataset with share-prices:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset \"us-shareprices-daily\" on disk (0 days old).\n",
      "- Loading from disk ... Done!\n",
      "CPU times: user 15.2 s, sys: 1.24 s, total: 16.4 s\n",
      "Wall time: 14.9 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_prices = hub.load_shareprices(variant='daily')"
   ]
  },
  {
   "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></th>\n",
       "      <th>SimFinId</th>\n",
       "      <th>Open</th>\n",
       "      <th>Low</th>\n",
       "      <th>High</th>\n",
       "      <th>Close</th>\n",
       "      <th>Adj. Close</th>\n",
       "      <th>Dividend</th>\n",
       "      <th>Volume</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2007-01-03</td>\n",
       "      <td>111052</td>\n",
       "      <td>12.3271</td>\n",
       "      <td>11.7000</td>\n",
       "      <td>12.3686</td>\n",
       "      <td>11.9714</td>\n",
       "      <td>10.42</td>\n",
       "      <td>NaN</td>\n",
       "      <td>44225700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-04</td>\n",
       "      <td>111052</td>\n",
       "      <td>12.0071</td>\n",
       "      <td>11.9743</td>\n",
       "      <td>12.2786</td>\n",
       "      <td>12.2371</td>\n",
       "      <td>10.65</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30259300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-05</td>\n",
       "      <td>111052</td>\n",
       "      <td>12.2529</td>\n",
       "      <td>12.0571</td>\n",
       "      <td>12.3143</td>\n",
       "      <td>12.1500</td>\n",
       "      <td>10.57</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29812200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-08</td>\n",
       "      <td>111052</td>\n",
       "      <td>12.2800</td>\n",
       "      <td>12.1829</td>\n",
       "      <td>12.3614</td>\n",
       "      <td>12.2100</td>\n",
       "      <td>10.62</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28468100</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-09</td>\n",
       "      <td>111052</td>\n",
       "      <td>12.3500</td>\n",
       "      <td>12.1643</td>\n",
       "      <td>13.2829</td>\n",
       "      <td>13.2243</td>\n",
       "      <td>11.51</td>\n",
       "      <td>NaN</td>\n",
       "      <td>119617800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   SimFinId     Open      Low     High    Close  Adj. Close  \\\n",
       "Ticker Date                                                                   \n",
       "AAPL   2007-01-03    111052  12.3271  11.7000  12.3686  11.9714       10.42   \n",
       "       2007-01-04    111052  12.0071  11.9743  12.2786  12.2371       10.65   \n",
       "       2007-01-05    111052  12.2529  12.0571  12.3143  12.1500       10.57   \n",
       "       2007-01-08    111052  12.2800  12.1829  12.3614  12.2100       10.62   \n",
       "       2007-01-09    111052  12.3500  12.1643  13.2829  13.2243       11.51   \n",
       "\n",
       "                   Dividend     Volume  \n",
       "Ticker Date                             \n",
       "AAPL   2007-01-03       NaN   44225700  \n",
       "       2007-01-04       NaN   30259300  \n",
       "       2007-01-05       NaN   29812200  \n",
       "       2007-01-08       NaN   28468100  \n",
       "       2007-01-09       NaN  119617800  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_prices.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note what happens if we call the exact same function again with the exact same argument - it returns the exact same result almost instantly. This is because the data-hub uses a RAM-cache for most of its functions, so the DataFrame is just taken from RAM instead of being reloaded from disk."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 8 µs, sys: 0 ns, total: 8 µs\n",
      "Wall time: 12.4 µs\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_prices = hub.load_shareprices(variant='daily')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also use the data-hub to load other datasets such as the TTM Income Statements:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset \"us-income-ttm\" on disk (6 days old).\n",
      "- Loading from disk ... Done!\n",
      "CPU times: user 307 ms, sys: 12.2 ms, total: 320 ms\n",
      "Wall time: 317 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_income_ttm = hub.load_income(variant='ttm')"
   ]
  },
  {
   "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></th>\n",
       "      <th>SimFinId</th>\n",
       "      <th>Currency</th>\n",
       "      <th>Fiscal Year</th>\n",
       "      <th>Fiscal Period</th>\n",
       "      <th>Publish Date</th>\n",
       "      <th>Shares (Basic)</th>\n",
       "      <th>Shares (Diluted)</th>\n",
       "      <th>Revenue</th>\n",
       "      <th>Cost of Revenue</th>\n",
       "      <th>Gross Profit</th>\n",
       "      <th>...</th>\n",
       "      <th>Non-Operating Income (Loss)</th>\n",
       "      <th>Interest Expense, Net</th>\n",
       "      <th>Pretax Income (Loss), Adj.</th>\n",
       "      <th>Abnormal Gains (Losses)</th>\n",
       "      <th>Pretax Income (Loss)</th>\n",
       "      <th>Income Tax (Expense) Benefit, Net</th>\n",
       "      <th>Income (Loss) from Continuing Operations</th>\n",
       "      <th>Net Extraordinary Gains (Losses)</th>\n",
       "      <th>Net Income</th>\n",
       "      <th>Net Income (Common)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\n",
       "      <th>Report 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",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2010-03-31</td>\n",
       "      <td>111052</td>\n",
       "      <td>USD</td>\n",
       "      <td>2010</td>\n",
       "      <td>Q2</td>\n",
       "      <td>2010-04-21</td>\n",
       "      <td>6.304956e+09</td>\n",
       "      <td>6.415840e+09</td>\n",
       "      <td>5.112300e+10</td>\n",
       "      <td>-2.999900e+10</td>\n",
       "      <td>2.112400e+10</td>\n",
       "      <td>...</td>\n",
       "      <td>188000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>15208000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>15208000000</td>\n",
       "      <td>-4.396000e+09</td>\n",
       "      <td>10812000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10812000000</td>\n",
       "      <td>10812000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-30</td>\n",
       "      <td>111052</td>\n",
       "      <td>USD</td>\n",
       "      <td>2010</td>\n",
       "      <td>Q3</td>\n",
       "      <td>2010-07-21</td>\n",
       "      <td>6.337305e+09</td>\n",
       "      <td>6.447691e+09</td>\n",
       "      <td>5.708900e+10</td>\n",
       "      <td>-3.381200e+10</td>\n",
       "      <td>2.327700e+10</td>\n",
       "      <td>...</td>\n",
       "      <td>186000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16808000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16808000000</td>\n",
       "      <td>-4.571000e+09</td>\n",
       "      <td>12237000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12237000000</td>\n",
       "      <td>12237000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-09-30</td>\n",
       "      <td>111052</td>\n",
       "      <td>USD</td>\n",
       "      <td>2010</td>\n",
       "      <td>Q4</td>\n",
       "      <td>2010-10-27</td>\n",
       "      <td>6.366227e+09</td>\n",
       "      <td>6.472984e+09</td>\n",
       "      <td>6.522500e+10</td>\n",
       "      <td>-3.954100e+10</td>\n",
       "      <td>2.568400e+10</td>\n",
       "      <td>...</td>\n",
       "      <td>155000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>18540000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>18540000000</td>\n",
       "      <td>-4.527000e+09</td>\n",
       "      <td>14013000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14013000000</td>\n",
       "      <td>14013000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-12-31</td>\n",
       "      <td>111052</td>\n",
       "      <td>USD</td>\n",
       "      <td>2011</td>\n",
       "      <td>Q1</td>\n",
       "      <td>2011-01-19</td>\n",
       "      <td>6.393793e+09</td>\n",
       "      <td>6.496383e+09</td>\n",
       "      <td>7.628300e+10</td>\n",
       "      <td>-4.671200e+10</td>\n",
       "      <td>2.957100e+10</td>\n",
       "      <td>...</td>\n",
       "      <td>258000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21745000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21745000000</td>\n",
       "      <td>-5.106000e+09</td>\n",
       "      <td>16639000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>16639000000</td>\n",
       "      <td>16639000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2011-03-31</td>\n",
       "      <td>111052</td>\n",
       "      <td>USD</td>\n",
       "      <td>2011</td>\n",
       "      <td>Q2</td>\n",
       "      <td>2011-04-21</td>\n",
       "      <td>6.421177e+09</td>\n",
       "      <td>6.519249e+09</td>\n",
       "      <td>8.745100e+10</td>\n",
       "      <td>-5.328700e+10</td>\n",
       "      <td>3.416400e+10</td>\n",
       "      <td>...</td>\n",
       "      <td>234000000.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>25616000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>25616000000</td>\n",
       "      <td>-6.064000e+09</td>\n",
       "      <td>19552000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>19552000000</td>\n",
       "      <td>19552000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    SimFinId Currency  Fiscal Year Fiscal Period Publish Date  \\\n",
       "Ticker Report Date                                                              \n",
       "AAPL   2010-03-31     111052      USD         2010            Q2   2010-04-21   \n",
       "       2010-06-30     111052      USD         2010            Q3   2010-07-21   \n",
       "       2010-09-30     111052      USD         2010            Q4   2010-10-27   \n",
       "       2010-12-31     111052      USD         2011            Q1   2011-01-19   \n",
       "       2011-03-31     111052      USD         2011            Q2   2011-04-21   \n",
       "\n",
       "                    Shares (Basic)  Shares (Diluted)       Revenue  \\\n",
       "Ticker Report Date                                                   \n",
       "AAPL   2010-03-31     6.304956e+09      6.415840e+09  5.112300e+10   \n",
       "       2010-06-30     6.337305e+09      6.447691e+09  5.708900e+10   \n",
       "       2010-09-30     6.366227e+09      6.472984e+09  6.522500e+10   \n",
       "       2010-12-31     6.393793e+09      6.496383e+09  7.628300e+10   \n",
       "       2011-03-31     6.421177e+09      6.519249e+09  8.745100e+10   \n",
       "\n",
       "                    Cost of Revenue  Gross Profit  ...  \\\n",
       "Ticker Report Date                                 ...   \n",
       "AAPL   2010-03-31     -2.999900e+10  2.112400e+10  ...   \n",
       "       2010-06-30     -3.381200e+10  2.327700e+10  ...   \n",
       "       2010-09-30     -3.954100e+10  2.568400e+10  ...   \n",
       "       2010-12-31     -4.671200e+10  2.957100e+10  ...   \n",
       "       2011-03-31     -5.328700e+10  3.416400e+10  ...   \n",
       "\n",
       "                    Non-Operating Income (Loss)  Interest Expense, Net  \\\n",
       "Ticker Report Date                                                       \n",
       "AAPL   2010-03-31                   188000000.0                    NaN   \n",
       "       2010-06-30                   186000000.0                    NaN   \n",
       "       2010-09-30                   155000000.0                    NaN   \n",
       "       2010-12-31                   258000000.0                    NaN   \n",
       "       2011-03-31                   234000000.0                    NaN   \n",
       "\n",
       "                    Pretax Income (Loss), Adj.  Abnormal Gains (Losses)  \\\n",
       "Ticker Report Date                                                        \n",
       "AAPL   2010-03-31                  15208000000                      NaN   \n",
       "       2010-06-30                  16808000000                      NaN   \n",
       "       2010-09-30                  18540000000                      NaN   \n",
       "       2010-12-31                  21745000000                      NaN   \n",
       "       2011-03-31                  25616000000                      NaN   \n",
       "\n",
       "                    Pretax Income (Loss)  Income Tax (Expense) Benefit, Net  \\\n",
       "Ticker Report Date                                                            \n",
       "AAPL   2010-03-31            15208000000                      -4.396000e+09   \n",
       "       2010-06-30            16808000000                      -4.571000e+09   \n",
       "       2010-09-30            18540000000                      -4.527000e+09   \n",
       "       2010-12-31            21745000000                      -5.106000e+09   \n",
       "       2011-03-31            25616000000                      -6.064000e+09   \n",
       "\n",
       "                    Income (Loss) from Continuing Operations  \\\n",
       "Ticker Report Date                                             \n",
       "AAPL   2010-03-31                                10812000000   \n",
       "       2010-06-30                                12237000000   \n",
       "       2010-09-30                                14013000000   \n",
       "       2010-12-31                                16639000000   \n",
       "       2011-03-31                                19552000000   \n",
       "\n",
       "                    Net Extraordinary Gains (Losses)   Net Income  \\\n",
       "Ticker Report Date                                                  \n",
       "AAPL   2010-03-31                                NaN  10812000000   \n",
       "       2010-06-30                                NaN  12237000000   \n",
       "       2010-09-30                                NaN  14013000000   \n",
       "       2010-12-31                                NaN  16639000000   \n",
       "       2011-03-31                                NaN  19552000000   \n",
       "\n",
       "                    Net Income (Common)  \n",
       "Ticker Report Date                       \n",
       "AAPL   2010-03-31           10812000000  \n",
       "       2010-06-30           12237000000  \n",
       "       2010-09-30           14013000000  \n",
       "       2010-12-31           16639000000  \n",
       "       2011-03-31           19552000000  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_income_ttm.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Again we see that the function returns nearly instantly when we call it the second time, provided the argument is exactly the same, because then the DataFrame is merely taken from the RAM-cache instead of being reloaded from disk:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 12 µs, sys: 1e+03 ns, total: 13 µs\n",
      "Wall time: 16.9 µs\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_income_ttm = hub.load_income(variant='ttm')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RAM-Cache Warning\n",
    "\n",
    "Because the data-hub uses a RAM-cache for its functions, it means that you should **NEVER** modify the DataFrames returned by the data-hub's functions, because that would modify the DataFrames stored in the RAM-cache, which may interfere with the data-hubs' other functions that use the same data. If you want to modify the DataFrames returned by the data-hub, you should **ALWAYS** make a copy first!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the share-prices.\n",
    "df_prices = hub.load_shareprices(variant='daily')\n",
    "\n",
    "# Now we want to set the Closing share-prices to zero.\n",
    "\n",
    "# This is BAD because it changes the contents of the RAM-cache\n",
    "# inside the hub-object.\n",
    "# df_prices[CLOSE] = 0.0\n",
    "\n",
    "# This is fine because it only changes a copy of the data, so\n",
    "# the RAM-cache inside the hub-object remains unchanged.\n",
    "df_prices2 = df_prices.copy()\n",
    "df_prices2[CLOSE] = 0.0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Although it is possible to clear the RAM-cache for the data-hubs' functions using the following command, it is recommended that you simply create a new data-hub and start over, if you accidentally modified any of the DataFrames returned by the data-hub."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Clear the RAM-cache for the function.\n",
    "# hub.load_shareprices.cache_clear()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate Signals\n",
    "\n",
    "The benefits of the data-hub are much greater for the functions that calculate signals such as valuation ratios, moving averages of the share-prices, sales and earnings growth, etc. The data-hub makes the syntax much simpler for these functions, because you only need to call a single function, which then loads the required datasets (or takes them from the RAM-cache if they have already been loaded), and then passes the datasets along with other relevant arguments to the lower-level functions that perform the actual computations.\n",
    "\n",
    "Furthermore, the data-hub also uses a disk-cache for its slow functions, so the results are automatically saved to a file on disk, which can be reloaded very quickly the next time you call the function. The cache-files are automatically updated by recomputing the slow functions, whenever the relevant datasets are downloaded again from the SimFin server.\n",
    "\n",
    "The disk-cache works in addition to the RAM-cache, so the data-hub uses two layers of caching to return the data as fast as possible, when the data does not have to be reloaded or calculated again.\n",
    "\n",
    "The filenames used by the disk-cache are automatically set from all the arguments to the data-hub and its functions, using a hashing-function that creates a short but nearly unique ID code, so the cache-files are kept separate when you call the same function but with different arguments."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Price Signals\n",
    "\n",
    "This is how easily we may calculate price signals such as Moving Averages. Note that the function finished very quickly because we are only using data for 3 stocks, so there is no need for the disk-cache here. But if we had not specified a list of tickers when creating the data-hub object, this function would calculate the signals for thousands of stocks, which takes a while to compute, so the disk-cache is often very useful."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'price_signals-d03458cf.pickle' not on disk.\n",
      "- Running function price_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 53.4 ms, sys: 100 µs, total: 53.5 ms\n",
      "Wall time: 55.5 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_price_signals = hub.price_signals()"
   ]
  },
  {
   "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></th>\n",
       "      <th>EMA</th>\n",
       "      <th>MACD</th>\n",
       "      <th>MACD-EMA</th>\n",
       "      <th>MAVG 20</th>\n",
       "      <th>MAVG 200</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\n",
       "      <th>Date</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",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2007-01-03</td>\n",
       "      <td>11.971400</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-04</td>\n",
       "      <td>12.110893</td>\n",
       "      <td>0.005961</td>\n",
       "      <td>0.003312</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-05</td>\n",
       "      <td>12.125253</td>\n",
       "      <td>0.004895</td>\n",
       "      <td>0.003961</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-08</td>\n",
       "      <td>12.149718</td>\n",
       "      <td>0.006456</td>\n",
       "      <td>0.004806</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-01-09</td>\n",
       "      <td>12.409650</td>\n",
       "      <td>0.047530</td>\n",
       "      <td>0.017515</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         EMA      MACD  MACD-EMA  MAVG 20  MAVG 200\n",
       "Ticker Date                                                        \n",
       "AAPL   2007-01-03  11.971400  0.000000  0.000000      NaN       NaN\n",
       "       2007-01-04  12.110893  0.005961  0.003312      NaN       NaN\n",
       "       2007-01-05  12.125253  0.004895  0.003961      NaN       NaN\n",
       "       2007-01-08  12.149718  0.006456  0.004806      NaN       NaN\n",
       "       2007-01-09  12.409650  0.047530  0.017515      NaN       NaN"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_price_signals.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Volume Signals\n",
    "\n",
    "This is how easily we may calculate the volume signals:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'volume_signals-0eb21bf8.pickle' not on disk.\n",
      "- Running function volume_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 164 ms, sys: 12 ms, total: 176 ms\n",
      "Wall time: 173 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_vol_signals = hub.volume_signals(window=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we call the same function with a different argument, then the data-hub automatically changes the cache-name so as to keep the different results in separate cache-files:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'volume_signals-6bc2fe28.pickle' not on disk.\n",
      "- Running function volume_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 141 ms, sys: 7.76 ms, total: 149 ms\n",
      "Wall time: 145 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_vol_signals2 = hub.volume_signals(window=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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></th>\n",
       "      <th>Relative Volume</th>\n",
       "      <th>Volume Market-Cap</th>\n",
       "      <th>Volume Turnover</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\n",
       "      <th>Date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2007-01-31</td>\n",
       "      <td>-0.463248</td>\n",
       "      <td>6.263896e+08</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-02-01</td>\n",
       "      <td>-0.695482</td>\n",
       "      <td>6.142787e+08</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-02-02</td>\n",
       "      <td>-0.753584</td>\n",
       "      <td>6.092018e+08</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-02-05</td>\n",
       "      <td>-0.814985</td>\n",
       "      <td>6.034860e+08</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2007-02-06</td>\n",
       "      <td>-0.416573</td>\n",
       "      <td>6.046619e+08</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Relative Volume  Volume Market-Cap  Volume Turnover\n",
       "Ticker Date                                                           \n",
       "AAPL   2007-01-31        -0.463248       6.263896e+08              NaN\n",
       "       2007-02-01        -0.695482       6.142787e+08              NaN\n",
       "       2007-02-02        -0.753584       6.092018e+08              NaN\n",
       "       2007-02-05        -0.814985       6.034860e+08              NaN\n",
       "       2007-02-06        -0.416573       6.046619e+08              NaN"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_vol_signals.dropna(how='all').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Financial Signals\n",
    "\n",
    "The following shows how easily we may calculate financial signals such as Current Ratio, Debt Ratio, Net Profit Margin, etc. The function's argument `variant` determines the frequency of the resulting data-points:\n",
    "\n",
    "- `variant=daily` means the result has daily data-points for the same dates as the share-prices. By default this uses forward-filling with the last-known values, but this can be changed with the `fill_method` argument when constructing the data-hub.\n",
    "\n",
    "- `variant=latest` means the result only has data-points for the last date of the share-prices. This is much faster to compute than daily data-points.\n",
    "\n",
    "- `variant=quarterly` means the result has 4 data-points per year for the same dates as the company's quarterly financial reports (with the report date shifted if the `offset` argument was given when the data-hub object was created)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset \"us-balance-ttm\" on disk (6 days old).\n",
      "- Loading from disk ... Done!\n",
      "Cache-file 'fin_signals-c549b5bd.pickle' not on disk.\n",
      "- Running function fin_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 408 ms, sys: 4.2 ms, total: 412 ms\n",
      "Wall time: 408 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_fin_signals_daily = hub.fin_signals(variant='daily')"
   ]
  },
  {
   "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></th>\n",
       "      <th>Current Ratio</th>\n",
       "      <th>Debt Ratio</th>\n",
       "      <th>Gross Profit Margin</th>\n",
       "      <th>Net Profit Margin</th>\n",
       "      <th>Return on Assets</th>\n",
       "      <th>Return on Equity</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\n",
       "      <th>Date</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",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2010-06-01</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.4132</td>\n",
       "      <td>0.21149</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-02</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.4132</td>\n",
       "      <td>0.21149</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-03</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.4132</td>\n",
       "      <td>0.21149</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-04</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.4132</td>\n",
       "      <td>0.21149</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-07</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.4132</td>\n",
       "      <td>0.21149</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Current Ratio  Debt Ratio  Gross Profit Margin  \\\n",
       "Ticker Date                                                         \n",
       "AAPL   2010-06-01       2.644206         NaN               0.4132   \n",
       "       2010-06-02       2.644206         NaN               0.4132   \n",
       "       2010-06-03       2.644206         NaN               0.4132   \n",
       "       2010-06-04       2.644206         NaN               0.4132   \n",
       "       2010-06-07       2.644206         NaN               0.4132   \n",
       "\n",
       "                   Net Profit Margin  Return on Assets  Return on Equity  \n",
       "Ticker Date                                                               \n",
       "AAPL   2010-06-01            0.21149               NaN               NaN  \n",
       "       2010-06-02            0.21149               NaN               NaN  \n",
       "       2010-06-03            0.21149               NaN               NaN  \n",
       "       2010-06-04            0.21149               NaN               NaN  \n",
       "       2010-06-07            0.21149               NaN               NaN  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_fin_signals_daily.dropna(how='all').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Growth Signals\n",
    "\n",
    "The following shows how easily we may calculate growth-signals. This function also takes the argument `variant` which has the same meaning as for the `fin_signals` function above. For example, when `variant=quarterly` it creates 4 data-points per yer:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset \"us-income-quarterly\" on disk (0 days old).\n",
      "- Loading from disk ... Done!\n",
      "Dataset \"us-cashflow-ttm\" on disk (0 days old).\n",
      "- Loading from disk ... Done!\n",
      "Dataset \"us-cashflow-quarterly\" on disk (26 days old).\n",
      "- Loading from disk ... Done!\n",
      "Cache-file 'growth_signals-70589cd4.pickle' not on disk.\n",
      "- Running function growth_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 1.01 s, sys: 19.7 ms, total: 1.03 s\n",
      "Wall time: 1.02 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_growth_signals_qrt = hub.growth_signals(variant='quarterly')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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></th>\n",
       "      <th>Earnings Growth</th>\n",
       "      <th>Earnings Growth QOQ</th>\n",
       "      <th>Earnings Growth YOY</th>\n",
       "      <th>FCF Growth</th>\n",
       "      <th>FCF Growth QOQ</th>\n",
       "      <th>FCF Growth YOY</th>\n",
       "      <th>Sales Growth</th>\n",
       "      <th>Sales Growth QOQ</th>\n",
       "      <th>Sales Growth YOY</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\n",
       "      <th>Report 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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2009-11-29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.385120</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.320320</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.254058</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-03-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.334123</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.047005</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.284755</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-05-30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.089994</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.624259</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-0.139259</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-08-29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.058230</td>\n",
       "      <td>0.779540</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.057664</td>\n",
       "      <td>1.089590</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.163049</td>\n",
       "      <td>0.612903</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-11-29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.324316</td>\n",
       "      <td>0.701422</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.166467</td>\n",
       "      <td>0.846096</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.295732</td>\n",
       "      <td>0.666503</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Earnings Growth  Earnings Growth QOQ  Earnings Growth YOY  \\\n",
       "Ticker Report Date                                                              \n",
       "AAPL   2009-11-29               NaN             0.385120                  NaN   \n",
       "       2010-03-01               NaN             0.334123                  NaN   \n",
       "       2010-05-30               NaN            -0.089994                  NaN   \n",
       "       2010-08-29               NaN             0.058230             0.779540   \n",
       "       2010-11-29               NaN             0.324316             0.701422   \n",
       "\n",
       "                    FCF Growth  FCF Growth QOQ  FCF Growth YOY  Sales Growth  \\\n",
       "Ticker Report Date                                                             \n",
       "AAPL   2009-11-29          NaN        0.320320             NaN           NaN   \n",
       "       2010-03-01          NaN        1.047005             NaN           NaN   \n",
       "       2010-05-30          NaN       -0.624259             NaN           NaN   \n",
       "       2010-08-29          NaN        1.057664        1.089590           NaN   \n",
       "       2010-11-29          NaN        0.166467        0.846096           NaN   \n",
       "\n",
       "                    Sales Growth QOQ  Sales Growth YOY  \n",
       "Ticker Report Date                                      \n",
       "AAPL   2009-11-29           0.254058               NaN  \n",
       "       2010-03-01           0.284755               NaN  \n",
       "       2010-05-30          -0.139259               NaN  \n",
       "       2010-08-29           0.163049          0.612903  \n",
       "       2010-11-29           0.295732          0.666503  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_growth_signals_qrt.dropna(how='all').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we change the argument to `variant='daily'` then we get daily data-points where the growth-rates have been forward-filled from the last-known values (this may be changed by setting the argument `fill_method` when constructing the data-hub object). Note that the cache-filename is different to keep the files separate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'growth_signals-c549b5bd.pickle' not on disk.\n",
      "- Running function growth_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 187 ms, sys: 4.5 ms, total: 192 ms\n",
      "Wall time: 186 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_growth_signals_daily = hub.growth_signals(variant='daily')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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></th>\n",
       "      <th>Earnings Growth</th>\n",
       "      <th>Earnings Growth QOQ</th>\n",
       "      <th>Earnings Growth YOY</th>\n",
       "      <th>FCF Growth</th>\n",
       "      <th>FCF Growth QOQ</th>\n",
       "      <th>FCF Growth YOY</th>\n",
       "      <th>Sales Growth</th>\n",
       "      <th>Sales Growth QOQ</th>\n",
       "      <th>Sales Growth YOY</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2009-11-30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32032</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.254058</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2009-12-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32032</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.254058</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2009-12-02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32032</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.254058</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2009-12-03</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32032</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.254058</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2009-12-04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38512</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32032</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.254058</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Earnings Growth  Earnings Growth QOQ  Earnings Growth YOY  \\\n",
       "Ticker Date                                                                    \n",
       "AAPL   2009-11-30              NaN              0.38512                  NaN   \n",
       "       2009-12-01              NaN              0.38512                  NaN   \n",
       "       2009-12-02              NaN              0.38512                  NaN   \n",
       "       2009-12-03              NaN              0.38512                  NaN   \n",
       "       2009-12-04              NaN              0.38512                  NaN   \n",
       "\n",
       "                   FCF Growth  FCF Growth QOQ  FCF Growth YOY  Sales Growth  \\\n",
       "Ticker Date                                                                   \n",
       "AAPL   2009-11-30         NaN         0.32032             NaN           NaN   \n",
       "       2009-12-01         NaN         0.32032             NaN           NaN   \n",
       "       2009-12-02         NaN         0.32032             NaN           NaN   \n",
       "       2009-12-03         NaN         0.32032             NaN           NaN   \n",
       "       2009-12-04         NaN         0.32032             NaN           NaN   \n",
       "\n",
       "                   Sales Growth QOQ  Sales Growth YOY  \n",
       "Ticker Date                                            \n",
       "AAPL   2009-11-30          0.254058               NaN  \n",
       "       2009-12-01          0.254058               NaN  \n",
       "       2009-12-02          0.254058               NaN  \n",
       "       2009-12-03          0.254058               NaN  \n",
       "       2009-12-04          0.254058               NaN  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_growth_signals_daily.dropna(how='all').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Valuation Signals\n",
    "\n",
    "The following shows how easily we may calculate valuation signals such as P/E and P/Sales ratios. This requires several datasets, but most of them have already been loaded by the data-hub in the function-calls we ran above, so the datasets are merely taken from the RAM-cache instead of being loaded again.\n",
    "\n",
    "By setting `variant='latest'` we only calculate the valuation signals for the latest day of the share-prices. This is much faster than calculating the signals for daily data-points, and is useful e.g. when making a stock-screener:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset \"us-shareprices-latest\" on disk (2 days old).\n",
      "- Loading from disk ... Done!\n",
      "Cache-file 'val_signals-cac8146c.pickle' not on disk.\n",
      "- Running function val_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 68.1 ms, sys: 284 µs, total: 68.4 ms\n",
      "Wall time: 66.2 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_val_signals_latest = hub.val_signals(variant='latest')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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></th>\n",
       "      <th>Dividend Yield</th>\n",
       "      <th>Earnings Yield</th>\n",
       "      <th>FCF Yield</th>\n",
       "      <th>Market-Cap</th>\n",
       "      <th>P/Book</th>\n",
       "      <th>P/E</th>\n",
       "      <th>P/FCF</th>\n",
       "      <th>P/NCAV</th>\n",
       "      <th>P/NetNet</th>\n",
       "      <th>P/Sales</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>AAPL</td>\n",
       "      <td>2019-12-20</td>\n",
       "      <td>0.010868</td>\n",
       "      <td>0.042534</td>\n",
       "      <td>0.045336</td>\n",
       "      <td>1.299092e+12</td>\n",
       "      <td>14.356514</td>\n",
       "      <td>23.510429</td>\n",
       "      <td>22.057394</td>\n",
       "      <td>-15.245951</td>\n",
       "      <td>-10.131468</td>\n",
       "      <td>4.993167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>AMZN</td>\n",
       "      <td>2019-12-20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.012640</td>\n",
       "      <td>0.022334</td>\n",
       "      <td>8.977162e+11</td>\n",
       "      <td>15.886534</td>\n",
       "      <td>79.114854</td>\n",
       "      <td>44.773878</td>\n",
       "      <td>-14.129031</td>\n",
       "      <td>-11.637230</td>\n",
       "      <td>3.381636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>MSFT</td>\n",
       "      <td>2019-12-20</td>\n",
       "      <td>0.011575</td>\n",
       "      <td>0.033733</td>\n",
       "      <td>0.031717</td>\n",
       "      <td>1.218196e+12</td>\n",
       "      <td>11.485805</td>\n",
       "      <td>29.644133</td>\n",
       "      <td>31.528443</td>\n",
       "      <td>-174.077735</td>\n",
       "      <td>-59.044724</td>\n",
       "      <td>9.384165</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Dividend Yield  Earnings Yield  FCF Yield    Market-Cap  \\\n",
       "Ticker Date                                                                  \n",
       "AAPL   2019-12-20        0.010868        0.042534   0.045336  1.299092e+12   \n",
       "AMZN   2019-12-20             NaN        0.012640   0.022334  8.977162e+11   \n",
       "MSFT   2019-12-20        0.011575        0.033733   0.031717  1.218196e+12   \n",
       "\n",
       "                      P/Book        P/E      P/FCF      P/NCAV   P/NetNet  \\\n",
       "Ticker Date                                                                 \n",
       "AAPL   2019-12-20  14.356514  23.510429  22.057394  -15.245951 -10.131468   \n",
       "AMZN   2019-12-20  15.886534  79.114854  44.773878  -14.129031 -11.637230   \n",
       "MSFT   2019-12-20  11.485805  29.644133  31.528443 -174.077735 -59.044724   \n",
       "\n",
       "                    P/Sales  \n",
       "Ticker Date                  \n",
       "AAPL   2019-12-20  4.993167  \n",
       "AMZN   2019-12-20  3.381636  \n",
       "MSFT   2019-12-20  9.384165  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_val_signals_latest.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We may also calculate the valuation signals for daily data-points, simply by setting the argument `variant='daily'`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'val_signals-a6127cfd.pickle' not on disk.\n",
      "- Running function val_signals() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 222 ms, sys: 11.8 ms, total: 234 ms\n",
      "Wall time: 229 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_val_signals_daily = hub.val_signals(variant='daily')"
   ]
  },
  {
   "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></th>\n",
       "      <th>Dividend Yield</th>\n",
       "      <th>Earnings Yield</th>\n",
       "      <th>FCF Yield</th>\n",
       "      <th>Market-Cap</th>\n",
       "      <th>P/Book</th>\n",
       "      <th>P/E</th>\n",
       "      <th>P/FCF</th>\n",
       "      <th>P/NCAV</th>\n",
       "      <th>P/NetNet</th>\n",
       "      <th>P/Sales</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2010-06-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.045227</td>\n",
       "      <td>0.050468</td>\n",
       "      <td>2.390632e+11</td>\n",
       "      <td>6.075612</td>\n",
       "      <td>22.110910</td>\n",
       "      <td>19.814601</td>\n",
       "      <td>16.343964</td>\n",
       "      <td>30.148580</td>\n",
       "      <td>4.676235</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.044692</td>\n",
       "      <td>0.049871</td>\n",
       "      <td>2.419227e+11</td>\n",
       "      <td>6.148285</td>\n",
       "      <td>22.375389</td>\n",
       "      <td>20.051612</td>\n",
       "      <td>16.539461</td>\n",
       "      <td>30.509200</td>\n",
       "      <td>4.732170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-03</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.044833</td>\n",
       "      <td>0.050029</td>\n",
       "      <td>2.411624e+11</td>\n",
       "      <td>6.128963</td>\n",
       "      <td>22.305071</td>\n",
       "      <td>19.988597</td>\n",
       "      <td>16.487484</td>\n",
       "      <td>30.413320</td>\n",
       "      <td>4.717298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.046086</td>\n",
       "      <td>0.051427</td>\n",
       "      <td>2.346042e+11</td>\n",
       "      <td>5.962289</td>\n",
       "      <td>21.698497</td>\n",
       "      <td>19.445019</td>\n",
       "      <td>16.039116</td>\n",
       "      <td>29.586248</td>\n",
       "      <td>4.589014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-07</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.047009</td>\n",
       "      <td>0.052457</td>\n",
       "      <td>2.299989e+11</td>\n",
       "      <td>5.845249</td>\n",
       "      <td>21.272555</td>\n",
       "      <td>19.063312</td>\n",
       "      <td>15.724268</td>\n",
       "      <td>29.005469</td>\n",
       "      <td>4.498931</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Dividend Yield  Earnings Yield  FCF Yield    Market-Cap  \\\n",
       "Ticker Date                                                                  \n",
       "AAPL   2010-06-01             NaN        0.045227   0.050468  2.390632e+11   \n",
       "       2010-06-02             NaN        0.044692   0.049871  2.419227e+11   \n",
       "       2010-06-03             NaN        0.044833   0.050029  2.411624e+11   \n",
       "       2010-06-04             NaN        0.046086   0.051427  2.346042e+11   \n",
       "       2010-06-07             NaN        0.047009   0.052457  2.299989e+11   \n",
       "\n",
       "                     P/Book        P/E      P/FCF     P/NCAV   P/NetNet  \\\n",
       "Ticker Date                                                               \n",
       "AAPL   2010-06-01  6.075612  22.110910  19.814601  16.343964  30.148580   \n",
       "       2010-06-02  6.148285  22.375389  20.051612  16.539461  30.509200   \n",
       "       2010-06-03  6.128963  22.305071  19.988597  16.487484  30.413320   \n",
       "       2010-06-04  5.962289  21.698497  19.445019  16.039116  29.586248   \n",
       "       2010-06-07  5.845249  21.272555  19.063312  15.724268  29.005469   \n",
       "\n",
       "                    P/Sales  \n",
       "Ticker Date                  \n",
       "AAPL   2010-06-01  4.676235  \n",
       "       2010-06-02  4.732170  \n",
       "       2010-06-03  4.717298  \n",
       "       2010-06-04  4.589014  \n",
       "       2010-06-07  4.498931  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_val_signals_daily.dropna(how='all').head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate Returns\n",
    "\n",
    "The data-hub also provides functions for easily calculating stock-returns for different investment durations. Let us first calculate the annualized future stock-returns for all 3-year investment periods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Name of the new Pandas Series with the stock-returns.\n",
    "TOTAL_RETURN_3Y = 'Total Return 3-Year'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'rel_change-2ab9d7a6.pickle' not on disk.\n",
      "- Running function rel_change() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 15.4 ms, sys: 0 ns, total: 15.4 ms\n",
      "Wall time: 14 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_returns_3y = hub.returns(name=TOTAL_RETURN_3Y, years=3,\n",
    "                            future=True, annualized=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For example, the calculated return for AAPL on 2007-01-03 is 0.359641, which means that between 2007-01-03 and 2010-01-03 (or the closest trading-day), the annualized return was about 35.96% per year. This is the so-called Total Return which assumes dividends are reinvested without having to pay any taxes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date\n",
       "2007-01-03    0.359641\n",
       "2007-01-04    0.356789\n",
       "2007-01-05    0.360885\n",
       "2007-01-08    0.351568\n",
       "2007-01-09    0.314955\n",
       "Name: Total Return 3-Year, dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_returns_3y.loc['AAPL']['2007-01-03':'2007-01-09']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also plot these 3-year annualized returns for the AAPL stock:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb76ffd3ef0>"
      ]
     },
     "execution_count": 34,
     "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": [
    "df_returns_3y.loc['AAPL'].dropna().plot(grid=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mean-Log Returns\n",
    "\n",
    "It is often useful to consider the average returns e.g. over 1-3 year periods. It is much faster to compute the mean-log returns, and the results are close to the mean returns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Name of the new Pandas Series with the stock-returns.\n",
    "TOTAL_RETURN_1_3Y = 'Total Return 1-3 Years'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cache-file 'mean_log_change-257ab99e.pickle' not on disk.\n",
      "- Running function mean_log_change() ... Done!\n",
      "- Saving cache-file to disk ... Done!\n",
      "CPU times: user 71.1 ms, sys: 199 µs, total: 71.3 ms\n",
      "Wall time: 69.3 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "df_returns_1_3y = \\\n",
    "    hub.mean_log_returns(name=TOTAL_RETURN_1_3Y,\n",
    "                         min_years=1, max_years=3,\n",
    "                         future=True, annualized=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For ticker AAPL, the mean-log return was 0.277523 for all 1-3 year investment periods between 2007-01-03 and 2010-01-03. This means that if you had bought AAPL stock on 2007-01-03 and sold them somewhere between 2008-01-03 and 2010-01-03, then on average you would have made roughly 28% per year, depending on exactly when you had sold the stocks in this 1-3 year period. Note that the mean-log return corresponds roughly to the (non-log) mean returns for values in this range."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Date\n",
       "2007-01-03    0.277523\n",
       "2007-01-04    0.265312\n",
       "2007-01-05    0.269260\n",
       "2007-01-08    0.266535\n",
       "2007-01-09    0.222106\n",
       "Name: Total Return 1-3 Years, dtype: float64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_returns_1_3y.loc['AAPL']['2007-01-03':'2007-01-09']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also plot the mean-log returns for 1-3 year investment periods for the AAPL stock:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb76f35a080>"
      ]
     },
     "execution_count": 38,
     "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": [
    "df_returns_1_3y.loc['AAPL'].dropna().plot(grid=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Combine Signals and Returns\n",
    "\n",
    "Let us now combine the stock-returns and the signals we are interested in, into a single DataFrame that is easier to work with:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# List of all the DataFrames with different signals.\n",
    "dfs = [df_price_signals, df_fin_signals_daily,\n",
    "       df_growth_signals_daily, df_val_signals_daily]\n",
    "\n",
    "# Combine all the signal DataFrames into a single DataFrame.\n",
    "df_all_signals = pd.concat(dfs, axis=1)\n",
    "\n",
    "# DataFrame with the signals we are interested in, for example.\n",
    "signals = [PE, PSALES, CURRENT_RATIO, MAVG_20]\n",
    "df_signals = df_all_signals[signals]\n",
    "\n",
    "# DataFrame with the signals and stock-returns.\n",
    "dfs = [df_signals, df_returns_3y, df_returns_1_3y]\n",
    "df_combined = pd.concat(dfs, axis=1)\n",
    "\n",
    "# Remove rows with any NaN values.\n",
    "df_combined = df_combined.dropna(how='any')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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></th>\n",
       "      <th>P/E</th>\n",
       "      <th>P/Sales</th>\n",
       "      <th>Current Ratio</th>\n",
       "      <th>MAVG 20</th>\n",
       "      <th>Total Return 3-Year</th>\n",
       "      <th>Total Return 1-3 Years</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ticker</th>\n",
       "      <th>Date</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",
       "      <td rowspan=\"5\" valign=\"top\">AAPL</td>\n",
       "      <td>2010-06-01</td>\n",
       "      <td>22.110910</td>\n",
       "      <td>4.676235</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>35.883590</td>\n",
       "      <td>0.207653</td>\n",
       "      <td>0.314334</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-02</td>\n",
       "      <td>22.375389</td>\n",
       "      <td>4.732170</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>35.921230</td>\n",
       "      <td>0.203691</td>\n",
       "      <td>0.308274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-03</td>\n",
       "      <td>22.305071</td>\n",
       "      <td>4.717298</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>35.972195</td>\n",
       "      <td>0.203844</td>\n",
       "      <td>0.310623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-04</td>\n",
       "      <td>21.698497</td>\n",
       "      <td>4.589014</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>36.041585</td>\n",
       "      <td>0.211123</td>\n",
       "      <td>0.326294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2010-06-07</td>\n",
       "      <td>21.272555</td>\n",
       "      <td>4.498931</td>\n",
       "      <td>2.644206</td>\n",
       "      <td>36.149300</td>\n",
       "      <td>0.213025</td>\n",
       "      <td>0.337643</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         P/E   P/Sales  Current Ratio    MAVG 20  \\\n",
       "Ticker Date                                                        \n",
       "AAPL   2010-06-01  22.110910  4.676235       2.644206  35.883590   \n",
       "       2010-06-02  22.375389  4.732170       2.644206  35.921230   \n",
       "       2010-06-03  22.305071  4.717298       2.644206  35.972195   \n",
       "       2010-06-04  21.698497  4.589014       2.644206  36.041585   \n",
       "       2010-06-07  21.272555  4.498931       2.644206  36.149300   \n",
       "\n",
       "                   Total Return 3-Year  Total Return 1-3 Years  \n",
       "Ticker Date                                                     \n",
       "AAPL   2010-06-01             0.207653                0.314334  \n",
       "       2010-06-02             0.203691                0.308274  \n",
       "       2010-06-03             0.203844                0.310623  \n",
       "       2010-06-04             0.211123                0.326294  \n",
       "       2010-06-07             0.213025                0.337643  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_combined.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can then plot e.g. the P/Sales ratios versus the future 1-3 year stock-returns:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb76d036f28>"
      ]
     },
     "execution_count": 41,
     "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": [
    "sns.scatterplot(x=PSALES, y=TOTAL_RETURN_1_3Y, hue=TICKER,\n",
    "                data=df_combined.reset_index())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### NumPy Arrays\n",
    "\n",
    "If you want to use the signals and stock-returns with functions that accept NumPy arrays instead of Pandas DataFrames, it is very easy to convert from Pandas to NumPy, because Pandas uses NumPy internally:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the numpy array with the signals and stock-returns\n",
    "# from the Pandas DataFrame.\n",
    "arr = df_combined.values\n",
    "\n",
    "# Number of signals in the DataFrame.\n",
    "num_signals = 4\n",
    "\n",
    "# Numpy array with the signals.\n",
    "arr_signals = arr[:, :num_signals]\n",
    "\n",
    "# Numpy array with the returns.\n",
    "arr_returns = arr[:, num_signals:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the NumPy array with the signals that we are interested in:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[22.11091028,  4.676235  ,  2.64420639, 35.88359   ],\n",
       "       [22.37538861,  4.73216951,  2.64420639, 35.92123   ],\n",
       "       [22.30507072,  4.71729798,  2.64420639, 35.972195  ],\n",
       "       ...,\n",
       "       [23.28061866,  5.35009978,  2.68507057, 61.045     ],\n",
       "       [23.77388378,  5.46345663,  2.68507057, 61.2085    ],\n",
       "       [23.74398892,  5.45658652,  2.68507057, 61.3425    ]])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arr_signals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4959, 4)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arr_signals.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the NumPy array with the two kinds of stock-returns that we calculated:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.20765287, 0.31433411],\n",
       "       [0.2036908 , 0.30827427],\n",
       "       [0.20384442, 0.31062266],\n",
       "       ...,\n",
       "       [0.37773349, 0.32155831],\n",
       "       [0.37209097, 0.31072557],\n",
       "       [0.37767608, 0.31214757]])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arr_returns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4959, 2)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arr_returns.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Because the two NumPy arrays have the same number of rows, they can easily be fed into different Machine Learning libraries, which we might do in a future tutorial."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## License (MIT)\n",
    "\n",
    "This is published under the\n",
    "[MIT License](https://github.com/simfin/simfin-tutorials/blob/master/LICENSE.txt)\n",
    "which allows very broad use for both academic and commercial purposes.\n",
    "\n",
    "You are very welcome to modify and use this source-code in your own project. Please keep a link to the [original repository](https://github.com/simfin/simfin-tutorials).\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}