{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Strategy Analysis with **Pandas TA** and AI/ML\n",
"* This is a **Work in Progress** and subject to change!\n",
"* Contributions are welcome and accepted!\n",
"* Examples below are for **educational purposes only**."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Required Packages\n",
"##### Uncomment the packages you need to install or are missing"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"#!pip install numpy\n",
"#!pip install pandas\n",
"#!pip install mplfinance\n",
"#!pip install pandas-datareader\n",
"#!pip install requests_cache\n",
"#!pip install alphaVantage-api"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Populating the interactive namespace from numpy and matplotlib\n",
"Numpy v1.18.3\n",
"Pandas v1.1.0\n",
"mplfinance v0.12.6a3\n",
"Pandas TA v0.2.12b\n"
]
}
],
"source": [
"%pylab inline\n",
"import datetime as dt\n",
"import random as rnd\n",
"from sys import float_info as sflt\n",
"\n",
"import numpy as np\n",
"import pandas as pd\n",
"pd.set_option('max_rows', 100)\n",
"pd.set_option('max_columns', 20)\n",
"\n",
"import mplfinance as mpf\n",
"\n",
"from alphaVantageAPI.alphavantage import AlphaVantage\n",
"import pandas_ta as ta\n",
"\n",
"from watchlist import colors, Watchlist\n",
"\n",
"print(f\"Numpy v{np.__version__}\")\n",
"print(f\"Pandas v{pd.__version__}\")\n",
"print(f\"mplfinance v{mpf.__version__}\")\n",
"print(f\"Pandas TA v{ta.version}\")\n",
"%matplotlib inline"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## MISC Functions"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"def recent_bars(df, tf: str = \"1y\"):\n",
" # All Data: 0, Last Four Years: 0.25, Last Two Years: 0.5, This Year: 1, Last Half Year: 2, Last Quarter: 4\n",
" yearly_divisor = {\"all\": 0, \"10y\": 0.1, \"5y\": 0.2, \"4y\": 0.25, \"3y\": 1./3, \"2y\": 0.5, \"1y\": 1, \"6mo\": 2, \"3mo\": 4}\n",
" yd = yearly_divisor[tf] if tf in yearly_divisor.keys() else 0\n",
" return int(ta.RATE[\"TRADING_DAYS_PER_YEAR\"] / yd) if yd > 0 else df.shape[0]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Collect some Data"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[!] Loading All: SPY, QQQ, AAPL, TSLA\n",
"[i] Loaded['D']: SPY_D.csv\n",
"[i] Runtime: 1812.7836 ms (1.8128 s)\n",
"[i] Loaded['D']: QQQ_D.csv\n",
"[i] Runtime: 1747.0058 ms (1.7470 s)\n",
"[i] Loaded['D']: AAPL_D.csv\n",
"[i] Runtime: 1105.7967 ms (1.1058 s)\n",
"[i] Loaded['D']: TSLA_D.csv\n",
"[i] Runtime: 906.7782 ms (0.9068 s)\n"
]
}
],
"source": [
"tf = \"D\"\n",
"tickers = [\"SPY\", \"QQQ\", \"AAPL\", \"TSLA\"]\n",
"watch = Watchlist(tickers, tf=tf)\n",
"watch.strategy = ta.CommonStrategy\n",
"watch.load(tickers, timed=True, analyze=True, verbose=False)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Select an Asset"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"AAPL (5241, 10)\n",
"Columns: open, high, low, close, volume, SMA_10, SMA_20, SMA_50, SMA_200, VOL_SMA_20\n"
]
}
],
"source": [
"ticker = tickers[2]\n",
"# watch.data[ticker].ta.constants(True, [0, 0, 0])\n",
"print(f\"{ticker} {watch.data[ticker].shape}\\nColumns: {', '.join(list(watch.data[ticker].columns))}\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Trim it"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"AAPL (252, 10)\n",
"Columns: open, high, low, close, volume, SMA_10, SMA_20, SMA_50, SMA_200, VOL_SMA_20\n"
]
}
],
"source": [
"duration = \"1y\"\n",
"recent = recent_bars(watch.data[ticker], duration)\n",
"asset = watch.data[ticker].copy().tail(recent)\n",
"print(f\"{ticker} {asset.shape}\\nColumns: {', '.join(list(asset.columns))}\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Create a Trend"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"date\n",
"2019-08-30 NaN\n",
"2019-09-03 NaN\n",
"2019-09-04 NaN\n",
"2019-09-05 NaN\n",
"2019-09-06 NaN\n",
" ... \n",
"2020-08-24 407.982918\n",
"2020-08-25 411.563980\n",
"2020-08-26 415.270883\n",
"2020-08-27 418.595162\n",
"2020-08-28 421.757313\n",
"Name: EMA_50, Length: 252, dtype: float64"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Example Long Trends\n",
"# long = ta.sma(asset.close, 10) < ta.sma(asset.close, 20) # SMA(10) > SMA(20)\n",
"# long = ta.ema(asset.close, 8) > ta.ema(asset.close, 21) # EMA(8) > EMA(21)\n",
"long = ta.increasing(ta.ema(asset.close, 50))\n",
"# long = ta.macd(asset.close).iloc[:,1] > 0 # MACD Histogram is positive\n",
"\n",
"asset.ta.ema(length=8, append=True)\n",
"asset.ta.ema(length=21, append=True)\n",
"asset.ta.ema(length=50, append=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Calculate Trend Returns from the long trend"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" CLTR | \n",
" TR_LOGRET | \n",
" CLTR_Trends | \n",
" CLTR_Trades | \n",
"
\n",
" \n",
" | date | \n",
" | \n",
" | \n",
" | \n",
" | \n",
"
\n",
" \n",
" \n",
" \n",
" | 2020-08-24 | \n",
" 0.629078 | \n",
" 0.011889 | \n",
" 1 | \n",
" 0 | \n",
"
\n",
" \n",
" | 2020-08-25 | \n",
" 0.620840 | \n",
" -0.008238 | \n",
" 1 | \n",
" 0 | \n",
"
\n",
" \n",
" | 2020-08-26 | \n",
" 0.634348 | \n",
" 0.013507 | \n",
" 1 | \n",
" 0 | \n",
"
\n",
" \n",
" | 2020-08-27 | \n",
" 0.622321 | \n",
" -0.012026 | \n",
" 1 | \n",
" 0 | \n",
"
\n",
" \n",
" | 2020-08-28 | \n",
" 0.620700 | \n",
" -0.001621 | \n",
" 1 | \n",
" 0 | \n",
"
\n",
" \n",
"
\n",
"