Technical Analysis Library in Python

A Technical Analysis library for financial time series datasets, useful for feature engineering. It is built on Pandas (Modin for speedup).

The library has implemented 31 indicators:

Volume

• Accumulation/Distribution Index (ADI)
• On-Balance Volume (OBV)
• Chaikin Money Flow (CMF)
• Force Index (FI)
• Ease of Movement (EoM, EMV)
• Volume-price Trend (VPT)
• Negative Volume Index (NVI)

Volatility

• Average True Range (ATR)
• Bollinger Bands (BB)
• Keltner Channel (KC)
• Donchian Channel (DC)

Trend

• Moving Average Convergence Divergence (MACD)
• Average Directional Movement Index (ADX)
• Vortex Indicator (VI)
• Trix (TRIX)
• Mass Index (MI)
• Commodity Channel Index (CCI)
• Detrended Price Oscillator (DPO)
• KST Oscillator (KST)
• Ichimoku Kinkō Hyō (Ichimoku)

Momentum

• Money Flow Index (MFI)
• Relative Strength Index (RSI)
• True strength index (TSI)
• Ultimate Oscillator (UO)
• Stochastic Oscillator (SR)
• Williams %R (WR)
• Awesome Oscillator (AO)

Others

• Daily Return (DR)
• Daily Log Return (DLR)
• Cumulative Return (CR)

Documentation

https://technical-analysis-library-in-python.readthedocs.io/en/latest/

Motivation to use

• English: https://towardsdatascience.com/technical-analysis-library-to-financial-datasets-with-pandas-python-4b2b390d3543
• Spanish: https://medium.com/datos-y-ciencia/biblioteca-de-an%C3%A1lisis-t%C3%A9cnico-sobre-series-temporales-financieras-para-machine-learning-con-cb28f9427d0

How to use (python 3.7)

$ pip install ta-modin

To use this library you should have a financial time series dataset including “Timestamp”, “Open”, “High”, “Low”, “Close” and “Volume” columns.

You should clean or fill NaN values in your dataset before add technical analysis features.

You can get code examples in examples_to_use folder.

You can visualize the features in this notebook.

Example adding all features

import platform

if platform.system() == 'Windows':
    import pandas as pd
else:
    import modin.pandas as pd

from ta import *

# Load datas
df = pd.read_csv('your-file.csv', sep=',')

# Clean NaN values
df = utils.dropna(df)

# Add ta features filling NaN values
df = add_all_ta_features(df, "Open", "High", "Low", "Close", "Volume_BTC", fillna=True)

Example adding individual features

import platform

if platform.system() == 'Windows':
    import pandas as pd
else:
    import modin.pandas as pd

from ta import *

# Load datas
df = pd.read_csv('your-file.csv', sep=',')

# Clean NaN values
df = utils.dropna(df)

# Add bollinger band high indicator filling NaN values
df['bb_high_indicator'] = bollinger_hband_indicator(df["Close"], n=20, ndev=2, fillna=True)

# Add bollinger band low indicator filling NaN values
df['bb_low_indicator'] = bollinger_lband_indicator(df["Close"], n=20, ndev=2, fillna=True)

Deploy for developers

$ git clone https://github.com/notadamking/ta-modin.git
$ cd ta-modin
$ pip install -r dev-requirements.txt
$ cd dev
$ python bollinger_band_features_example.py

Based on:

• https://en.wikipedia.org/wiki/Technical_analysis
• https://pandas.pydata.org
• https://github.com/FreddieWitherden/ta
• https://github.com/femtotrader/pandas_talib

Credits:

Developed by Darío López Padial (aka Bukosabino) and other contributors: https://github.com/bukosabino/ta/graphs/contributors

Please, let him know about any comments or feedback.