intro-algo

Introduction to algorithmic trading. This project is strictly for education purpose and is not financial advice.

Anaconda - Setting up a Python environment

Install Anaconda, then open a Anaconda Prompt to create a Python 3.7 environment with the name py37-smu by:

conda create -n py37-smu python=3.7

Multiple environments of higher or lower version can be created for other projects. For now we will use this 3.7 environment by activating it:

conda activate py37-smu

To see a list of Anaconda environments and the installation directory:

conda env list
or
conda info --envs

The installation directory contains the python.exe, and we will need this path for configuring PyCharm later. Next install the following packages:

conda install numpy
conda install matplotlib
conda install -c conda-forge python-dotenv
conda install -c anaconda requests
conda install -c conda-forge websockets
pip install python-gnupg
pip install dydx-v3-python
pip install web3

Please note that dydx and web3 requires C++ compiler, so download and install the community version from the official website: https://visualstudio.microsoft.com/downloads/

Github - Download the code

Install Git, then open a Git Shell to download this repository to your local computer by:

git clone https://github.com/aiver-workshop/intro-algo.git

This command will download the codes in a folder called intro-algo, where you will find a .\intro-algo\python subfolder we call the python working directory.

PyCharm - Development IDE

1. Open the project by File -> Open... -> select python working directory
2. Configure the Python interpreter by File -> Settings... -> Project: intro-algo -> Python Interpreter -> Add -> Conda Environment -> Existing environment -> select Anaconda environment directory of py37-smu

PyCharm - Testing framework

PyCharm supports multiple testing framework such as pytest and unittests. Please configure the default test runner by File -> Settings... -> Tools -> Python Integrated Tools -> Testing -> Default test runner -> Unittests

Environment variable - Secret info management

The simplest way and the wrong way to handle credentials is to hardcode it in our code.

A safe way to handle your secret keys/password is saving them in envirnoment variables. We create a folder c:\vault to store secret files with sensitive information like api credentials. The key/value pairs in the file will be read as environment variable when Python program runs. All the secret filename will start with a dot such as .keys.

⚠️ This folder contains sensitive information! Never share or check-in the files

In script, we can read the values as follow:

import os
from dotenv import load_dotenv

dotenv_path = 'c:/vault/.keys'
load_dotenv(dotenv_path=dotenv_path)
API_KEY = os.getenv('API_KEY')
API_SECRET = os.getenv('API_SECRET')

File encryption - GnuPG

An alternative way to protect sensitive data is to encrypt the file with a password that is only known to you. You are required to enter the password to decrypt and retrieve the credential in order to start the process. See lib\gpg_crpto.py for instruction on how to encrypt a file and to decrypt in Python.