Backtest With Jupyter Notebook Anaconda 3.5.0.1
Just walkthrough the whole process of creating Python backtest in Anaconda, Windows 10.
The process is broken down into 5 major steps:
00x | Install Anaconda
10x | Get online stock (e.g. AAPL) | Market Data
20x | Create trade strategy | Strategy Class
30x | Generate trade signal | Signal Class
40x | Create investment portfolio | Portfolio Class
50x | Perfomance review | PnL + Graphing
A timeline is provided below for easy navigation:
In case you just want to use the Anaconda root setup, skip step 'Ana002'
mm:ss
00:00 - 03:02 ||Ana001| Install Anaconda
03:04 - 10:02 ||Ana002|Setup different environment other than root - optional step*
10:04 - 19:02 ||Ana101| Get Apple stock data from Yahoo Finance, save to local drive as Excel
19:04 - 29:11 ||Ana201| Create Strategy - calculate moving averages
29:12 - 35:53 ||Ana202| Create Strategy - get stock trend
35:55 - 41:22 ||Ana301| Get Trade Signal and Generate Order
41:24 - 49:50 ||Ana401402| Create Investment Portfolio and Trade Details
49:52 - 59:03 ||Ana501502| Calculate Profit and Loss and Graphing
* install optional Python packages that your project requires
** the strategy mentioned assumes zero borrowing cost so that whenever short position occurs,
we can borrow unlimited amount of money without needing to pay a penny of interest.
This is not real in real world. So just be aware of that.
Here's the 59-min video(Complete Walkthrough - Jupyter Notebook Anaconda):
*** this video is rendered in 1.05x speed; feel free to visit the original speed videos below.
Install Anaconda 3.5.0.1(Ana001)
Setup Environment In Anaconda 3.5.0.1(Ana002) - Optional
Get Online Stock Data + Save To Local Drive(Ana101)
Form Trade Strategy - Calculate Moving Average (Ana201)
Form Trade Strategy - Get Stock Trend (Ana202)
Generate Trade Signal - Get Stock Trend (Ana301)
Create Investment Portfolio (Ana401402)
Measuring Performance & Graphing (Ana501502)
Backtest With Python Default Editor
My Thought
Comparing to buidling Python strategy backtest in default Python editor,the Anaconda Jupyter Notebook takes a bit of time to get used to.
It takes time to load up the Jupyter Notebook while such load time is minimal in Python default editor.
But from layman's perspective, setting up multiple environments and installing packages
through the Anaconda Navigator are intuitive.
Anaconda basically organizes the folders under the 'env' folder.
That's good folder management.
In the future, I'll try using Sublime Text 3 to route to the Anaconda's environment folder.
See if I can take advantage of Sublime Text' speed while maintaining proper environment folder
management.
If you find this article helpful, feel free to discuss on my Facebook:
www.facebook.com/clueple
Disclaimer:
Everything shown in this video is just the author's personal experience with Anaconda.
All mentioned assumptions and tactics are used as demonstration for education purpose.
Trade details, not limited to, leverage level, deposit amount, transactions fees, and
any costs associated with the strategy are not meant to be complete.
No investment decision should be based on any information provided in this video.
No professional advice are provided or implied whatsoever.
Viewers are solely responsible for any loss associated to consumption of this video.
No comments:
Post a Comment