g***e
发帖数: 577 | 1 【 以下文字转载自 Stock 讨论区 】
发信人: geome (老网友), 信区: Stock
标 题: free back test tool in python - for quant algos
发信站: BBS 未名空间站 (Sat Mar 4 15:10:49 2017, 美东)
Hi,
Our team has made a free back test tool in below github:
https://github.com/geome-mitbbs/QTS_Research
It is simple to use yet has great flexibility:
To use:
1. System requirements: you need to have python3 + numpy + pandas + (
matplotlib,optional)
2. Download files into any folder.
3. Modify Sample_Test.py and run. Currently Sample_Test.py contains a moving
average trend following algo: if 10 day average > 25 day average, buy. if 10
day average < 25 day average, sell. you can either test this algo with your
own parameters ( ticker, average period etc ) or you can write your own
strategy. The strategy is simply a part of python code consists of the
functions in Quant_Indicators.py
Anyone interested in quant trading strategy research project/product. Please
email me in private. Thanks. | l*****t
发帖数: 1080 | | g***e
发帖数: 577 | 3 Thanks for asking.
I am posting here for 2 reasons.
1. For people willing to do open-source projects. This might be a good one
which can mix finance/python/machine learning etc to help you build skills
or broad eyesight.
2. People interested in the product. This will be interest for myself.
So I guess it somehow helps the 1st group.
【在 l*****t 的大作中提到】
: 大概看了一下，但是这个策略是对是错还未知啊。
| g***e
发帖数: 577 | 4 This is just a sample, it has not much analysis on the trend following
strategy but it shows how concise the code/platform is. (If you used similar
software or website you know what I mean ).
This is on-going project. We aim to add more research result, not just tool
to do research.
【在 l*****t 的大作中提到】
: 大概看了一下，但是这个策略是对是错还未知啊。
| r********g
发帖数: 868 | 5 系统要求里还需要quandl吧
【在 g***e 的大作中提到】
: This is just a sample, it has not much analysis on the trend following
: strategy but it shows how concise the code/platform is. (If you used similar
: software or website you know what I mean ).
: This is on-going project. We aim to add more research result, not just tool
: to do research.
|
|
