Меню

Portfolio / Open-source

Uberbacklog

I’m doing my research in Jupyter lab environment, and here is an Uberbacklog thing. It’s a catalog of small research sprints which typically take about 1-4 weeks each. Since 2018 I’ve implemented more than #150 sprints on various topics: US market, RU market, options, futures, volatility, risks, you name it…

I have my proprietary framework for backtesting, proprietary datafeed for futures/stocks and options, plethora of models and researches.

A picture is worth a thousand words

Open-source projects

Yet Another Universal Backtesting Engine Release (YAUBER) – Algo Lib

https://github.com/alexveden/yauber-algo

This package is a collection of standalone algorithms for financial time series analysis.

Highlights:

  • It heavily uses Numba for improving performance of the code. It’s generally faster than comparable Pandas algorithms.
  • It’s build based on strict principles: future reference free, consistent when starting point of history changes, NaN friendly, built-in data validation checks
  • It’s stable and well tested, this means that logic of algorithms won’t change in the future and any algorithm in this package is 100% covered by unit tests.

Yet Another Universal Backtesting Engine Release (YAUBER) – Backtester

https://github.com/alexveden/yauber-backtester

yauber-backtester is bare-bone portfolio backtesting engine that:

  • supports various portfolio management techniques: asset ranking, basket trading, portfolio rebalancing, etc.
  • intended to work on large asset universes (like 2000-3000 US Stocks EOD), or small intraday asset universes (like futures or forex, 1h timeframe).
  • supports meta-strategies, building and managing a portfolio of other trading strategies
  • allows simulating capital allocations, costs, margin trading, etc.

Yet Another Universal Backtesting Engine Release (YAUBER) – Executor

https://github.com/alexveden/yauber-executor

A boilerplate project for building asyncio Python distributed infrastructure.

Cython Tools: toolkit for efficient Cython development

https://github.com/alexveden/cython-tools

Key features

  • Keeping all development tools in one place with minimal efforts
  • Built-in Cython debugger (including low-level cdef / C code, based on GDB)
  • Cython code unit testing with coverage reports
  • Line-profiler
  • Easy running .pyx files by entry-point functions
  • Cython annotations and index file for all project
  • Cython project boilerplate
  • Cython snippets for low-level C-code and debugging tricks

Testimonials

Since the Upwork rejected to work with all Russians, and suspended access to my profile, here are screenshots of my projects’ feedbacks and the profile page as well.