So I wanted to get into algo-trading last year, more specifically in crypto, but I didn't want to simply spot buy and sell crypto assets, imo those markets lack liquidity, and liquidity is king
So I decided to build a bot for leveraged trading of Bitcoin. Long story short, I saw a gap in the market: no one had built a bot framework for leveraged trading in crypto.
So thats just what Ive been doing in my free time over the past year, and since it's almost done. I think now is the time for fellow enthusiast to take a look and rip it apart.
A bit of context, on what the tool does and is:
Background
GoatFish is meant to be an infrastructure that allows you to dynamically run custom code in a vacuum (a docker container). while taking care of all the other elements of algotrading for you. All the streaming information from the exchange, all the order setting, backtesting, paperTrading.... is carried out an an application layer and the user doesn't have to worry about it.
Exchanges
Currently it only supports Bitmex, but it is structured to be able to add any an all exchanges a user may want to add with very little effort(as long as its leveraged trading).
Strategies
All you need to deploy a strategy is:
- A bot name
- A unique set of API keys
- A strategy, which can come in the form of a single file or several folders
- A preferred exchange
- A preferred timeframe [defaults to all timeframes if not specified]
This information will then be used to keep track of all your positions all of your orders and automatically connect the strategy to the chosen exchange, and submit/cancel/ammend orders when necessary.
Additionally the bot functionality allows you to dynamically switch between actions. A bot can just as easily:
- livetrade
- papertrade
- backtest
- stop
Analytics
The application provides real time analytics and scoring your strategies against each other.
Progress
As it stand the application is a series of dockerised endpoints that communicate through kafka. They each all work together and individually to deploy and run strategies. How ever usage at the moment is
limited o how knowledgable you are of the endpoints and their use, so I will be starting the development of a CLI with accompanying documentation to make it more accessible.
Im also very much aware the code is not perfect, but it does fairly well, a test strategy has been running a trade/min for the last 22 days wit
... keep reading on reddit ➡
I love SQLite for small apps and things with not much concurrency, but I didn't want to use a heavy ORM and have to create models. For some small projects, you just want to store some dicts with fast retrieval.
For that, I made Goatfish. It's a small schemaless layer over SQLite, where you basically create a class of your objects and they get .save() .find() and other such methods.
The library also supports indexing on arbitrary keys, for fast lookups.
Probably it
https://preview.redd.it/v3pqltbzgde11.png?width=960&format=png&auto=webp&s=789bb495e89dce5ad9996e5d59d86b41bc0b460b
So I wanted to get into algo-trading last year, more specifically in crypto, but I didn't want to simply spot buy and sell crypto assets, imo those markets lack liquidity, and liquidity is king
So I decided to build a bot for leveraged trading of Bitcoin. Long story short, I saw a gap in the market: no one had built a bot framework for leveraged trading in crypto.
So thats just what Ive been doing in my free time over the past year, and since it's almost done. I think now is the time for fellow enthusiast to take a look and rip it apart.
A bit of context, on what the tool does and is:
Background
GoatFish is meant to be an infrastructure that allows you to dynamically run custom code in a vacuum (a docker container). while taking care of all the other elements of algotrading for you. All the streaming information from the exchange, all the order setting, backtesting, paperTrading.... is carried out an an application layer and the user doesn't have to worry about it.
Exchanges
Currently it only supports Bitmex, but it is structured to be able to add any an all exchanges a user may want to add with very little effort(as long as its leveraged trading).
Strategies
All you need to deploy a strategy is:
- A bot name
- A unique set of API keys
- A strategy, which can come in the form of a single file or several folders
- A preferred exchange
- A preferred timeframe [defaults to all timeframes if not specified]
This information will then be used to keep track of all your positions all of your orders and automatically connect the strategy to the chosen exchange, and submit/cancel/ammend orders when necessary.
Additionally the bot functionality allows you to dynamically switch between actions. A bot can just as easily:
- livetrade
- papertrade
- backtest
- stop
Analytics
The application provides real time analytics and scoring your strategies against each other.
Progress
As it stand the application is a series of dockerised endpoints that communicate through kafka. They each all work together and individually to deploy and run strategies. How ever usage at the moment is
limited o how knowledgable you are of the endpoints and their use, so I will be starting the development of a CLI with accompanying documentation to make it more accessible.
Im also very much aware the code is not perfect, but it does fairly well, a test strategy has been running a trade/min for the last 22 days wit
... keep reading on reddit ➡So I wanted to get into algo-trading last year, more specifically in crypto, but I didn't want to simply spot buy and sell crypto assets, imo those markets lack liquidity, and liquidity is king
So I decided to build a bot for leveraged trading of Bitcoin. Long story short, I saw a gap in the market: no one had built a bot framework for leveraged trading in crypto.
So thats just what Ive been doing in my free time over the past year, and since it's almost done. I think now is the time for fellow enthusiast to take a look and rip it apart.
A bit of context, on what the tool does and is:
Background
GoatFish is meant to be an infrastructure that allows you to dynamically run custom code in a vacuum (a docker container). while taking care of all the other elements of algotrading for you. All the streaming information from the exchange, all the order setting, backtesting, paperTrading.... is carried out an an application layer and the user doesn't have to worry about it.
Exchanges
Currently it only supports Bitmex, but it is structured to be able to add any an all exchanges a user may want to add with very little effort(as long as its leveraged trading).
Strategies
All you need to deploy a strategy is:
- A bot name
- A unique set of API keys
- A strategy, which can come in the form of a single file or several folders
- A preferred exchange
- A preferred timeframe [defaults to all timeframes if not specified]
This information will then be used to keep track of all your positions all of your orders and automatically connect the strategy to the chosen exchange, and submit/cancel/ammend orders when necessary.
Additionally the bot functionality allows you to dynamically switch between actions. A bot can just as easily:
- livetrade
- papertrade
- backtest
- stop
Analytics
The application provides real time analytics and scoring your strategies against each other.
Progress
As it stand the application is a series of dockerised endpoints that communicate through kafka. They each all work together and individually to deploy and run strategies. How ever usage at the moment is
limited o how knowledgable you are of the endpoints and their use, so I will be starting the development of a CLI with accompanying documentation to make it more accessible.
Im also very much aware the code is not perfect, but it does fairly well, a test strategy has been running a trade/min for the last 22 days wit
... keep reading on reddit ➡So I wanted to get into algo-trading last year, more specifically in crypto, but I didn't want to simply spot buy and sell crypto assets, imo those markets lack liquidity, and liquidity is king
So I decided to build a bot for leveraged trading of Bitcoin. Long story short, I saw a gap in the market: no one had built a bot framework for leveraged trading in crypto.
So thats just what Ive been doing in my free time over the past year, and since it's almost done. I think now is the time for fellow enthusiast to take a look and rip it apart.
A bit of context, on what the tool does and is:
Background
GoatFish is meant to be an infrastructure that allows you to dynamically run custom code in a vacuum (a docker container). while taking care of all the other elements of algotrading for you. All the streaming information from the exchange, all the order setting, backtesting, paperTrading.... is carried out an an application layer and the user doesn't have to worry about it.
Exchanges
Currently it only supports Bitmex, but it is structured to be able to add any an all exchanges a user may want to add with very little effort(as long as its leveraged trading).
Strategies
All you need to deploy a strategy is:
- A bot name
- A unique set of API keys
- A strategy, which can come in the form of a single file or several folders
- A preferred exchange
- A preferred timeframe [defaults to all timeframes if not specified]
This information will then be used to keep track of all your positions all of your orders and automatically connect the strategy to the chosen exchange, and submit/cancel/ammend orders when necessary.
Additionally the bot functionality allows you to dynamically switch between actions. A bot can just as easily:
- livetrade
- papertrade
- backtest
- stop
Analytics
The application provides real time analytics and scoring your strategies against each other.
Progress
As it stand the application is a series of dockerised endpoints that communicate through kafka. They each all work together and individually to deploy and run strategies. How ever usage at the moment is
limited o how knowledgable you are of the endpoints and their use, so I will be starting the development of a CLI with accompanying documentation to make it more accessible.
Im also very much aware the code is not perfect, but it does fairly well, a test strategy has been running a trade/min for the last 22 days with out any downtime.
That being sai
