Morioh
Login
Wilford Pagac

Wilford Pagac

2 years ago

A Freqtrade Framework & Strategy with Python

MoniGoMani

MoniGoMani
a Freqtrade Framework & Strategy

⚠️ Warning: Pre-release / Experimental ⚠️
It’s not recommended to use this strategy live already, due to it still being under heavy development!
If you’d like to see this project progress faster then please help out where you can here

Motivation

Isn’t that what we all want? Our money to go many? Well that’s what this Freqtrade framework/strategy hopes to do for you by enabling you to HyperOpt a lot of signals to alter the weight from.

Big thank you to xmatthias and everyone who helped on MoniGoMani, Freqtrade Discord support was also really helpful, so thank you too!

⚠️ Disclaimer

  • This strategy is under development. It is not recommended running it live at this moment.
  • Always test this strategy before using it!
  • I am in no way responsible for your live results! This strategy is still experimental and under heavy development!
  • MoniGoMani should always be re-optimized after doing manual changes!
  • You need to optimize your own copy of MoniGoMani while thinking logically, don’t follow your computer blindly!

Contents

The Idea & Theory

MoniGoMani aims to be more than just a conventional strategy, it’s a framework to “easily” find a profitable strategy configuration in any market! Without the need to do any programming. However, you will need to know some Technical Analysis and be able to pull your own conclusions from your test-results, this is not just an easy copy/paste.

MGM (MoniGoMani) derives itself from other strategies by its use of something I called “weighted signals”. Each signal has its own weight allocated to it & a total buy/sell signal needed is defined too. MGM will loop through all signals, if they trigger it will add up the weight and eventually it will check if it’s bigger than what’s needed in total over a candle lookback window (to take previous signals into consideration). If the grand total of the sum of weighted signals is bigger then what is required it will buy/sell. The signals used here are implemented, so they can easily be changed by a developer to further improve upon them.

The beauty lies in using MGM in combination with HyperOpting (= A form of machine learning where you BackTest a timerange a lot of times to find the most ideal values), since all weighted signals have been made HyperOptable it can be used to find the most “ideal” weight divisions. Also will it teach us what works where & what doesn’t since MoniGoMani first detects Downwards/Sideways/Upwards trends and then does all the above individually for each kind of trend (Creating basically 3 individual strategies, 1 for each kind of trend).

Further it will do various HyperOptable checks upon the open trades to see if there are “bad” ones to unclog while running.

Features

  • Partially Automated Optimization Process
  • All HyperOpt Results can easily be applied and removed with the use of some Go-To Commands
  • Configurable Buy/Sell Signal Weight Influence Tables for Downwards/Sideways/Upwards trends, each table currently has 9 Buy & 9 Sell signals implemented (HyperOptable!):
  • Configurable Total Buy/Sell Signal Percentages for Downwards/Sideways/Upwards trends (HyperOptable!)
  • Configurable LookBack Windows for Total Buy/Sell Signal Percentages for Downwards/Sideways/Upwards trends (HyperOptable!)
  • Configurable Trading on Downwards/Sideways/Upwards trends for Buys/Sells (HyperOptable!)
  • Settings to Enable/Disable HyperOpting for individual buy_params & sell_params and setting them to a static value through HyperOpt Setting Overrides
  • Configurable Open Trade Unclogger, if enabled it attempts to unclog the bot when it’s stuck with losing trades & unable to trade more new trades (HyperOptable!) 🚀
  • TimeFrame-Zoom during BackTesting/HyperOpting to prevent profit exploitation! (Read: BackTesting-Traps)
  • Custom Long Continuously decreasing ROI Table generation with configurable roi_table_step_size
  • Precision Setting to alter the step-size used during HyperOpting
  • 2 Custom HyperLoss Functions:
    • WinRatioAndProfitRatioLoss: Attempts to optimise for the best profit and stability (Returns smaller number for better results)
    • UncloggedWinRatioAndProfitRatioLoss: Same as WinRatioAndProfitRatioLoss but has a configurable Percentage of loss to ignore while HyperOpting (Small losses are a by-product of the Unclogger)
  • Top Volume & All Tradable StaticPairList Downloading to easily fetch a good StaticPairList
  • Total Overall Signal Importance Calculator for Total Average Signal Importance Calculation upon the HyperOpt Results (With some really handy subcommands)
  • Pre-Configured Main/Sub Plot Configurations for visualisation of all indicators used in FreqUI
  • Turn On/Off All Individual Weighted Signal DataFrame entries for easy debugging in an IDE or better speed while dry/live running or HyperOpting

Support/Updates for Legacy versions stopped since Auto-HyperOptable Strategies are merged into the official Freqtrade Development Branch! Please switch to the new MoniGoManiHyperStrategy!

Getting Started

Take a good read at the MGM_DOCUMENTATION.md, the current place where you can find all MoniGoMani Documentation!

Go-To Commands

Hyper Opting MoniGoManiHyperStrategy.py:

freqtrade hyperopt -s MoniGoManiHyperStrategy -c ./user_data/mgm-config.json -c ./user_data/mgm-config-private.json --hyperopt-loss WinRatioAndProfitRatioLoss --spaces all -e 1000 --timerange 20210101-20210316

Apply HyperOpt Results from a <epoch of choice>:

freqtrade hyperopt-show -n <epoch of choice> -c ./user_data/mgm-config.json -c ./user_data/mgm-config-private.json --no-header --print-json | tail -n 1 | jq '.' > ./user_data/mgm-config-hyperopt.json

Reset HyperOpt Results:

rm ./user_data/mgm-config-hyperopt.json

Back Testing MoniGoManiHyperStrategy.py:

freqtrade backtesting -s MoniGoManiHyperStrategy -c ./user_data/mgm-config.json -c ./user_data/mgm-config-private.json --timerange 20210101-20210316

Total Average Signal Importance Calculation (with the Total-Overall-Signal-Importance-Calculator.py):

python ./user_data/mgm_tools/Total-Overall-Signal-Importance-Calculator.py -sc USDT -lf ./user_data/mgm-config-hyperopt.json -cf ./user_data/Total-Average-Signal-Importance-Report.log

Retrieve a current Binance-Top-Volume-StaticPairList.json file (using Binance-Retrieve-Top-Volume-StaticPairList.json):

freqtrade test-pairlist -c ./user_data/mgm_tools/Binance-Retrieve-Top-Volume-StaticPairList.json --quote USDT --print-json | tail -n 1 | jq '.|{exchange: { pair_whitelist: .}}' > ./user_data/mgm_pair_lists/Binance-USDT-Top-Volume-StaticPairList.json
# Don't forget to open the downloaded '...-StaticPairList.json' and copy the PairList Data into your own 'mgm-config.json' file to start using it!

Download Candle Data:

freqtrade download-data --timerange 20201201-20210316 -t 5m 1h -c ./user_data/mgm-config.json -c ./user_data/mgm-config-private.json

Contributing

  • Feel free to join our community CryptoStonksShallRise on Discord, there you can follow/participate in the Official MoniGoMani Channels:
    • #moni-go-mani-announcements
    • #moni-go-mani-development (Only available to MoniGoMani Developers!)
    • #moni-go-mani-updates
    • #moni-go-mani-testing
    • #moni-go-mani-help (Be sure to take your time to read the Documentation thoroughly before asking questions though!)
    • #moni-go-mani-cluster-releases (Only available to the MoniGoMani Cluster Maintainer & MoniGoMani Developers, for now!)
    • #moni-go-mani-setup-releases

More general chats for Technical Analysis, Freqtrade, Iconomi and Random discussion are also available there 🙂

Planned

Ordered by current schedule/priority

  • More global/broader trend detection using zoomed out indicator data
  • Sigmoid Buy/Sell Signals
  • Other & Better indicators! MoniGoMani has been designed so signals can easily be inserted / swapped out Please use the Total-Overall-Signal-Importance-Calculator.py to find out which signals do best and report your results to the Discord Server, so we can improve! 🚀
  • HyperOpt over 3 separate timeranges (one representing each individual kind of trend, downwards/sideways/upwards, a timeframe that represents a corresponding trend should be picked)
  • A method to pull a Static Averaged Volume PairList (Calculated by summing up the top volume pairlists for each candle over the period of the timerange to hyperopt upon & then dividing by the total amount of candles in the timerange, to create an averaged “volume” pairlist that can be used during BackTesting/HyperOpting which should lead to a more “realistic” pairlist to test upon when using a VolumePairList when dry/live-running)
  • Improve upon bot loop speed (Try to improve code to reach reduction in HyperOpting time needed)
  • Huge refactor that should improve the codebase reducing a lot of duplicate code & making implementing new weighted signals even easier
  • MultiProcessed DataFrame indicator checking if possible for speed improvements
  • HyperOpt over a timerange through a few simple Telegram commands, review the results and choose if and which new epoch should be applied.

Changelog

Updates and releases can be found in under Releases

View the (old) legacy Changelog here.

Freqtrade

Freqtrade is the well known open source crypto day-trading bot that makes this strategy possible! It’s completely free to use and alter and has many amazing features. Big thank you to xmatthias and everyone who helped on it!

ICONOMI

Can’t wait until MoniGoMani is fully on point? Or is this all too technical for you? Check out ICONOMI!

Instead of buying loose individual crypto manually like you usually do on exchanges, this platform has mostly been created to buy & hodl Investment Strategies. ICONOMI strategies are owned by Strategy Managers, these are often day-trades / technical analysts by profession so in general they have quite a good idea what they are doing. Each investment strategy contains up to ±20 different coins with a percentage allocated to each one. The managers will often re-balance these percentages towards coins they’ll think will be profitable.

There are fees tied to each strategy, and it’s up to the manager of each strategy to pick the percentages of fees for his/her strategy. Usually strategies that are re-balanced often (aka market being watched more actively) or larger strategies with a good reputation ask higher fees. However, fees are only charged if new profits have been made, so they are quite in the benefit of the user. More info on fees in general can be found here, and more info on Performance fees can be found here.

In general this is a good platform to invest into when you still need to start learning Technical Analysis, when you don’t have time to monitor the status of the market or when you don’t feel confident trading your own funds. Since here you have strategy owners “doing the day-trading for you” by re-balancing the strategies & the percentages of coins in them.

If you join please use my referral link! => (https://www.iconomi.com/register?ref=JdFzz) 🙏 (Then a percentage of your fees that you have to pay anyways to the strategy owners and ICONOMI will go to me instead, which is a neat win-win way for us both to support me for my work on MGM!)

Recommended ICONOMI Strategies

  • Crypto Knowledge Pool (CKP): A community influenced strategy
  • CKP’s Telegram Chat: If you want to vote if the coins will go up or down and hear about interesting news or ask questions. When I wrote this they we’re right about 65% of the time. The manager will take the results into consideration when altering the strategy.
  • Knepala: The personal strategy of the owner of CKP, most of the time it does even better than CKP itself.
  • Look on ICONOMI for more strategies you deem interesting :slight_smile:

Download Details:

Author: Rikj000
Download Link: Download The Source Code
Official Website: https://github.com/Rikj000/MoniGoMani
License: GPL-3.0

#python #freqtrade

A Freqtrade Framework & Strategy with Python
19.60 GEEK