Settings¶
Here’s a list of settings specific to tradero, for Django settings refer here.
All can be set with environmental variables with the TRADERO_ prefix (i.e. TRADERO_QUOTE_ASSET for QUOTE_ASSET).
QUOTE_ASSET¶
Default: BUSD
The Quote Asset for the Symbols to be retrieved.
TIME_INTERVAL¶
Default: 5
The Time Resolution in minutes to be used for the data, models, and indicators.
SYNC_EXECUTION¶
Default: False
Synchronous Execution of Threads (No Threads). Meant only to be used when running tests.
EXECUTOR_THREADS¶
Default: None
Amount of Threads to be used when parallelizing code. A way of limiting CPU core usage. It corresponds to the max_workers parameter of the
ThreadPoolExecutor.
USE_TASKS¶
Default: False
Use tasks instead of threads for scheduling. A task for updating each Symbol or Bot will be submitted to the workers.
SYMBOLS_QUANTITY¶
Default: cpu_count() * 4
Amount of Symbols to be tracked once they have been ranked by model score and filtered by market size.
WARM_UP_PERIODS¶
Default: 1
Number of periods of data to be retrieved by the warm_and_ready command. Each period corrresponds to 1000 time intervals (i.e. 2 periods with a time interval of 5 correspond to the data of the last 10000 minutes - 2000 k-lines of 5 minutes).
MARKET_SIZE_THRESHOLD¶
Default: 180000
Minimum Last 24h Volume of Quote Asset of a Symbol. Symbols below this threshold will not be tracked and listed to the User.
MODEL_SCORE_THRESHOLD¶
Default: 0.3
Minimum model score of a Symbol. Symbols below this threshold will not be tracked and listed to the User.
CLEANING_WINDOW¶
Default: 1000
Amount of Klines and Training Data Time Intervals (observations) to be left for each Symbol after updating indicators. Controls how much data is fed to the Prediction Model. A value of 1500 will feed the Prediction Model with at most the last 1500 Training Data and “clean” the older ones.
CLEANING_WINDOW_BOTS_LOGS¶
Default: 3600
Amount of TraderoBot Logsto be left for each bot after updating (0 to disable logrotate).
PREDICTION_MODEL_CLASS¶
Default: base.DecisionTreeRegressor
Prediction Model Class in dotted path format to be used.
PREDICTION_ENABLED¶
Default: True
Enables core prediction of the next time interval functionality.
OUTLIERS_MODEL_CLASS¶
Default: base.OutliersSVC
Outliers Model Class in dotted path format to be used.
OUTLIERS_ENABLED¶
Default: True
Enables core outliers detection functionality.
OUTLIERS_THRESHOLD¶
Default: 0.05
Proportion of Symbol observations (prices) to be considered as atypical by the outliers detection functionality.
INDICATORS¶
Default: __all__
Indicators to be enabled (calculated and shown). A string of comma-separated indicators’ slugs - i.e. macdcg,stp - or __all__. For indicator-specific settings, see the indicator’s documentation.
BOT_USER_QUOTA¶
Default: 0
Default Bot Quota (Maximum number of Bots) for Users (0 for no quota).
DUMMY_USER_ENABLED¶
Default: False
Enables the Dummy User.
DUMMY_USER_SYMBOL¶
Default: KEY
Base Asset of the Symbol to be used at start for Bots.
DUMMY_USER_BOTS¶
Default: 50
Amount of Bots to create for the Dummy User.
DUMMY_USER_BOT_QUOTA¶
Default: 100
Bot Quota for the Dummy User.