Signal Maker
Signals define the conditions for entering and exiting positions in a backtest. They also adjust backtest configurations for conditional pyramiding, leverage, stop-loss, and take-profit. In other words, a signal is a set of rules—either user-defined or generated by a machine learning model—used in backtesting.
Signal Maker allows you to combine various market data features using AND and OR operators to create custom signals. If you’re looking to generate signals using machine learning models, refer to Signal Generator.
To create a signal, you need to complete the fields listed in the table below.
Metric | Description |
|---|---|
Signal Name | The name assigned to the signal. |
Market Data | The market data containing the conditions involved in the signal. |
Timeframe | The time period for which the signal is defined. |
Signal | The conditions that define and trigger the signal. |
How to Define a Signal?
Signals are generated from columns containing labels. To create such columns, use processing functions like Advanced Labeling, Cross, etc.
Once you have the necessary column(s) for the signal, follow these steps:
Click ADD.
Select the column in the Categorical Feature box.
Open the Label box and choose the value that should trigger the signal.
If your signal involves multiple columns, use AND and OR operators to combine them.
To learn more about creating signals, read this tutorial.