The probe postprocessor is a testing tool that allows you to override the data sources in the system so that instead of reading from a data unit, they send out the values set in the probe. This allows for a quick way to run the system without having to import data or deploying the component. Learning is turned off when the probe is run.
|Probe main GUI|
In the right pane you will see the output of the probed component. Bottom right is a plot that continuously samples the output as you change the values. You can change the values by either moving the sliders or by entering numbers in the text box. For the sliders to be available, the data source that you override must have scaling data.
You have the following choices:
- Source: The data source you wish to override. Synapse tries to automatically find and select the input data source.
- Probe At: Defines which component to probe. Synapse tries to automatically find and select the output block.
- Reset Overrides: If you switch source, the probe keeps the overrides of the previous source. You can reset that override by pressing this button.
Loading external samples
Above the sliders you will find a "Load.." button. Clicking on it opens up the external sample dialog:
|External sample dialog|
- From Data Source: Loads a sample into the probe from the active data source. You can choose the data set and sample number. You will see a preview of the sample in the bottom text box of the dialog.
- By Manual Entry: Allows you to manually enter a sample in comma, semicolon, space or tab separated format. No other delimiters and no strings are allowed.
Clicking on the "OK" button loads the sample into the probe.
- Be aware that unless all your features are completely independent that you can easily make unrealistic or impossible combinations with the probe.
- The probe can only be used for static systems, i.e that have no short term memories/time dependencies.
- Tutorial 4 - Tutorial 4 ("Cows") has an in-depth practical example of using the probe.