Weight Slider Controls
When you select a node from the logic template workspace, a weight slider control appears in the right hand pane. Additionally, if the node you selected had other nodes connected to it, their respective weight sliders will also appear.
The weight slider control gives you fine-tuned control over the output values of the nodes, allowing you to apply inversions (NOT operation), fuzzy logic hedges (VERY and SOMEWHAT), an output multiplier and range constraints. The weight slider also has user customization controls, including name and a plot color.
The weight slider control features a text box that displays the current name for this node. You can change this value to update the name. The name does not affect the output calculation but is useful to help you label the node in a context that you understand.
Logic nodes can be enabled or disabled by right-clicking the weight slider control and bringing up the context menu. Click on the Enabled menu item to toggle the enabled/disabled status. Disabled nodes will be treated in BloodHound as if they do not exist in the chart. Any nodes that are connected to a disabled node will be ignore the output from the disabled node. This is useful for eliminating the influence of a logic structure or solver when trying to diagnose an unexpected output. Note: this enabled/disabled menu setting overrides the solver’s enabled/disabled setting if set.
Solver nodes have a plot color property that when set, will plot the raw solver value in the BloodHound chart pane. Setting the plot color is useful for visualizing the solver output behavior in isolation. Use this this to understand what the solver is doing and how it contributes to the final BloodHound output.
The plot color only applies to solver nodes, so begin by selecting a solver node (plot color is not supported by Logic nodes such as AND, OR, Ratio or Additive). By clicking the gray shaded area on the left side of the weight slider control, a pop-up menu will appear. Select the plot color menu and choose a desired color for this solver. Finally click on the chart and click F5 to update it. You will notice a line plot in the BloodHound pane representing the solver value historically for each bar. The line plot is displayed for both the long (above the 0 line) and short values (below the 0 line) with the same color.
The range slider allows you to put constraints of the node’s output range. By default the range is 0 to 1.0. By sliding the left hand or right hand slider handle you can change the minimum output value and maximum output value respectively. Changing the range to a more narrow range will linearly map the original output to the new output range.
Changing the range is useful for constructing point based systems using the Additive node. For more information on how this is accomplished, see the documentation section for the Additive Node.
For example, suppose you wanted to constrain the node’s output to 0.30 to 0.80. After setting it, an original value of 0.5 (half way from 0 to 1.0) will be now mapped to 0.55 which is half-way between 0.30 and 0.80.
The formula for mapping values to the constraints are as follows:
where is the original output value, is the minimum value in the range, and is maximum value in the range.
The output multiplier multiplies the original value by the factor specified by the output multiplier. This effectively allows you to boost or reduce how favorable the output of this node.
The output multiplier is especially useful when used in conjunction with the Ratio node. When attaching a node to a Ratio node, the output multiplier works in the context of a numerical weight, which is directly factored into the averaging calculation. For more information, see the Ratio node description.
Right-click the weight slider control to bring up the context menu, from there click on Hedge. The sub-menu will provide you with the option to apply a NOT, SOMEWHAT and VERY modification to the node’s output. For an explanation of what the hedges (and inversion) does, see the section on Fuzzy Logic Hedges in the logic template documentation.