Inference

Inference Data objects are used during the process of inferring missing data values in the network. The Inference Data object contains a set of rules that controls the inference process.

Missing network values can be inferred from existing data, allowing you to fill in gaps in your data with reasonable values. For example, if the width of a pipe is missing but you have the width of the pipe immediately upstream, it is reasonable to assume that the missing width is the same. InfoWorks WS Pro includes a set of inference rules that you can apply to all or part of the network. The use of a common set of rules within the software means that all missing data can be inferred to the same standard, rather than depending on who does the inference or when it is done.

Inference rules are viewed and edited in the Inference Data Editor.

InfoWorks WS Pro can infer the following missing values:

• Link Parameters (Length, Diameter, Roughness Coefficient, Material, Year)
• Link Water Quality Parameters (Bulk Coefficient and Wall Coefficient)
• Node Parameters (Coordinates, Elevation, Ground Level, Nominal Pressure and Area Code
• Area Codes for links and nodes
• Customer Point Parameters (Elevation, Specific Consumption and Average Demand)
• Pump Parameters (Nominal Flow, Nominal Speed and Maximum Speed)
• Pump Station Duty Head
• Reservoir Parameters (Top and Bottom Water Level)

Merging of nodes and links via Proximity tests can also be carried out using inference rules.

You can infer values for either the currently selected objects or the whole network. You can also choose which missing values are to be inferred and select the user-defined data flag that is to be applied to all inferred values. The use of a data flag is important for auditing purposes, as it allows you to distinguish between inferred data and data created in other ways.

InfoWorks WS Pro has built-in inference rules for link parameters, node parameters, customer point elevations, pump and pumping station parameters, reservoir parameters and proximity testing.

Inferring missing values

Before you can infer missing data, you must create an Inference Data object in the database. This is stored in an Inference Data group. See Creating Groups and Creating an Item Within a Group for more information on creating the empty object. You need to set up the inference parameters you require in this object before initiating the inference process.

To set up an Inference Data object:

1. Create the Inference Data object in an Inference Data group.
2. Drag the object onto the main window background (or right-click and choose Open). This displays the Inference Data Editor.
3. Set up the rules for each individual inference task. The tasks are listed down the left-hand side of the dialog.
• You can change the order of the rules by using the Up and Down buttons. Click a rule to view parameters  on the right hand side of the dialog. The top pane contains a description of the inference task, so the editor should be easy to use.
• You can switch a particular task on or off by toggling the tick box next to the name.
• You can set a flag for all inferred values.
4. When you have completed editing the inference data object, simply close the window. You will be prompted to save any unsaved changes. Use the Save button at any time to ensure you do not accidentally lose changes you have already made.

To infer missing values:

1. If you want inference applied to only part of the network, select the required objects. See Selecting Network Objects.
2. To start the inference process, drag your Inference Data object onto the GeoPlan Window for your checked out network. InfoWorks WS Pro carries out the data inference process for all selected objects. If there is no current selection, an option will be displayed to carry out inference for the entire network.