Creating and deleting virtual variables

Creating virtual variables

Use the Dataflow Toolbox in the Dataflow window to create new virtual variables, virtual lines or virtual surfaces and other dataflow objects.

During creation, the displayed operators are filtered by the type of objects in the Dataflow window and are further filtered by the presence of a pre-selected object(s). Pre-selected objects are connected as operands. When there are no pre-selected objects, a virtual variable is created without connected operands. It is worthwhile to review the suitability of the operands for a new virtual variable.

Notes:

New virtual variables by default inherit their Calibration Source. See Calibration settings for virtual variables for further information.
New virtual variables copy some general properties from their operand(s). See Inheritance of Properties for more information.
If you often create the same virtual variables in your EV files, you can create a template EV file with the virtual variables and their properties already defined. See About templates for more information.
It is also possible to Import virtual variables from one EV file into another.
You can Copy-and-Paste a virtual variable or a section of a virtual variable chain to another EV file or the same EV file.
Alternatively, under a suitable active echogram, the Line and Surface tool can create virtual lines/time series or virtual surfaces. Under the Echogram menu, selecting Export, then New Analysis Exporter, offers supported analysis exporters.

Blank echograms and graphs

A blank echogram or graph may indicate that Echoview is unable to calculate output sample values for a virtual variable. A Warning icon beside a dataflow object indicates that a warning or error occurred during its calculation. Point to the icon to display the associated message, which may help you troubleshoot the issue.

A blank echogram or graph can occur without a Warning icon. If no icon is displayed, or the associated message does not identify the cause, check the input data, operator settings, and upstream objects.

Issues may include:

Unsuitable operand(s) for the operator. Review the data types of the supported operands of your operator.
A displayed calculation may be selectable but when evaluated the output is undefined (e.g., causes a divide by zero) or is unsuited for the operand data type. Review operator caveats and equations. An example is where Standard deviation is selected for boolean data, such as when using the XxY Statistic or XxYxZ Statistic operators, which do not support this calculation for boolean inputs.
A mismatch in ping geometry, at the platform or transducer level. Review the Platforms and Transducers of the operands. The Match Geometry operator is useful to change the ping geometry of an operand.
A mismatch in operand ping times. The Match ping times operator can output a virtual variable composed of pings with matching times.
The operator algorithm may output no data values under particular conditions. Review the operator algorithm.
Code operator error handling.
Scripting errors.

Deleting virtual variables

Complete the following steps to delete a virtual variable.

Click Dataflow to display the Dataflow window.
Click the virtual variable you wish to delete.
On the Dataflow menu, choose Delete

Note: Delete is not available if the selected variable is a raw variable.

Complete the following steps to delete all virtual variables.

Click Dataflow to display the Dataflow window.
On the Dataflow menu, click Delete All Virtual Variables.
Note: This will also delete any unavailable virtual variables imported with a Template.

Adding new data to an existing virtual variable

Some virtual variables need to recompute the entire dataset when new data is added, i.e., they do not simply append new values.

This leads to a lag in performance while Echoview updates the results, especially when triggering operators further down the dataflow chain to recompute.

The following operators recompute their entire dataset when new data becomes available. This has performance consequences for live viewing.

During live viewing

These operators, which use a centered window, recompute the entire dataset when new data is added. However, to improve performance during Live viewing, they only reprocess existing data in the first half of the window. Downstream operators will still recompute the entire dataset.