NaN, or "not a number", is a common acronym used to represent values that are undefined. In seismic data and other volumes, NaN values are used to represent non-existent data or missing values. Some processes and formulas may generate NaN values when the calculation is invalid.
NaN removal is a trace by trace process and is independent for each trace. It does not consider neighboring traces for its calculation. Samples for interpolation, extrapolation and nearest values are selected from the same trace.
- In the Control Panel, open the Process tab.
- At the tab header, click the Add icon and select New Process.
- Scroll down and double-click on NaN Removal.
- Type a name for the process and click OK.
Configure the NaN Removal process
- In the Details Panel, select the input Volume.
- Replace With Constant: Samples are replaced with the value.
- Linear Interpolation: Samples are interpolated linearly between the nearest non-NaN values of the same trace. Neighbouring traces are not used. Traces are extrapolated up and down using the first and last valid samples of the same trace. Neighbouring traces are not used.
- Nearest Neighbour: Sample values are copied from nearest non-NaN neighbour of the same trace. Neighbouring traces are not used.
If trace is all NaNs:
- Drop trace: Discards the entire trace.
- Fill with constant value: Replace all samples the value.
- Error: Warns interactively, but fails with an error in a workflow.
NaN replacement value:
- Use this value for Replace With Constant and/or Fill with constant value.
Treat zeros as NaN:
- Also replace/update samples with a value of 0.