I wanted to add that NaN is only viable for floating point numbers. If you are using an integer, you will have to designate a bad value. For instance, many LabVIEW functions use -1 in a signed 32 bit number repensentation to designate default values when the only valid numbers are positive.
You can also change the default value of the numeric indicators by creating a new control. The new control will look and act just like a numeric except it will be modified to have -1 or NaN as the default values. This way you can build an array of your numeric control which when grown will have the default values of your choosing.
You can learn more on how to do this by reading "LabVIEW Custom Controls, Indicators, and Type Definitions". You can find it o
nline but it is also a shipping document. Open LabVIEW 6.1 >> help >> search the LabVIEW bookshelf. This opens a pdf of all the shipping documents which are a great place just to learn more about how LabVIEW works. Look toward the end of the pdf for the link to the custom controls one.