Normally, data only flows once through a wire. It is generated once at a source, and then flows through the wire to be processed at the next vi.
If you want to store data, you need a special construct.
Typically, you can will use shift registers for that. Your program will probably have a main execution loop. The shift register of this loop then holds the data. See the first attached example program . Let it run, with highlight execution on (the lamp in the toolbar of the block diagram) That will probably give you a good feeling for how Labview works.
The second example program is similar to the first. But it uses an event loop, to handle the different parts of the program.
There's other ways of storing data. Often, you simply let the data flow to an indicator, and 'store' your information in the indicator. When you need the data later, your read the indicator using a property node. (right-click, create property node). You can also use the property node to update the data in the indicator.
In the second example program, I showed both methods. One where the data is updated using the shift register. The second, where it is update using the property nodes. (and also feed into the shift register, so that I don't break the rest of the program)
