LabVIEW
measurement slowed by counter
From what I can tell, you are reading single samples of N channels in a tight loop, seperating the channels, and stuffing each individual into its own queue. In a parallel loop, you are building these samples into an array and if the size is 100, you average them, stream them to a file, and zero the intermediary arrays. All arrays grow at the same rate, so why all these separate size checks and case structures?
This is all way too complicated and inefficient. Why not read 100 samples at once, average the channels, place the array of averaged channels into a single queue, and write them all at once to tdms?
You also seem to be unaware of common simplifications. Index array is resizable. Most primitives can operate on arrays. For example on the left, you explicitly index out four individual string elements, see if they are empty strings, then invert the boolean and form an array of four elements. You would get exactly the same resulting boolean array by taking the four element array subset, checking for empty, and inverting.
I think with little effort, you code could be reduced to 20% of its current size and it would run much more efficiently and would be easier to maintain and debug. Try it! 🙂
