Hi All,
I see that there are two schools of thought. One is to go with 3D, 4D arrays. The other to Use Cluster and Keep your Arrays Separate.
With 3d Arrays, the problem is it will have a size such that it can acomodate the largest 2d Array. So other smaller 2d Arrays will be tagged with Zeroes. One Way to get around this is to initialize your 3d Array with NAN. The problem is you must again know before hand the size of this 3d array. I am attaching an example how 3d Array initialized with NaN can populate 2d Arrays.
I do not Like Cluster Structure as it gets Cumbersome and Hogs a Lot of Memory. Not to Mention the various bundling and Unbudling Operations. Also if you want to save this data for use in spread sheet you have to "twist and shout". May
be a wrong metaphor?
What I like to do is no matter what Save all my data in 2d array. So I can represent a 3d,4d array in 2d Array form just by using First columns to show the index of various 2d Arrays. (1 column as index for 3D, 2 Columns for 4D etc.) I am attaching another example showing this method. The Biggest advantage is this array can be saved in spread sheet file. Also I think arrays can be manipulated easily to extract whichever elements or Subarrays needed for use. The Problem remains however what to do with the tagged Zeroes? Unfortunately Labview does this and may need to remedy this. Of Course You can Logically add NAN instead of tagged zeroes. This becomes some what easy because you are handling a 2d Array rather than 3d+ array.
Having said all this. I like the 2d Array approach as I have used this in the past and also I donot need to know the size of final data array. I can save all this valuable data in a hurry.
Good Luck!!
Mache
Good Luck!
Mache