Integer Variables?

Share icon

Hello!
Maybe i am just a bit stupid, but i have a small problem that _should_ be trivial, but i cannot seem to solve it:

I have to import a binary file from a custom experiment. This uses packed bits to store 2 values in an unsigned long(32bit) int.
Seperating them should be easy by using binary operators. But loading it into a "variable", its automatically cast into a float (messing things up).
Is there any way to define variables strictly as interger in a procedure? Or am i just doing something wrong and am using the wrong method?
Igor variables are strictly doubles. However, a wave's type can be integer (long or short, unsigned or signed). You'll probably want to just create a wave with as many points as you have integer variables. If necessary you can use dimension labels to refer to different points of the wave by a name. Try looking in the command help for the Make operation and the SetDimLabel operation and go from there.

April 22, 2008 at 05:28 pm - Permalink

You don't say how you are loading the data but I suspect you are using FBinRead. You can tell FBinRead that you are reading unsigned long using the /U/F=3 flags. This will read an unsigned long into the variable you pass as the last parameter to FBinRead and will convert the unsigned long to double. You can then break the two values apart with something like:

Variable lowPart = value & 0xFFFF
Variable highPart = value / 2^16

April 22, 2008 at 09:18 pm - Permalink