Data are separated into four categories.
1) Raw data
The raw data (usually referred to as "Study Data") are the raw MRI scans and physiology. These are files than cannot be replicated. The idea is to have a location which should be backed up rigorously (often and in several places), as distinct from the analysis data which can be re-created.
Choose a medium-sized drive (1TB / 100 subjects is a rough guide). Make sure you have a back-up protocol for this drive.
2) Analysis data
Analysis data are automatically generated by the CSPM software, and so do not need to be backed up as carefully (or even at all). The software analyses the raw data, and saves the generated files in the Analysis data location.
Choose the largest space you have (3TB / 100 subjects is a rough guide). It is best to have a short path to this location, ideally the root folder of a drive (I use V:\).
3) LinuxShare: Data within Linux
LinuxShare is a folder "seen" in the Linux virtual machine and Windows, and is where files are swapped between the two operating systems. Choose something with a short path. It can be within the Analysis data space (I use V:\LinuxShare).
If you don't put this within Analysis data, choose a medium-sized drive (500+GB/100 subjects).
4) Physiology data
If you have physiology data, choose another location for the processed files. These will be larger than the raw data, but size will vary greatly according to the physiology recorded (which can range from ~10MB/subject to 2+GB/subject). These data include some manual processing, so they will not be replicable, and therefore this folder should also be backed up carefully.