You can use make_dyadic_vectors to calculate the dispersion (and also the cone of uncertainty, see the usage). The only requirement is that your input should be 4D volumes of samples of theta/phi as described on the bedpostx doc.
> On 3 Aug 2017, at 14:54, Max Wichmann <[log in to unmask]> wrote:
> Hello Fsl Team,
> in my work I analyse different DWI acquisition schemes. Therefore I do different measurements. I repeat every measurement 5 times to get some statistic. The mean directions for every measurement I calculate via the make_dyadic_vectors command. But is there also a command for calculating the cone of uncertainty for my own sample. I know that bedpostx calculates it directly. But in my case I take the bedpostx output files of 5 samples and want to calculate again the uncertainty (dispersion). I could not find a single command in the directory. Maybe somebody knows how I can calculate it or where I can find it in the bedpost script?
> Thank you for your help!