Hello-

I am trying to run the 3dhistog command on a set of Beta-valued data to see if I can determine what a good cut-off point would be for outliers (i.e. Betas can often get inflated due to neighboring motion or TR pulse artifact). I have a script that I ran to exclude Beta outliers more extreme than +/- 500, however I wanted to see if there were a better more stringent cut-off for this, and I wanted a visual representation of the data to see if there were a good point in which to do this. Since I will be using these Betas to calculate standard deviations for each participant, getting rid of outlying Bet-values is important, while at the same time not excluding data that should remain in there to capture true variability (that isn't related to things like motion artifact or TR pulse inflations, etc).

I tried specifying number of bins (the default is 100, but I even tried to make it 100, to see if this would change anything), however no matter what I do when I plot using the 1dRplot or the 1dplot I get a very similar looking graphical representation, which is not really giving me what I want to look at. I know the data is in the file I want because it will bin and show frequencies within the text output.

I have attached a screenshot that shows the frequencies in text form as well as the command line 3dhistog command I used to generate those values, along with the plot function I tried to execute. I can't seem to get the bins in the histogram to do what I want. The data being input is float data, but from the documentation it seems like this command should be able to handle it. Maybe there is a flag I need to put in that I am missing?

Anywhoo, any help would be appreciated!!
-AK