Hi - I've only just discovered that one can take FSL vector data, and not only transform it into a for renderable by caret, but scale the vectors by any arbitrary volume (e.g. by some FSL model property, like dispersion, or indeed even by tracking results).
It's all great, I can't believe I missed that. It works very well, although I suspect it wasn't so much a core feature.. It would be nice to map the vector magnitudes back into a scalar volume file. and you provide a way to do this. I get the result: $ caret_command -volume-copy-vector-mag dyads1_d1_inverse.LR.vector.gii f1.nii.gz dyads1_d1_inverse.LR.vector.gii.nii.gz VOLUME COPY VECTOR FILE MAGNITUDE INTO VOXELS ERROR: Error dyads1_d1_inverse.LR.vector.gii: is an XML file but XML is not supported for this type of file. Perhaps you need a new version of Caret. except that XML is the only kind of format for giftii, at least I think so and it won't let me convert to anything else. I get the same with caret_command -volume-convert-vector-to-volume possibly those commands were devloped for something else. possibly I shouldn't be using a mac. I'm using lion. I could try RHEL 6 and ubuntu 10.4 or soething. anyway, it would be great to map those magnitudes back (that's really quite neat. you can do a lot with that actually. mapping the magnitudes back is a bonus. so, if it can work great. if not fine. -- someone mentioned now might be a good time for feature requests. In general the thing I find most resitrictive about caret is that even though you can show a different metric on each surface, there is only one threshold and only one pallete (you can get round it somewhat by calling your metric a shape file, but it's limited). It would be great to find a way to show different palettes and thresholds for different viewing windows. I appreciate that's a bit non-trivial. possibly the data manipulated by "metric settings" could be stored as part of the metric itself, at least as an option? also: it would be nice to be able to show more that one functional overlay, and have more than one anatomical underlay. at the moment, there is one anatomical underlay, and it's always black and white (assumed to be a T1w image) and one functional overlay. the functional overlay is tied into the metric settings, and that can be very disruptive. even if the functional volume is actually the data from which you derived the metric being displayed, it's often very hard to get the both thresholded right. it would be nice at least if the volume and surface palettes and thresholding were separate. All of this can be more than a bagatelle. At the moment I want to show very many metrics in groups of scenes. very many. and then crunch through that for all the data sets there's a script. it processes the data. it makes a spec file. the spec is equivalent in all data sets. so the same scene file applies to all. adjust. print. if caret were more flexible about palete and thresholding on a per window (or per metric would work. that's enough room to move easily) basis that would be pretty much all I can think of to improve. I think so actually. everything else is doable, that I can think of, one way or another. Clearly there will be a switchover here. So far, for the few (the one) things I've tried, I've found wb_command and caret to be interoperable. it's all great. those are feature requests, but they aren't today's. the palette/metric/threshold/multiple window/volume stuff is a general bugbear. really the only one actually. you can't get round it. I mean you can, but not with caret. you have to print the windows individually. and that isn't great if you're working with the gui, comparing this and that, seeing whats what in the data. thanks! Colin
_______________________________________________ caret-users mailing list caret-users@brainvis.wustl.edu http://brainvis.wustl.edu/mailman/listinfo/caret-users
