Hi Matthew, I have modified the file but now I get this error five times: Welcome to FABBER v2.0 Logfile started: /tmp/fsl_V0k80d_ox_asl/basil/step1/logfile Start time: Tue Oct 13 14:28:52 2015 FABBER release v2.0 Forward Model version: $Id: fwdmodel_asl_grase.cc,v 1.22 2013/09/04 15:13:00 chappell Exp $ Loading mask data from '/tmp/fsl_V0k80d_ox_asl/mask' Loading data from '/home/daan/Desktop/PIN43218_t0//nifti/diffdata.nii.gz' Invalid_option exception caught in fabber: An exception has been thrown Runtime error:- Data length (10) does not match model's output length (9)!Trace: VariationalBayesInferenceTechnique::DoCalculations; FABBER main(); FABBER main (outer). Usage: fabber <arguments> Arguments are mandatory unless they appear in [brackets]. Use -@ argfile to read additional arguments from a text file. [--help] : print this usage message --output=/path/to/output : put output here (including logfile) --method={vb|spatialvb} : use VB (or VB with spatial priors) [--max-iterations=NN] : number of iterations of VB to use (default: 10) [--data-order={interleave|concatenate|singlefile}] : should time points from multiple data be interleaved (e.g. TE1/TE2) or left in order? (default: interleave) --data1=file1, [--data2=file2]. (use --data=file instead if --data-order=singlefile) --mask=maskfile : inference will only be performed where mask value > 0 --model={quipss2|q2tips-dualecho|pcasl-dualecho} : forward model to use. For model parameters use fabber --help --model=<model_of_interest> --noise={ar1|white} : Noise model to use ar1: two AR(1) models (optional cross-linking between TE1 & TE2) [--ar1-cross-terms={dual|same|none}] : two types of cross-linking, or none (default: dual) white: white noise model, optionally with different noise variances at some data points [--noise-pattern=<phi_index_pattern>] : repeating pattern of noise variances for each data point (e.g. --noise-pattern=12 gives odd and even data points different noise variances) [--save-model-fit] and [--save-residuals] : Save model fit/residuals files [--print-free-energy] : Calculate & dump F to the logfile after each update [--allow-bad-voxels] : Skip to next voxel if a numerical exception occurs (don't stop) For spatial priors (using --method=spatialvb): --param-spatial-priors=<choice_of_prior_forms>: Specify a type of prior to use for each forward model parameter. One letter per parameter. S=spatial, N=nonspatial, D=Gaussian-process-based combined prior --fwd-initial-prior=<prior_vest_file>: specify the nonspatial prior distributions on the forward model parameters. The vest file is the covariance matrix supplemented by the prior means; see the documentation for details. Very important if 'D' prior is used. Any suggestions? With kind regards, Daan On Tue, Oct 13, 2015 at 2:18 PM, Matthew Webster < [log in to unmask]> wrote: > Hello - this might be a regression from 508, can you modify the first line > of /usr/lib/fsl/5.0/oxford_asl to read: > > #!/bin/bash > > and let me know if it fixes your issue. > > Kind Regards > Matthew > > Hello all, > > > > Since having updated fsl in Ubuntu yesterday I have been getting error > messages when I try to use oxford_asl. I have not changed any settings or > values. The error message reads: > > > > /usr/lib/fsl/5.0/oxford_asl: 1176: /usr/lib/fsl/5.0/oxford_asl: > alltis[0]=0.1: not found > > /usr/lib/fsl/5.0/oxford_asl: 1176: /usr/lib/fsl/5.0/oxford_asl: > alltis[1]=1.5: not found > > Number of inversion times: 2 > > Number of timepoints in data: 10 > > Number of repeats in data: 5 > > /usr/lib/fsl/5.0/oxford_asl: 1190: /usr/lib/fsl/5.0/oxford_asl: Syntax > error: Bad for loop variable > > > > If I use bash before calling oxford_asl it seems to work, but I am > wondering if something has changed in the oxford_asl pipeline? I could not > find anything on the FSL site. And am I supposed to use bash before the > command? Thank you. > > > > With kind regards, > > > > Daan van den Oever >