[Camino-users] ActiveAx and AxCaliber fit

[stefania oliviero]

Hello all,
I simulated the diffusion signal coming from a synthetic substrate by using two acquisition schemes: one with 300 acquisitions, the other one with 3900 acquisitions. I am trying to execute the fit of such a signal with the models ActiveAx and AxCaliber by using the following commands:

for ActiveAx:

cat ${FNAMEINFO[i]}_NOISE.Bfloat | modelfit -fitmodel mmwmdfixed -fitalgorithm mcmc -burnin 1000 -samples 100 \
-interval 500 -noisemodel rician -sigma ${SIGMA} -schemefile ${SCHEMEFILE} | chunkstats -chunksize 16 -samples 100 \
-mean > ${FNAMEINFO[i]}_MMWMD.Bfloat

for AxCaliber:

cat ${FNAMEINFO[i]}_NOISE.Bfloat | modelfit -fitmodel TensorGDRCylinders -fitalgorithm MULTIRUNLM -samples 1000 -burnin 1000 \
-interval 500 -noisemodel rician -sigma ${SIGMA} -schemefile ${SCHEMEFILE} | chunkstats -chunksize 16 -samples 1000 \
-mean > ${FNAMEINFO[i]}_AxCaliber.Bfloat

With both the acquisition schemes

In the case of ActiveAx the output is only:

mar 17, 2021 1:04:21 PM optimizers.LM_Minimizer minimise
AVVERTENZA: optimizers.MarquardtMinimiserNonConvergenceException: Exceeded maximum number of iterations: 5000

In the case of AxCaliber:
mar 15, 2021 11:13:06 AM optimizers.LM_Minimizer minimise
AVVERTENZA: optimizers.MarquardtMinimiserException: Infinite objective function in mrqcof.

How can I fix such issues?

Thanks in advance for your help
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.nitrc.org/pipermail/camino-users/attachments/20210317/fbf80156/attachment.htm>