FWE correction in 3dClustSim

I ran this command in ++ Compile date = Oct 26 2017 {AFNI_17.3.01:linux_xorg7}

3dClustSim -mask mask_temp+tlrc -NN 1 -fwhmxyz ${xx} ${yy} ${zz} -pthr 0.005 0.001 -athr 0.05 0.01 0.005 -iter 10000 -quiet -prefix all_sb_${var_name}_std_FWHM_alsim

Of course, this command was old. And it echo errors in this new version of AFNI. The errors were:

++ Number of voxels in mask = 186996
++ start ACF calculations out to radius = 34.09 mm

  • ACF done (0.00 CPU s thus far)
    ++ ACF 1D file [radius ACF mixed_model gaussian_NEWmodel] written to 3dFWHMx.1D
    ++ 1dplot: AFNI version=AFNI_17.3.01 (Oct 26 2017) [32-bit]
    ++ Authored by: RWC et al.
  • and 1dplot-ed to file 3dFWHMx.1D.png
    ++ 3dClustSim: AFNI version=AFNI_17.3.01 (Oct 26 2017) [32-bit]
    ++ Authored by: RW Cox and BD Ward
    ++ 186996 voxels in mask (32.81% of total)
    *+ WARNING: -NN option is no longer supported! All NN cases are computed now :slight_smile:
    *+ WARNING: Use of -fwhmxyz is not advised; consider using -acf instead
    FMRI data does not have Gaussian-shaped smoothness!
    ** FATAL ERROR: 3dClustSim – unknown option ‘8.09099’
    ** Program compile date = Oct 26 2017
    I wonder how to calculate the values for -acf option; and how to fix the error " – unknown option ‘8.09099’"

Hi all
Can anyone help me?

Would you please put an “echo” in front of the command
so that we can see how the parameters expand? Perhaps
one of your variables has multiple values.

  • rick

Hi rick
thanks for your help. I have performed the scripts. I was surprised that the cluster size in 1-sided thresholding was bigger that those in 2-sided or bi-sided thresholding. To my knowledge, one sided test is less stringent that two sided test. So why the cluster size in 1-sided is the biggest?