Collinearity Warning with 3dREMLfit

AFNI Message Board
1

Hello AFNI Team,
I’m running into a collinearity issue when running 3dREMLfit. I’ve pasted the warning message and the 3dDeconvolve + 3dREMLfit commands below.

To look into this, I’ve checked my timing files and there are no duplicated onsets across my conditions. I’ve set up the timing files as local times and some runs don’t have any onsets. I’ve added a * for those rows/runs without onsets. I also tried to plot the design matrix made with 3dDeconvolve using 1dplot but it is too big (warning from 3dDeconvolve: + WARNING: Can’t plot 3dDecon_rf_aCompCorr_con/rf_matrix.jpg – matrix size 3912x185 exceeds max=3333). Do you have any guidance on how to troubleshoot this problem?

Thank you – Catherine

Collinearity warning message from 3dREML

  • X matrix: 19.436% of elements are nonzero
    *+ WARNING: -----
    *+ WARNING: QR decomposition of [R]^(-1/2) [X] had 5 collinearity problems
    *+ WARNING: -----

3dDeconvolve command to make design matrix
3dDeconvolve -input
${subject}_ses-sen_task-sentences_rec-topup_run-1_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz
${subject}_ses-sen_task-sentences_rec-topup_run-2_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz
${subject}_ses-sen_task-sentences_rec-topup_run-3_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz
${subject}_ses-sen_task-sentences_rec-topup_run-4_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz
${subject}_ses-sen_task-sentences_rec-topup_run-5_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz
${subject}_ses-sen_task-sentences_rec-topup_run-6_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz
-polort A
-num_stimts 35
-mask /usr/local/mridata/Smith_Merit/Processed_Data/derivatives/FMRI_Sentences/Group_Level/tpl-MNI152NLin2009aAsym_res-1_desc-brain_mask_dil2mm_2mm_master.nii.gz -local_times
-x1D_stop
-stim_file 1 confounds/${subject}_rot_x.txt -stim_base 1 -stim_label 1 roll
-stim_file 2 confounds/${subject}_rot_y.txt -stim_base 2 -stim_label 2 pitch
-stim_file 3 confounds/${subject}_rot_z.txt -stim_base 3 -stim_label 3 yaw
-stim_file 4 confounds/${subject}_trans_x.txt -stim_base 4 -stim_label 4 dS
-stim_file 5 confounds/${subject}_trans_y.txt -stim_base 5 -stim_label 5 dL
-stim_file 6 confounds/${subject}_trans_z.txt -stim_base 6 -stim_label 6 dP
-stim_file 7 confounds/${subject}_rot_x_derivative1.txt -stim_base 7 -stim_label 7 rolldx
-stim_file 8 confounds/${subject}_rot_y_derivative1.txt -stim_base 8 -stim_label 8 pitchdx
-stim_file 9 confounds/${subject}_rot_z_derivative1.txt -stim_base 9 -stim_label 9 yawdx
-stim_file 10 confounds/${subject}_trans_x_derivative1.txt -stim_base 10 -stim_label 10 dSdx
-stim_file 11 confounds/${subject}_trans_y_derivative1.txt -stim_base 11 -stim_label 11 dLdx
-stim_file 12 confounds/${subject}_trans_z_derivative1.txt -stim_base 12 -stim_label 12 dPdx
-stim_file 13 confounds/${subject}_rot_x_power2.txt -stim_base 13 -stim_label 13 roll_sq
-stim_file 14 confounds/${subject}_rot_y_power2.txt -stim_base 14 -stim_label 14 pitch_sq
-stim_file 15 confounds/${subject}_rot_z_power2.txt -stim_base 15 -stim_label 15 yaw_sq
-stim_file 16 confounds/${subject}_trans_x_power2.txt -stim_base 16 -stim_label 16 dS_sq
-stim_file 17 confounds/${subject}_trans_y_power2.txt -stim_base 17 -stim_label 17 dL_sq
-stim_file 18 confounds/${subject}_trans_z_power2.txt -stim_base 18 -stim_label 18 dP_sq
-stim_file 19 confounds/${subject}_rot_x_derivative1_power2.txt -stim_base 19 -stim_label 19 rolldx_sq
-stim_file 20 confounds/${subject}_rot_y_derivative1_power2.txt -stim_base 20 -stim_label 20 pitchdx_sq
-stim_file 21 confounds/${subject}_rot_z_derivative1_power2.txt -stim_base 21 -stim_label 21 yawdx_sq
-stim_file 22 confounds/${subject}_trans_x_derivative1_power2.txt -stim_base 22 -stim_label 22 dSdx_sq
-stim_file 23 confounds/${subject}_trans_y_derivative1_power2.txt -stim_base 23 -stim_label 23 dLdx_sq
-stim_file 24 confounds/${subject}_trans_z_derivative1_power2.txt -stim_base 24 -stim_label 24 dPdx_sq
-stim_file 25 confounds/${subject}_a_comp_cor_00.txt -stim_base 25 -stim_label 25 a_comp_cor_00
-stim_file 26 confounds/${subject}_a_comp_cor_01.txt -stim_base 26 -stim_label 26 a_comp_cor_01
-stim_file 27 confounds/${subject}_a_comp_cor_02.txt -stim_base 27 -stim_label 27 a_comp_cor_02
-stim_file 28 confounds/${subject}_a_comp_cor_03.txt -stim_base 28 -stim_label 28 a_comp_cor_03
-stim_file 29 confounds/${subject}_a_comp_cor_04.txt -stim_base 29 -stim_label 29 a_comp_cor_04
-stim_times 30 /usr/local/mridata/Smith_Merit/Raw_Data/Sentences/Randomization_1/stimes.3567_01_MonthList_sub12.1D ‘TENT(0,16,21)’ -stim_label 30 month_targets
-stim_times 31 /usr/local/mridata/Smith_Merit/Raw_Data/Sentences/Randomization_1/stimes.3567_02_WeekList_sub12.1D ‘TENT(0,16,21)’ -stim_label 31 week_targets
-stim_times 32 /usr/local/mridata/Smith_Merit/Raw_Data/Sentences/Randomization_1/stimes.3567_03_DayList_sub12.1D ‘TENT(0,16,21)’ -stim_label 32 day_targets
-stim_times 33 /usr/local/mridata/Smith_Merit/Raw_Data/Sentences/Randomization_1/stimes.3567_04_HourList_sub12.1D ‘TENT(0,16,21)’ -stim_label 33 hour_targets
-stim_times 34 /usr/local/mridata/Smith_Merit/Processed_Data/nifti/${subject}/ses-sen/func/time_files/remember_foils.1D ‘TENT(0,16,21)’ -stim_label 34 remembered_foils
-stim_times 35 /usr/local/mridata/Smith_Merit/Processed_Data/nifti/${subject}/ses-sen/func/time_files/forgotten_foils.1D ‘TENT(0,16,21)’ -stim_label 35 forgotten_foils
-num_glt 3
-gltsym ‘SYM: +.5remembered_foils[8…12] +.5forgotten_foils[8…12]’
-glt_label 1 “peak_all_foils”
-gltsym ‘SYM: +remembered_foils[8…12]’
-glt_label 2 “peak_remembered_foils”
-gltsym ‘SYM: +forgotten_foils[8…12]’
-glt_label 3 “peak_forgotten_foils”
-jobs 16
-errts 3dDecon_aCompCorr_rf/errts_rf
-fout -tout -x1D 3dDecon_aCompCorr_rf/rf_matrix.1D -xjpeg 3dDecon_aCompCorr_rf/rf_matrix.jpg
-bucket 3dDecon_aCompCorr_rf/${subject}_decon_results_rf
-cbucket 3dDecon_aCompCorr_rf/${subject}_decon_results_rf_cbucket \

3dREMLfit command
3dREMLfit -matrix 3dDecon_rf_aCompCorr_con/rf_matrix.1D
-input “${subject}_ses-sen_task-sentences_rec-topup_run-1_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz ${subject}_ses-sen_task-sentences_rec-topup_run-2_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz ${subject}_ses-sen_task-sentences_rec-topup_run-3_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz ${subject}_ses-sen_task-sentences_rec-topup_run-4_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz ${subject}_ses-sen_task-sentences_rec-topup_run-5_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz ${subject}_ses-sen_task-sentences_rec-topup_run-6_space-MNI152NLin2009cAsym_desc-preproc_bold_blur4mm_scaled.nii.gz”
-mask /usr/local/mridata/Smith_Merit/Processed_Data/derivatives/FMRI_Sentences/Group_Level/tpl-MNI152NLin2009aAsym_res-1_desc-brain_mask_dil2mm_2mm_master.nii.gz
-fout -tout -Rbuck 3dDecon_rf_aCompCorr_con/${subject}_decon_results_rf_REML -Rvar 3dDecon_rf_aCompCorr_con/${subject}_decon_results_rf_REMLvar -Rerrts 3dDecon_rf_aCompCorr_con/errts_rf_REML -GOFORIT -verb

2

What is the output of:

1d_tool.py -show_mmms -infile 3dDecon_aCompCorr_rf/rf_matrix.1D
1d_tool.py -show_cormat_warnings -infile 3dDecon_aCompCorr_rf/rf_matrix.1D

  • rick
3

Thank you for your response!

1d_tool.py -show_mmms -infile rf_matrix.1D
file rf_matrix.1D (len 3912)
col 0: min = 0.0000, mean = 0.1667, max = 1.0000, stdev = 0.3727
col 1: min = -1.0000, mean = -0.0000, max = 1.0000, stdev = 0.2361
col 2: min = -0.5015, mean = -0.0000, max = 0.9985, stdev = 0.1832
col 3: min = -1.0000, mean = 0.0000, max = 1.0000, stdev = 0.1550
col 4: min = -0.4301, mean = -0.0000, max = 0.9985, stdev = 0.1369
col 5: min = 0.0000, mean = 0.1667, max = 1.0000, stdev = 0.3727
col 6: min = -1.0000, mean = -0.0000, max = 1.0000, stdev = 0.2361
col 7: min = -0.5015, mean = -0.0000, max = 0.9985, stdev = 0.1832
col 8: min = -1.0000, mean = 0.0000, max = 1.0000, stdev = 0.1550
col 9: min = -0.4301, mean = -0.0000, max = 0.9985, stdev = 0.1369
col 10: min = 0.0000, mean = 0.1667, max = 1.0000, stdev = 0.3727
col 11: min = -1.0000, mean = -0.0000, max = 1.0000, stdev = 0.2361
col 12: min = -0.5015, mean = -0.0000, max = 0.9985, stdev = 0.1832
col 13: min = -1.0000, mean = 0.0000, max = 1.0000, stdev = 0.1550
col 14: min = -0.4301, mean = -0.0000, max = 0.9985, stdev = 0.1369
col 15: min = 0.0000, mean = 0.1667, max = 1.0000, stdev = 0.3727
col 16: min = -1.0000, mean = -0.0000, max = 1.0000, stdev = 0.2361
col 17: min = -0.5015, mean = -0.0000, max = 0.9985, stdev = 0.1832
col 18: min = -1.0000, mean = 0.0000, max = 1.0000, stdev = 0.1550
col 19: min = -0.4301, mean = -0.0000, max = 0.9985, stdev = 0.1369
col 20: min = 0.0000, mean = 0.1667, max = 1.0000, stdev = 0.3727
col 21: min = -1.0000, mean = -0.0000, max = 1.0000, stdev = 0.2361
col 22: min = -0.5015, mean = -0.0000, max = 0.9985, stdev = 0.1832
col 23: min = -1.0000, mean = 0.0000, max = 1.0000, stdev = 0.1550
col 24: min = -0.4301, mean = -0.0000, max = 0.9985, stdev = 0.1369
col 25: min = 0.0000, mean = 0.1667, max = 1.0000, stdev = 0.3727
col 26: min = -1.0000, mean = -0.0000, max = 1.0000, stdev = 0.2361
col 27: min = -0.5015, mean = -0.0000, max = 0.9985, stdev = 0.1832
col 28: min = -1.0000, mean = 0.0000, max = 1.0000, stdev = 0.1550
col 29: min = -0.4301, mean = -0.0000, max = 0.9985, stdev = 0.1369
col 30: min = -0.0057, mean = 0.0000, max = 0.0085, stdev = 0.0012
col 31: min = -0.0049, mean = -0.0000, max = 0.0049, stdev = 0.0010
col 32: min = -0.0024, mean = -0.0000, max = 0.0023, stdev = 0.0011
col 33: min = -0.0743, mean = 0.0000, max = 0.0885, stdev = 0.0227
col 34: min = -0.1487, mean = -0.0000, max = 0.1587, stdev = 0.0558
col 35: min = -0.1891, mean = -0.0000, max = 0.2491, stdev = 0.0598
col 36: min = -0.0077, mean = -0.0000, max = 0.0098, stdev = 0.0010
col 37: min = -0.0044, mean = 0.0000, max = 0.0053, stdev = 0.0005
col 38: min = -0.0036, mean = -0.0000, max = 0.0036, stdev = 0.0004
col 39: min = -0.1375, mean = -0.0000, max = 0.0798, stdev = 0.0117
col 40: min = -0.1949, mean = -0.0000, max = 0.2110, stdev = 0.0279
col 41: min = -0.2460, mean = 0.0000, max = 0.2939, stdev = 0.0343
col 42: min = -0.0000, mean = 0.0000, max = 0.0001, stdev = 0.0000
col 43: min = -0.0000, mean = -0.0000, max = 0.0000, stdev = 0.0000
col 44: min = -0.0000, mean = -0.0000, max = 0.0000, stdev = 0.0000
col 45: min = -0.0009, mean = 0.0000, max = 0.0081, stdev = 0.0010
col 46: min = -0.0035, mean = 0.0000, max = 0.0247, stdev = 0.0037
col 47: min = -0.0052, mean = 0.0000, max = 0.0473, stdev = 0.0070
col 48: min = -0.0000, mean = 0.0000, max = 0.0001, stdev = 0.0000
col 49: min = -0.0000, mean = 0.0000, max = 0.0000, stdev = 0.0000
col 50: min = -0.0000, mean = 0.0000, max = 0.0000, stdev = 0.0000
col 51: min = -0.0001, mean = 0.0000, max = 0.0188, stdev = 0.0005
col 52: min = -0.0008, mean = 0.0000, max = 0.0438, stdev = 0.0025
col 53: min = -0.0012, mean = -0.0000, max = 0.0852, stdev = 0.0031
col 54: min = -0.0841, mean = -0.0000, max = 0.1422, stdev = 0.0392
col 55: min = -0.1368, mean = 0.0000, max = 0.2856, stdev = 0.0392
col 56: min = -0.2614, mean = 0.0000, max = 0.2713, stdev = 0.0392
col 57: min = -0.2458, mean = -0.0000, max = 0.5938, stdev = 0.0392
col 58: min = -0.2976, mean = 0.0000, max = 0.4403, stdev = 0.0392
col 59: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 60: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 61: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 62: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 63: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 64: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 65: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 66: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 67: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 68: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 69: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 70: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 71: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 72: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 73: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 74: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 75: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 76: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 77: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 78: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 79: min = 0.0000, mean = 0.0151, max = 1.0000, stdev = 0.1219
col 80: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 81: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 82: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 83: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 84: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 85: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 86: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 87: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 88: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 89: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 90: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 91: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 92: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 93: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 94: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 95: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 96: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 97: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 98: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 99: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 100: min = 0.0000, mean = 0.0151, max = 1.0000, stdev = 0.1218
col 101: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 102: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 103: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 104: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 105: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 106: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 107: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 108: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 109: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 110: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 111: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 112: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 113: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 114: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 115: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 116: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 117: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 118: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 119: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 120: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 121: min = 0.0000, mean = 0.0151, max = 1.0000, stdev = 0.1219
col 122: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 123: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 124: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 125: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 126: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 127: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 128: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 129: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 130: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 131: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 132: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 133: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 134: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 135: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 136: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 137: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 138: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 139: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 140: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 141: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 142: min = 0.0000, mean = 0.0153, max = 1.0000, stdev = 0.1229
col 143: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 144: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 145: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 146: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 147: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 148: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 149: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 150: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 151: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 152: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 153: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 154: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 155: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 156: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 157: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 158: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 159: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 160: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 161: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 162: min = 0.0000, mean = 0.0294, max = 1.0000, stdev = 0.1689
col 163: min = 0.0000, mean = 0.0289, max = 1.0000, stdev = 0.1674
col 164: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 165: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 166: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 167: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 168: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 169: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 170: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 171: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 172: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 173: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 174: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 175: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 176: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 177: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 178: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 179: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 180: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 181: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 182: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 183: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320
col 184: min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0318

1d_tool.py -show_cormat_warnings -infile rf_matrix.1D
– no warnings for correlation matrix (cut = 0.400) –

4

Hello AFNI Team,

I’m bumping this to see if there might be a solution to our error?

Warm Regards,
Catherine

5

Hi Catherine,

Thank you for the reminder.

I don’t recall what I was thinking back then (big surprise), but there are some notable points from that mmms output.

There are 4 matrix columns that show min = max = stdev = 0, and 2 that are very close to that. What do these 6 columns look like?
I suppose 43, 44, 49 and 50 are all zero. But 42 and 48 are surely close to that. Any all-zero columns will lead to warnings.

col 42: min = -0.0000, mean = 0.0000, max = 0.0001, stdev = 0.0000
col 43: min = -0.0000, mean = -0.0000, max = 0.0000, stdev = 0.0000
col 44: min = -0.0000, mean = -0.0000, max = 0.0000, stdev = 0.0000
col 48: min = -0.0000, mean = 0.0000, max = 0.0001, stdev = 0.0000
col 49: min = -0.0000, mean = 0.0000, max = 0.0000, stdev = 0.0000
col 50: min = -0.0000, mean = 0.0000, max = 0.0000, stdev = 0.0000

There are also many columns that have identical mmms results, though that does not imply they are the same. But moving the “col xxx:” text to the end of the line and sorting, we see these all have the same output (for example, there are other sets as well):

min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 164:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 165:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 166:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 167:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 168:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 169:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 170:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 171:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 172:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 173:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 174:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 175:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 176:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 177:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 178:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 179:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 180:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 181:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 182:
min = 0.0000, mean = 0.0010, max = 1.0000, stdev = 0.0320 col 183:

So start by looking at how these were created, and see if there is anything peculiar in those timing files. Please feel free to post timing files that you are concerned about.

  • rick
6

Hello Rick,

Thank you very much for your help! I went back to the timing files and did find that there are duplicate values between our timing files. I will fix that error and will see if the collinearity issue is fixed.

-Catherine

7

Hello Rick,

The timing files are now correct and we are still getting the collinearity error.

For stim_times 34 & 35 of our model, it’s common to have runs without an onset. Especially stim_times 35. Empty runs are denoted as a * in the timing files. Is it possible this is causing the issue?

-stim_times 34 /usr/local/mridata/Smith_Merit/Processed_Data/nifti/${subject}/ses-sen/func/time_files/remember_foils.1D ‘TENT(0,16,21)’ -stim_label 34 remembered_foils
-stim_times 35 /usr/local/mridata/Smith_Merit/Processed_Data/nifti/${subject}/ses-sen/func/time_files/forgotten_foils.1D ‘TENT(0,16,21)’ -stim_label 35 forgotten_foils \

Thank you,
Catherine

8

Hi Catherine,

Yes, the all-zero columns will still result in those collinearity problem warnings, since the program cannot distinguish the cases.
I see you did have -GOFORIT in the original command

If you want, it would be okay to send me the matrix file via email (click on my name) so that I could investigate it locally.

  • rick
9

Hi Catherine,

Thank you for the X-matrix file. The collinearity issues are not sooooo bad, but they seem to be coming from the motion parameters, not the regressors of interest. Try running these commands:

xmat_tool.py -load_xmat rf_matrix_tosend.1D -no_gui -show_conds
1d_tool.py -infile rf_matrix_tosend.1D -show_cormat_warnings_full

The first shows small condition numbers for the main and (non-motion) baseline regressors, but a very large condition number for motion (and motion+others).

Looking at pairwise correlations from the second command, the high correlations seem to be between each motion parameter and its square.

So these are not bad things, but as they are correlated enough, the programs are whining about them. You might want to run these commands regularly, and verify that the high condition numbers are only due to motion squares.

Does that seem reasonable?

  • rick