History of AFNI updates  

|
February 10, 2003 12:25PM
Doug or others,

Help me out with what should be a quick question. I know I'm missing something -- 18 degrees of freedom -- yet I'm not sure where. Granted, it's a minor iss.ue, but I'm bugged by it.

The specifics:
360 timepoints split into 3 runs (-concat flag)
3 stimulus vectors, each at 11 timepoints
6 motion vectors (put into baseline for "overall F" with -stim_base)
3 offset / drift vectors.

In the .HEAD file of the output from 3dDeconvolve, I've got 33 and 282 degrees of freedom in this F (double-checked that these really are the # of df AFNI is using to calculate p). The 33 is clear, as my "overall F" is really "F for the vectors of interest" (everything else is in the "baseline"): 3 stim vects at 11 timepoints = 33.

The 282 has me more perplexed. OK, so these are the number of timepoints (360) minus the number of degrees of freedom I've used. 360-282 = 78. 33 for stimuli of interest, 9 for offset / drift (3 runs with 3 each), 18? for motion (3 runs with 6 each) and we're at 60. 18 left and I can't think of where they've gone. What am I missing Doug? I looked at another example and I'm off by the same amount (also 3 runs, but 9 stimulus vectors, and 540 timepoints).

3dDeconvolve call is below. Thanks in advance,

Craig


3dDeconvolve -input run_study+orig -polort 2 -num_stimts 9 \
-stim_file 1 "study_behav.txt[1]" -stim_label 1 "SSI" \
-stim_file 2 "study_behav.txt[2]" -stim_label 2 "WSI" \
-stim_file 3 "study_behav.txt[3]" -stim_label 3 "FORGOT" \
-stim_file 4 "motion_study[0]" -stim_label 4 "Roll" \
-stim_file 5 "motion_study[1]" -stim_label 5 "Pitch" \
-stim_file 6 "motion_study[2]" -stim_label 6 "Yaw" \
-stim_file 7 "motion_study[3]" -stim_label 7 "dS" \
-stim_file 8 "motion_study[4]" -stim_label 8 "dL" \
-stim_file 9 "motion_study[5]" -stim_label 9 "dP" \
-stim_minlag 1 0 -stim_maxlag 1 10 \
-stim_minlag 2 0 -stim_maxlag 2 10 \
-stim_minlag 3 0 -stim_maxlag 3 10 \
-stim_minlag 4 0 -stim_maxlag 4 0 \
-stim_minlag 5 0 -stim_maxlag 5 0 \
-stim_minlag 6 0 -stim_maxlag 6 0 \
-stim_minlag 7 0 -stim_maxlag 7 0 \
-stim_minlag 8 0 -stim_maxlag 8 0 \
-stim_minlag 9 0 -stim_maxlag 9 0 \
-stim_base 4 \
-stim_base 5 \
-stim_base 6 \
-stim_base 7 \
-stim_base 8 \
-stim_base 9 \
-concat ../run_study_timepoints \
-num_glt 11 \
-glt 1 ../study_SSIvWSI.mat \
-glt_label 1 "SSI-WSI" \
-glt 1 ../study_SSIvFORGOT.mat \
-glt_label 2 "SSI-FORGOT" \
-glt 1 ../study_WSIvFORGOT.mat \
-glt_label 3 "WSI-FORGOT" \
-glt 1 ../study_SSI.mat \
-glt_label 4 "SSI" \
-glt 1 ../study_WSI.mat \
-glt_label 5 "WSI" \
-glt 1 ../study_FORGOT.mat \
-iresp 1 iresp2_SSI \
-iresp 2 iresp2_WSI \
-iresp 3 iresp2_FORGOT \
-mask run_study_mask+orig -nocout \
-fout -tout -bucket Deconv2s
Subject Author Posted

df calculation in 3dDeconvolve

Craig Stark February 10, 2003 12:25PM

Re: df calculation in 3dDeconvolve

B. Douglas Ward February 10, 2003 01:15PM

Re: df calculation in 3dDeconvolve

Craig Stark February 10, 2003 01:39PM