waver weirdness

AFNI Message Board
1

I expect to get hollered at for asking about a program that’s no longer supported, but waver seems a little better suited to my purpose (writing a simple HRF demo in Python for a class) than 3dDeconvolve.

‘waver -GAM -TR 1 -peak 1 -undershoot 0.2 -tstim 0x0.75:1 15x0.75:1’ inexplicably gives me this output:

0
0
0
0.0106119
0.110515
0.238478
0.360398
0.400881
0.336269
0.237694
0.134989
0.0331447
-0.0533985
-0.0745188
-0.0268729
0
0
0
0.141155
0.375
0.608845
0.75
0.663465
0.488886
0.3
0.111114
-0.0634648
-0.15
-0.075
0

whereas ‘waver -GAM -TR 1 -peak 1 -undershoot 0.2 -tstim 1x0.75:1 15x0.75:1’ gives me this:

0
0
0
0.141155
0.375
0.608845
0.75
0.663465
0.488886
0.3
0.111114
-0.0634648
-0.15
-0.075
0
0
0
0.141155
0.375
0.608845
0.75
0.663465
0.488886
0.3
0.111114
-0.0634648
-0.15
-0.075
0

In both cases, the two responses should be the same size, but for some reason, if I start the first stimulus at time point zero, its response is smaller than it ought to be.

waver_HRFs.PNG

2

Hi Phil,

That format is not quite correct. It should be START:ENDxHEIGHT.
Having the end time (or duration?) at the end is not quite right, and
may confuse the reading function.

So replace ‘0x0.75:1 15x0.75:1’ with ‘0:1x0.75 15x0.75’. And note
that the :1 is not necessary, since it equals the TR. With -TR 2,
the 1 second duration would make it smaller. So you might really
just want ‘0x0.75 15x0.75’

Note that the case of ‘1x0.75:1’ should mean the same as ‘1x0.75’.
Also 15x0.75:1 should not even be correct, since 1 would not be a
valid end time.

  • rick
3

Thanks! It’s working as it should now.