Back to the main page.
Bug 769 - improve answer to faq
| Status |
CLOSED FIXED |
| Reported |
2011-06-20 09:12:00 +0200 |
| Modified |
2011-06-22 14:17:40 +0200 |
| Product: |
FieldTrip |
| Component: |
core |
| Version: |
unspecified |
| Hardware: |
PC |
| Operating System: |
Mac OS |
| Importance: |
P1 normal |
| Assigned to: |
Jan-Mathijs Schoffelen |
| URL: |
|
| Tags: |
|
| Depends on: |
|
| Blocks: |
|
| See also: |
|
Jan-Mathijs Schoffelen - 2011-06-20 09:12:44 +0200
Dear Field-trippers
I recently ran some time frequency analysis using the wavelet method.
I was surprise to see that whereas I asked for a step of 1Hz, it was not
the case every alternate step (for foi=2:1:10 ; I got [2 3.2 4 5.2 6 7.2 8
9.2 10]), and it was also the case when I changed the toi (2.5, 1.5 with
constant trials length of 2.5 sec and sampling freq 1000Hz).
I found a kind of explanation following the faq
(http://fieldtrip.fcdonders.nl/faq/why_does_my_output.freq_not_match_my_cfg.foi_when_using_wavelet_formerly_wltconvol_in_ft_freqanalyis).
Still it is not clear to me why this refers to Fourier transform.
From my (none specialist) point of view it is possible to produce a
wavelet of any size (Hz;Dt).
It is clear that there is some limitations based on the duration of the
epochs and the number of time points. But in the present case it shouldn't
be a problem, is it?
Would be please if someone can help me with this issue.
Thanks in advance
Manuel
-> this also holds for the corresponding faq about mtmconvol