0todd0000
commented
3 years ago Hi Eline,
The numbers in A and SUBJ are OK, but the order is not. The order should look something like this:
A = [0, 0, 0, ... 1, 1, 1, .... 2, 2, 2, ...]
SUBJ = [2, 3, 6, ... 2, 3, 6, .... 2, 3, 6, ...]or something like this:
A = [0, 1, 2, 0, 1, 2, 0, 1, 2, ...]
SUBJ = [2, 2, 2, 3, 3, 3, 6, 6, 6, ....]In both cases above, each subject has all three cases.
But in the printed data you pasted above, this is not the case:
A = [0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2,...]
SUBJ = [2, 3, 6, 7, 10, 11, 12, 18, 19, 2, 3, 6, ...]Notice that subject 2, for example, has three values of A=0, and no values of A=1 or A=2. For a balanced design, each subject must have ALL of A=0, A=1 and A=2.
ElineNijmeijer1
Hi Todd,
I'm trying to run a repeated measures ANOVA in python. Therefore I checked the design and it should be balanced (9 participants, 3 conditions each).
However, I still go the error that something went wrong. The first row are the conditions (9x 3 different conditions) The second row consists of the 9 subjects (measured 3 times for each condition once)
And then the shape of the data (Y) is shown, which is (27,60) (60 timepoints).
Do you have any idea what's going wrong?
Regards, Eline
[0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2] [ 2 3 6 7 10 11 12 18 19 2 3 6 7 10 11 12 18 19 2 3 6 7 10 11 12 18 19] 27 27 (27, 60) Traceback (most recent call last):
File "....\SPM test balanced designs.py", line 165, in
Frm = spm1d.stats.anova1rm(YLARGE, ALARGE, SUBJ, equal_var)
File "....spm1d\stats\anova\ui.py", line 116, in anova1rm design = designs.ANOVA1rm(A, SUBJ)
File "....spm1d\stats\anova\designs.py", line 154, in init self._check_balanced()
File "....spm1d\stats\anova\designs.py", line 178, in _check_balanced raise( ValueError('Design must be balanced.') )
ValueError: Design must be balanced.