Python module version? What are the versions of each module that you used? about eeglearn HOT 6 CLOSED

From the looks of it the error is caused by lasagne and probably due to differences in versions.

The code was tested with the following:

python 2.7 
theano 0.8 
lasagne 0.2.dev1 
CUDA 8.0

I updated the setup file to included these.

from eeglearn.

Thanks a lot! I tried with exactly the above version.
It can run Loading data...
Generating images...
Interpolating 2594/2670Interpolating 2611/2670Interpolating 2630/2670Interpolating 2648/2670Interpolating 2665/2670

Training the CNN Model...
fold 1/13
Traceback (most recent call last):
And then give me the following error:

File "C:\Users\myUser\AppData\Local\Continuum\anaconda2\lib\site-packages\theano\tensor\type.py", line 47, in init
# broadcastable is immutable, and all elements are either

AttributeError: 'TheanoConfigParser' object has no attribute 'floatX'

from eeglearn.

I see. It's weird that having the same package versions you are still getting this error. Try setting this ENV variable when running the script.

THEANO_FLAGS='floatX=float32' python <script_name.py>

from eeglearn.

Finally, it works now! Thanks so much!!

For the topology-preserving projections of electrode locations (3D-->2D), I am just wondering the input for the following function is in 3D Cartesian Coordinates (x, y, z), right? That means your sample data Neuroscan_locs_orig.mat is a Neuroscan 64 EEG channels system in Cartesian Coordinates (x, y, z), right?

def azim_proj(position):
"""
Computes the Azimuthal Equidistant Projection of input point in 3D Cartesian Coordinates.

If I change the input "position" by using let's say Biosemi (another EEG system) 128 channel location files in Cartesian Coordinates. It should also work, right?

Thanks for your help and clarification.

from eeglearn.

Yes, the inputs are in Cartesian coordinates. It should work if you replace the channel coordinates with another one of your choice (in Cartesian coordinates).

from eeglearn.

I see. Thanks so much. Really appreciate your help.

from eeglearn.