Datasets
Models
Downloads: 1
57 lines (52 with data), 5.4 kB
B
Co^4 ã @ s d Z ddlmZ ddlT ddlZddlmZ ddlm Z ddl
m Z ddl Z
dd ZG dd deZdd
d ZdddZdd ZdS )a
Created by Victor Delvigne
ISIA Lab, Faculty of Engineering University of Mons, Mons (Belgium)
victor.delvigne@umons.ac.be
Source: Bashivan, et al."Learning Representations from EEG with Deep Recurrent-Convolutional Neural Networks." International conference on learning representations (2016).
Copyright (C) 2019 - UMons
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
This library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
é )ÚDataset)Ú*Nc C s t | ¡}t j |¡ t| | }g }g }x^t|D ]R}| ||| |d |
¡ | t |d||
||d | d
g¡¡ q4W ||fS )Né r éÿÿÿÿ)ÚnpÚarangeÚrandomÚshuffleÚintÚrangeÚappendÚhstack)ÚlengthZn_foldZtot_idZlen_foldZtrain_idZtest_idÚi© r ú5/home/vdelv/PycharmProjects/EEGLearn_Pytorch/Utils.pyÚkfold# s
4r c @ s( e Zd ZdZdd Zdd Zdd ZdS ) ÚEEGImagesDatasetz!EEGLearn Images Dataset from EEG.c C s || _ || _d S )N)ÚlabelÚImages)Úselfr Úimager r r Ú__init__2 s zEEGImagesDataset.__init__c C s
t | jS )N)Úlenr )r r r r Ú__len__6 s zEEGImagesDataset.__len__c C s2 t |¡r| ¡ }| j| }| j| }||f}|S )N)ÚtorchÚ is_tensorÚtolistr r )r Úidxr r Úsampler r r Ú __getitem__9 s
z EEGImagesDataset.__getitem__N)Ú__name__Ú
__module__Ú __qualname__Ú__doc__r r r r r r r r / s r Tc C s¼ d}g }xt |dD ]r\}}|\}} | tj¡}|r<| ¡ }| |}
t |
¡ jd¡\} } | | | k ¡ ¡ ||
| ¡ }
||
¡ 7 }qW ||d }dd |D }t |t | }||fS )Ng r r c S s g | ]}|D ]}|q qS r r )Ú.0ÚsublistÚitemr r r ú
<listcomp>R s z Test_Model.<locals>.<listcomp>)
Ú enumerateÚtor Úfloat32ÚcudaÚmaxÚcpuÚdatar r r' Úsumr )ÚnetÚ
TestloaderÚ criterionÚis_cudaÚ running_lossÚ
evaluationr r/ Z input_imgÚlabelsÚoutputsÚ_Ú predictedÚlossÚ running_accr r r Ú
Test_ModelD s r= é ÚSGDç-C ëâ6?é Fc C s |r| ¡ } n| } t ¡ }
|dkr:tj| ¡ |d} n,|dkrVtj| ¡ |d} ntdt d xt|D ]} d}
g }xt |dD ]~\}}|\}}|
¡ | | t j
¡ ¡ }t | ¡ jd¡\}}| ||k ¡ ¡ |
|| ¡ }| ¡ | ¡ |
| ¡ 7 }
qW |
|d }
d d
|D }t|t| }t| ||
d \}}| | |d krrtd | d ||
|||f qrW |rtd
|
|||f |
|||fS )Nr? )ÚlrÚAdamz Optimizer: z not implemented.g r r c S s g | ]}|D ]}|q qS r r )r% r& r' r r r r( z s z#TrainTest_Model.<locals>.<listcomp>TzH[%d, %3d] loss: %.3f Accuracy : %.3f val-loss: %.3f val-Accuracy : %.3fzRFinished Training
loss: %.3f Accuracy : %.3f val-loss: %.3f val-Accuracy : %.3f)r, ÚnnÚCrossEntropyLossÚoptimr? Ú
parametersrC Úprintr r) Ú zero_gradr* r r+ r- r. r/ r r ÚbackwardÚstepr' r0 r r= )ÚmodelZ trainloaderZ
testloaderÚn_epochÚoptiÚ
learning_rater4 Ú print_epochÚverboser1 r3 Ú optimizerÚepochr5 r6 r r/ Úinputsr7 r8 r9 r: r; r< Zvalidation_lossZvalidation_accr r r ÚTrainTest_ModelW sB rU c s t d¡d t d¡} | d }g x|D ]} t|¡ q*W t fddtt jd d D ¡}t d d
|i¡ t
d d S ) Nz"Sample Data/FeatureMat_timeWin.matÚfeaturesz#Sample Data/Neuroscan_locs_orig.matÚAc s> g | ]6}t t ¡ d d
|d |d d
f dddqS )NéÀ r é T)Ú normalize)Z
gen_imagesr Úarray)r% r )ÚfeatsÚlocs_2dr r r( s z create_img.<locals>.<listcomp>r rX zSample Data/images_time.matÚimgz1Images Created and Save in Sample Dat/images_time) ÚsioÚloadmatr Z azim_projr r[ r r
ÚshapeÚsavematrH )ÚlocsZlocs_3dÚeZimages_timewinr )r\ r] r Ú
create_img s
re )T)r> r? r@ TrA F)r$ Útorch.utils.data.datasetr ZUtils_Bashivanr Úscipy.ioÚior_ Ú torch.optimrF Útorch.nnrD Únumpyr r r r= rU re r r r r Ú<module> s
2