Data loading and trial extraction

[mhmmd-hsn]

Hello, hope you are doing great.
I'm trying to code what you have done in "Decoding Bilingual EEG Signals With Complex Semantics Using Adaptive Graph Attention Convolutional Network" and I have some questions about the dataset. based on what has been said in Fig2 in the paper you distinguish signals based on numbers 1 to 9 or 7 for different classes, numbers 21 and 22 for reading, and 30 and 31 for the speaking part. but I cannot find these numbers in the evt.bdf files! this is what I have printed from 2 different evt files.

I can't understand what these numbers even mean, It doesn't match with anything!

how can I extract reading trials from raw data?

[cfli20]

Hello, hope you are doing great. I'm trying to code what you have done in "Decoding Bilingual EEG Signals With Complex Semantics Using Adaptive Graph Attention Convolutional Network" and I have some questions about the dataset. based on what has been said in Fig2 in the paper you distinguish signals based on numbers 1 to 9 or 7 for different classes, numbers 21 and 22 for reading, and 30 and 31 for the speaking part. but I cannot find these numbers in the evt.bdf files! this is what I have printed from 2 different evt files.

I can't understand what these numbers even mean, It doesn't match with anything!

how can I extract reading trials from raw data?

The readout is not very complete when using Python; there may be encoding or formatting problems.

We recommend that you use MATLAB.
You need to install the plugin and select both data.bdf and evt.bdf when reading.
You can refer to https://blog.csdn.net/goolete/article/details/129500501

Read the labels and then crop the individual tasks to a uniform size.

[mhmmd-hsn]

I tried what you said, but I have a problem understanding the explanations in the link (I have no experience with Chinese) and I even can not download neuracleEEGFileReader; guess there are some problems with sign up and login.

can you share the processed data with me? Is shared data processed at all? or it is complete raw data?

[cfli20]

I tried what you said, but I have a problem understanding the explanations in the link (I have no experience with Chinese) and I even can not download neuracleEEGFileReader; guess there are some problems with sign up and login.

can you share the processed data with me? Is shared data processed at all? or it is complete raw data?

The shared data is raw data.
I have uploaded the plugin. You can download the plugin from https://drive.google.com/drive/folders/1vSg8RZJ0iVUdv5QvBLvrSFiQGyCg_AtE.
The release note contains detailed installation instructions in English.

[mhmmd-hsn]

I will be happy if you give me access. Just so you know, I have already asked for permission.
Also, I have another question. When I try to load and plot events (with Python and Matplotlib), I see these event markers in the file. It sounds correct, but after testing different files, all I see is class 1 to 3 in each event file, which form different data! All the files look kind of the same. Based on what has been written in the paper, there should be 9 different classes, but I can not see anything more than 3 classes in the dataset.

[cfli20]

I will be happy if you give me access. Just so you know, I have already asked for permission. Also, I have another question. When I try to load and plot events (with Python and Matplotlib), I see these event markers in the file. It sounds correct, but after testing different files, all I see is class 1 to 3 in each event file, which form different data! All the files look kind of the same. Based on what has been written in the paper, there should be 9 different classes, but I can not see anything more than 3 classes in the dataset.

As I mentioned above. The readout is not very complete when using Python or Matploylib; there may be encoding or formatting problems. We recommend that you use MATLAB.
You need to install the plugin and select both data.bdf and evt.bdf when reading.
You can refer to https://blog.csdn.net/goolete/article/details/129500501

Read the labels and then crop the individual tasks to a uniform size.

[mhmmd-hsn]

you said, but you did'nt give me permission to drive!
Also, that problem is solved, I just load data in matlab then save it in csv format so I can use it in Python. And I'm sure that it works fine.