Create parameter page
+ + + + + + + + + +
+
+
+
+
+
+
+
+
+
+In [ ]:
+
+
+
+
+
+
+
+
+
+Copied!
+
+
+#Create a latex table with one column being name and the other column being description
+import pandas as pd
+from eeggan.helpers.system_inputs import default_inputs_training_autoencoder, default_inputs_training_gan, default_inputs_visualize, default_inputs_generate_samples
+from collections import OrderedDict
+#Create a latex table with one column being name and the other column being description
+import pandas as pd
+from eeggan.helpers.system_inputs import default_inputs_training_autoencoder, default_inputs_training_gan, default_inputs_visualize, default_inputs_generate_samples
+from collections import OrderedDict
+
+
+
+
+
+
+
+
+
+TODO: THE DESCRIPTIONS WERE AUTO-GENERATED, SO WE NEED TO WRITE THEM OURSELVES.
+
+
+
+
+
+
+
+
+
+In [ ]:
+
+
+
+
+
+
+
+
+
+Copied!
+
+
+#Create a latex table from a dictionary
+def create_latex_table(kw_dict):
+ #Create a dictionary with the keys being the name of the variable and the values being the description
+ kw_dict = OrderedDict(sorted(kw_dict.items(), key=lambda x: x[0]))
+ kw_dict = {k: [v[1], v[2]] for k, v in kw_dict.items()}
+ #Create a dataframe from the dictionary
+ df = pd.DataFrame.from_dict(kw_dict, orient='index', columns=['Description', 'Default'])
+
+ return df
+#Create a latex table from a dictionary
+def create_latex_table(kw_dict):
+ #Create a dictionary with the keys being the name of the variable and the values being the description
+ kw_dict = OrderedDict(sorted(kw_dict.items(), key=lambda x: x[0]))
+ kw_dict = {k: [v[1], v[2]] for k, v in kw_dict.items()}
+ #Create a dataframe from the dictionary
+ df = pd.DataFrame.from_dict(kw_dict, orient='index', columns=['Description', 'Default'])
+
+ return df
+
+
+
+
+
+
+
+
+
+In [ ]:
+
+
+
+
+
+
+
+
+
+Copied!
+
+
+#Overwrite the descriptions of the default inputs for the GAN, autoencoder, visualization, and generation of samples functions
+def overwrite_descriptions(function = 'GAN', kw_dict=None):
+
+ '''
+ Overwrite the descriptions of the default inputs for the GAN, autoencoder, visualization, and generation of samples functions.
+
+ '''
+ if function.lower() == 'gan':
+ definitions = {
+ 'ddp': 'Whether to use DistributedDataParallel for training the GAN.',
+ 'seed': 'The seed to use for reproducibility.',
+ 'n_epochs': 'The number of epochs to train the GAN.',
+ 'batch_size': 'The batch size to use during training.',
+ 'sample_interval': 'The interval at which to save samples of the generated data.',
+ 'hidden_dim': 'The dimension of the hidden layers in the generator and discriminator.',
+ 'num_layers': 'The number of layers in the generator and discriminator.',
+ 'patch_size': 'The size of the patches used for the PatchGAN discriminator.',
+ 'discriminator_lr': 'The learning rate for the discriminator.',
+ 'generator_lr': 'The learning rate for the generator.',
+ 'data': 'The EEG data to train the GAN on.',
+ 'checkpoint': 'The directory to save the model checkpoints.',
+ 'autoencoder': 'The trained autoencoder to use for training the GAN.',
+ 'kw_conditions': 'The conditions to use for training the GAN.',
+ 'kw_time': 'The time to use for training the GAN.',
+ 'kw_channel': 'The channel to use for training the GAN.',
+ 'save_name': 'The name to save the trained GAN.',
+ }
+
+ elif function.lower() == 'autoencoder':
+ definitions = {
+ 'ddp': 'Whether to use DistributedDataParallel for training the autoencoder.',
+ 'load_checkpoint': 'Whether to load a checkpoint for the autoencoder.',
+ 'seed': 'The seed to use for reproducibility.',
+ 'data': 'The EEG data to train the autoencoder on.',
+ 'checkpoint': 'The directory to save the model checkpoints.',
+ 'save_name': 'The name to save the trained autoencoder.',
+ 'target': 'The target to use for training the autoencoder.',
+ 'kw_time': 'The time to use for training the autoencoder.',
+ 'kw_channel': 'The channel to use for training the autoencoder.',
+ 'activation': 'The activation function to use in the autoencoder.',
+ 'channels_out': 'The number of channels in the output of the autoencoder.',
+ 'time_out': 'The number of time points in the output of the autoencoder.',
+ 'n_epochs': 'The number of epochs to train the autoencoder.',
+ 'batch_size': 'The batch size to use during training.',
+ 'sample_interval': 'The interval at which to save samples of the generated data.',
+ 'hidden_dim': 'The dimension of the hidden layers in the autoencoder.',
+ 'num_layers': 'The number of layers in the autoencoder.',
+ 'num_heads': 'The number of attention heads to use in the transformer.',
+ 'train_ratio': 'The ratio of the data to use for training the autoencoder.',
+ 'learning_rate': 'The learning rate for the autoencoder.',
+ }
+
+ elif function.lower() == 'visualization':
+ definitions = {
+ 'loss': 'Whether to plot the loss of the model.',
+ 'average': 'Whether to plot the average of the data.',
+ 'pca': 'Whether to use PCA for visualization.',
+ 'tsne': 'Whether to use t-SNE for visualization.',
+ 'spectogram': 'Whether to use a spectogram for visualization.',
+ 'fft': 'Whether to use FFT for visualization.',
+ 'channel_plots': 'Whether to plot the channels.',
+ 'model': 'The trained model to visualize.',
+ 'data': 'The data to visualize.',
+ 'comp_data': 'The compressed data to visualize.',
+ 'kw_conditions': 'The conditions to use for visualization.',
+ 'kw_time': 'The time to use for visualization.',
+ 'kw_channel': 'The channel to use for visualization.',
+ 'n_samples': 'The number of samples to visualize.',
+ 'channel_index': 'The index of the channel to visualize.',
+ 'tsne_perplexity': 'The perplexity to use for t-SNE iterations.',
+ 'tsne_iterations': 'The number of iterations to use for t-SNE.',
+ }
+
+ elif function.lower() == 'generate_samples':
+ definitions = {
+ 'seed': 'The seed to use for reproducibility.',
+ 'model': 'The trained model to use for generating samples.',
+ 'save_name': 'The name to save the generated samples.',
+ 'kw_time': 'The time to use for generating samples.',
+ 'sequence_length': 'The length of the sequence to generate.',
+ 'num_samples_total': 'The total number of samples to generate.',
+ 'num_samples_parallel': 'The number of samples to generate in parallel.',
+ 'conditions': 'The conditions to use for generating samples.',
+ }
+
+ else:
+ print('Invalid function name. Please enter either "GAN", "autoencoder", "visualization", or "generate_samples".')
+
+ #Overwrite the descriptions of the default inputs
+ for key in kw_dict.keys():
+ if key in definitions.keys():
+ kw_dict[key][1] = definitions[key]
+
+ return kw_dict
+#Overwrite the descriptions of the default inputs for the GAN, autoencoder, visualization, and generation of samples functions
+def overwrite_descriptions(function = 'GAN', kw_dict=None):
+
+ '''
+ Overwrite the descriptions of the default inputs for the GAN, autoencoder, visualization, and generation of samples functions.
+
+ '''
+ if function.lower() == 'gan':
+ definitions = {
+ 'ddp': 'Whether to use DistributedDataParallel for training the GAN.',
+ 'seed': 'The seed to use for reproducibility.',
+ 'n_epochs': 'The number of epochs to train the GAN.',
+ 'batch_size': 'The batch size to use during training.',
+ 'sample_interval': 'The interval at which to save samples of the generated data.',
+ 'hidden_dim': 'The dimension of the hidden layers in the generator and discriminator.',
+ 'num_layers': 'The number of layers in the generator and discriminator.',
+ 'patch_size': 'The size of the patches used for the PatchGAN discriminator.',
+ 'discriminator_lr': 'The learning rate for the discriminator.',
+ 'generator_lr': 'The learning rate for the generator.',
+ 'data': 'The EEG data to train the GAN on.',
+ 'checkpoint': 'The directory to save the model checkpoints.',
+ 'autoencoder': 'The trained autoencoder to use for training the GAN.',
+ 'kw_conditions': 'The conditions to use for training the GAN.',
+ 'kw_time': 'The time to use for training the GAN.',
+ 'kw_channel': 'The channel to use for training the GAN.',
+ 'save_name': 'The name to save the trained GAN.',
+ }
+
+ elif function.lower() == 'autoencoder':
+ definitions = {
+ 'ddp': 'Whether to use DistributedDataParallel for training the autoencoder.',
+ 'load_checkpoint': 'Whether to load a checkpoint for the autoencoder.',
+ 'seed': 'The seed to use for reproducibility.',
+ 'data': 'The EEG data to train the autoencoder on.',
+ 'checkpoint': 'The directory to save the model checkpoints.',
+ 'save_name': 'The name to save the trained autoencoder.',
+ 'target': 'The target to use for training the autoencoder.',
+ 'kw_time': 'The time to use for training the autoencoder.',
+ 'kw_channel': 'The channel to use for training the autoencoder.',
+ 'activation': 'The activation function to use in the autoencoder.',
+ 'channels_out': 'The number of channels in the output of the autoencoder.',
+ 'time_out': 'The number of time points in the output of the autoencoder.',
+ 'n_epochs': 'The number of epochs to train the autoencoder.',
+ 'batch_size': 'The batch size to use during training.',
+ 'sample_interval': 'The interval at which to save samples of the generated data.',
+ 'hidden_dim': 'The dimension of the hidden layers in the autoencoder.',
+ 'num_layers': 'The number of layers in the autoencoder.',
+ 'num_heads': 'The number of attention heads to use in the transformer.',
+ 'train_ratio': 'The ratio of the data to use for training the autoencoder.',
+ 'learning_rate': 'The learning rate for the autoencoder.',
+ }
+
+ elif function.lower() == 'visualization':
+ definitions = {
+ 'loss': 'Whether to plot the loss of the model.',
+ 'average': 'Whether to plot the average of the data.',
+ 'pca': 'Whether to use PCA for visualization.',
+ 'tsne': 'Whether to use t-SNE for visualization.',
+ 'spectogram': 'Whether to use a spectogram for visualization.',
+ 'fft': 'Whether to use FFT for visualization.',
+ 'channel_plots': 'Whether to plot the channels.',
+ 'model': 'The trained model to visualize.',
+ 'data': 'The data to visualize.',
+ 'comp_data': 'The compressed data to visualize.',
+ 'kw_conditions': 'The conditions to use for visualization.',
+ 'kw_time': 'The time to use for visualization.',
+ 'kw_channel': 'The channel to use for visualization.',
+ 'n_samples': 'The number of samples to visualize.',
+ 'channel_index': 'The index of the channel to visualize.',
+ 'tsne_perplexity': 'The perplexity to use for t-SNE iterations.',
+ 'tsne_iterations': 'The number of iterations to use for t-SNE.',
+ }
+
+ elif function.lower() == 'generate_samples':
+ definitions = {
+ 'seed': 'The seed to use for reproducibility.',
+ 'model': 'The trained model to use for generating samples.',
+ 'save_name': 'The name to save the generated samples.',
+ 'kw_time': 'The time to use for generating samples.',
+ 'sequence_length': 'The length of the sequence to generate.',
+ 'num_samples_total': 'The total number of samples to generate.',
+ 'num_samples_parallel': 'The number of samples to generate in parallel.',
+ 'conditions': 'The conditions to use for generating samples.',
+ }
+
+ else:
+ print('Invalid function name. Please enter either "GAN", "autoencoder", "visualization", or "generate_samples".')
+
+ #Overwrite the descriptions of the default inputs
+ for key in kw_dict.keys():
+ if key in definitions.keys():
+ kw_dict[key][1] = definitions[key]
+
+ return kw_dict
+
+
+
+
+
+
+
+
+
+In [ ]:
+
+
+
+
+
+
+
+
+
+Copied!
+
+
+#Main function
+def main(overwrite_desc=True):
+ #Get the default inputs for the training of the autoencoder
+ kw_dict = default_inputs_training_autoencoder()
+ #Overwrite the descriptions of the default inputs for the autoencoder
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('autoencoder', kw_dict)
+ #Create a latex table from the dictionaryw
+ ae_table = create_latex_table(kw_dict)
+
+ #Get the default inputs for the training of the GAN
+ kw_dict = default_inputs_training_gan()
+ #Overwrite the descriptions of the default inputs for the GAN
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('GAN', kw_dict)
+ #Create a latex table from the dictionary
+ gan_table = create_latex_table(kw_dict)
+
+ #Get the default inputs for the visualization
+ kw_dict = default_inputs_visualize()
+ #Overwrite the descriptions of the default inputs for the visualization
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('visualization', kw_dict)
+ #Create a latex table from the dictionary
+ visualize_table = create_latex_table(kw_dict)
+
+ #Get the default inputs for the generation of samples
+ kw_dict = default_inputs_generate_samples()
+ #Overwrite the descriptions of the default inputs for the generation of samples
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('generate_samples', kw_dict)
+ #Create a latex table from the dictionary
+ generate_table = create_latex_table(kw_dict)
+
+ #Load and Save Tables as an .md file
+ with open("docs/HowToUse/parameters.md", "w") as f:
+ #Format the markdown file
+ f.write(f"---")
+ f.write(f"hide:")
+ f.write(f" -toc")
+ f.write(f"---")
+
+ #Write the title of the markdown file
+ f.write(f"# Function Parameters\n\n")
+ f.write(f"This page contains the default parameters for the functions in the eeggan package. The parameters are organized by function and are listed in a table with the parameter name and a description of the parameter.\n\n")
+
+ #Write the tables for the autoencoder, GAN, visualization, and generation of samples
+ f.write(f"## Autoencoder Training\n\n")
+ f.write(f"The autoencoder training function is used to train an autoencoder on EEG data. The autoencoder is trained to learn a compressed representation of the data. It can be incorporated into a GAN to improve the quality of the generated samples. The function has the following parameters:\n\n")
+ f.write(f"{ae_table.to_markdown()}\n\n")
+
+ f.write(f"## GAN Training\n\n")
+ f.write(f"The GAN training function is used to train a Generative Adversarial Network (GAN) on EEG data. The GAN is trained to generate realistic samples of EEG data. The function has the following parameters:\n\n")
+ f.write(f"{gan_table.to_markdown()}\n\n")
+
+ f.write(f"## Visualization\n\n")
+ f.write(f"The visualization function is used to visualize the results of training an autoencoder or GAN. The function has the following parameters:\n\n")
+ f.write(f"{visualize_table.to_markdown()}\n\n")
+
+ f.write(f"## Generate Samples\n\n")
+ f.write(f"The generate samples function is used to generate samples of EEG data using a trained GAN. The function has the following parameters:\n\n")
+ f.write(f"{generate_table.to_markdown()}\n\n")
+#Main function
+def main(overwrite_desc=True):
+ #Get the default inputs for the training of the autoencoder
+ kw_dict = default_inputs_training_autoencoder()
+ #Overwrite the descriptions of the default inputs for the autoencoder
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('autoencoder', kw_dict)
+ #Create a latex table from the dictionaryw
+ ae_table = create_latex_table(kw_dict)
+
+ #Get the default inputs for the training of the GAN
+ kw_dict = default_inputs_training_gan()
+ #Overwrite the descriptions of the default inputs for the GAN
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('GAN', kw_dict)
+ #Create a latex table from the dictionary
+ gan_table = create_latex_table(kw_dict)
+
+ #Get the default inputs for the visualization
+ kw_dict = default_inputs_visualize()
+ #Overwrite the descriptions of the default inputs for the visualization
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('visualization', kw_dict)
+ #Create a latex table from the dictionary
+ visualize_table = create_latex_table(kw_dict)
+
+ #Get the default inputs for the generation of samples
+ kw_dict = default_inputs_generate_samples()
+ #Overwrite the descriptions of the default inputs for the generation of samples
+ if overwrite_desc:
+ kw_dict = overwrite_descriptions('generate_samples', kw_dict)
+ #Create a latex table from the dictionary
+ generate_table = create_latex_table(kw_dict)
+
+ #Load and Save Tables as an .md file
+ with open("docs/HowToUse/parameters.md", "w") as f:
+ #Format the markdown file
+ f.write(f"---")
+ f.write(f"hide:")
+ f.write(f" -toc")
+ f.write(f"---")
+
+ #Write the title of the markdown file
+ f.write(f"# Function Parameters\n\n")
+ f.write(f"This page contains the default parameters for the functions in the eeggan package. The parameters are organized by function and are listed in a table with the parameter name and a description of the parameter.\n\n")
+
+ #Write the tables for the autoencoder, GAN, visualization, and generation of samples
+ f.write(f"## Autoencoder Training\n\n")
+ f.write(f"The autoencoder training function is used to train an autoencoder on EEG data. The autoencoder is trained to learn a compressed representation of the data. It can be incorporated into a GAN to improve the quality of the generated samples. The function has the following parameters:\n\n")
+ f.write(f"{ae_table.to_markdown()}\n\n")
+
+ f.write(f"## GAN Training\n\n")
+ f.write(f"The GAN training function is used to train a Generative Adversarial Network (GAN) on EEG data. The GAN is trained to generate realistic samples of EEG data. The function has the following parameters:\n\n")
+ f.write(f"{gan_table.to_markdown()}\n\n")
+
+ f.write(f"## Visualization\n\n")
+ f.write(f"The visualization function is used to visualize the results of training an autoencoder or GAN. The function has the following parameters:\n\n")
+ f.write(f"{visualize_table.to_markdown()}\n\n")
+
+ f.write(f"## Generate Samples\n\n")
+ f.write(f"The generate samples function is used to generate samples of EEG data using a trained GAN. The function has the following parameters:\n\n")
+ f.write(f"{generate_table.to_markdown()}\n\n")
+
+
+
+
+
+
+
+
+
+
+In [ ]:
+
+
+
+
+
+
+
+
+
+Copied!
+
+
+if __name__ == '__main__':
+ overwrite_desc = False #Whether to overwrite the descriptions of the default inputs
+ main(overwrite_desc=overwrite_desc)
+if __name__ == '__main__':
+ overwrite_desc = False #Whether to overwrite the descriptions of the default inputs
+ main(overwrite_desc=overwrite_desc)
+
-