searchep.eval#
Quick evaluation of trained models and new calculations
- searchep.eval.data_kappa_ev_ts_file(filename: str, training_data: Data)#
Kappa-, eigenvalues and training steps are written to a csv file
Only the eigenvalues belonging to the EP are written to the csv. Usually used at the end of training to facilitate access to data.
- Parameters:
filename (str) – Name of the crated csv file
training_data (data.Data) – Data class which contains all relevant values
- searchep.eval.eigenvalue_space_plotly(filename: str)#
Plotly plot of the energy / eigenvalue space
Reads an output file to get the kappa-, eigen-, and phi-values. Plot is used to identify exchange behavior of the eigenvalues which belong to an EP.
- Parameters:
filename (str) – Name of the output file to be read
- searchep.eval.kappa_space_training_plotly(filename: str, training_data: Data, show: bool = False)#
Plotly plot of trained kappa space
Creates a html file in which the plot is saved. Usually used at the end of the training to instantly evaluate the results.
- Parameters:
filename (str) – Name of the created html file
training_data (data.Data) – Data class which contains all relevant values
show (bool, optional) – If True, the plot is displayed
- searchep.eval.save_kernel_evs(filename: str, training_data: Data)#
Writes all kernel eigenvalues in every training step to a new pickle file
Usually used after a training loop for further evaluation of the training process.
- Parameters:
filename (str) – Name of the created pickle file
training_data (data.Data) – Data class which contains all relevant values
- searchep.eval.write_new_dataset(filename: str, init_data: Data)#
Writes kappa-, eigen- and phi-values to a new csv file
Usually used after the initial_dataset function to save the obtained eigenvalues belonging to the EP.
- Parameters:
filename (str) – Name of the created csv file
init_data (data.Data) – Data class which contains all relevant values