yaml_helpers module

class itrails.yaml_helpers.FlowSeq(iterable=(), /)[source]

Bases: list

A list subclass that indicates the sequence should be represented inline in YAML.

This class is used to signal that the list is to be serialized using flow style (i.e. inline) when dumped by PyYAML.

itrails.yaml_helpers.flow_seq_representer(dumper, data)[source]

Represent a FlowSeq instance in YAML flow style.

This custom representer instructs PyYAML to dump FlowSeq instances using a flow-style sequence.

Parameters:

  • dumper – The YAML dumper instance.

  • data – The FlowSeq data to be represented.

Returns:

A YAML node representing the sequence in flow style.

itrails.yaml_helpers.load_config(config_file)[source]

Load a YAML configuration file.

Reads the YAML file specified by config_file and returns the parsed configuration. If an error occurs during file reading or parsing, the error is printed to stderr and the process exits.

Parameters:

config_file (str) – Path to the YAML configuration file.

Returns:

Parsed configuration data.

Return type:

Any

itrails.yaml_helpers.update_best_model(best_model_yaml, optim_variables, current_optim_params, current_result, iteration)[source]

Update the YAML file storing the best model information.

This function compares the current log likelihood with the stored value in the YAML file and updates the file if the current value is better (i.e. higher than the stored value). It updates the following fields in the YAML data:

  • optimized_parameters: A dictionary mapping optimized parameter names to their adjusted values. The adjustment is done by scaling each parameter using the fixed parameter mu found in the YAML file’s fixed_parameters field. For the parameter named “r”, the value is multiplied by mu; for all other parameters, the value is divided by mu.

  • results: A dictionary containing the best log likelihood value and the corresponding iteration number.

Note: The YAML file is expected to exist and contain the required fields (fixed_parameters and results). If the file does not exist or cannot be read, the function raises an error or exits.

Parameters:

  • best_model_yaml (str) – Path to the YAML file (e.g., “best_model.yaml”).

  • optim_variables (list) – List of names for optimized parameters.

  • current_optim_params (list) – List of current optimized parameter values corresponding to optim_variables.

  • current_result (float) – The current log likelihood value to compare against the stored value.

  • iteration (int) – The current iteration number in the model training process.