specless.inference.edsm.AutomataInferenceAlgorithm

class specless.inference.edsm.AutomataInferenceAlgorithm(binary_location: str = 'dfasat/flexfringe', output_directory: str = './')

Bases: InferenceAlgorithm

The inference algorithm for inferring an automaton from a list of Traces, where trace is defined as a sequence of symbols, i.e. a set of strings. For example, ${a, b, c}$

Parameters:

InferenceAlgorithm (_type_) – _description_

Methods

draw_IPython	Draws the dot file data in a way compatible with a jupyter / IPython notebook
draw_initial_model	Draws the initial (prefix-tree) model
draw_learned_model	Draws the final, learned model
infer	calls the flexfringe binary given the data in the training file

Attributes

initial_model_filepath	the output filename for the unlearned, initial model, as this is a different from the inputted "output-dir".
learned_model_filepath	the output filename for the fully learned model, as this is a different from the inputted "output-dir"
output_filepath	The output filepath for the results of learning the model

draw_IPython(filename: str) → None

Draws the dot file data in a way compatible with a jupyter / IPython notebook

Parameters:

filename (str) – The learned model dot file data

draw_initial_model() → None

Draws the initial (prefix-tree) model

draw_learned_model() → None

Draws the final, learned model

infer(dataset: BaseDataset, get_help: bool = False, record_time: bool = True, go_fast: bool = False, **kwargs) → Specification | Exception

calls the flexfringe binary given the data in the training file

Parameters:

• dataset (Dataset) – dataset that contains the path to the training data

• get_help (bool, optional) – Whether or not to print the flexfringe usage help memu. Defaults to False.

• record_time (bool, optional) – _description_. Defaults to True.

• go_fast (bool, optional) – optimizes this call to make it as fast as possible, at the expensive of usability. use for benchmarking / Hyperparam optimization. Defaults to False.

• kwargs (dict, optional) – controlling the learning process

Raises:

Exception – _description_

Return type:

Union[Specification, Exception]

property initial_model_filepath: str

the output filename for the unlearned, initial model, as this is a different from the inputted “output-dir”. In this case, it will be a prefix tree from the given learning data.

Returns:

The initial model filepath.

Return type:

str

property learned_model_filepath: str

the output filename for the fully learned model, as this is a different from the inputted “output-dir”

Returns:

The learned model filepath.

Return type:

str

property output_filepath: str

The output filepath for the results of learning the model