polychron.models.MCMCData Module

MCMCData dataclass

Dataclass containing all of the MCMC results.

This enables lighter-weight "save" behavior

Note

Optional[foo] must be used rather than from __future__ import annotations and foo | None in python 3.9 due to use of get_type_hints

contexts class-attribute instance-attribute

contexts: List[str] = field(default_factory=list)

List of contexts passed in to and returned by run_MCMC, stored in topological order

accept_samples_context class-attribute instance-attribute

accept_samples_context: List[List[float]] = field(default_factory=list)

A list of accepted samples from the MCMC process.

The number of accepted samples is important.

Formerly StartPage.ACCEPT

accept_samples_phi class-attribute instance-attribute

accept_samples_phi: List[List[float]] = field(default_factory=list)

Accepted group boundaries from MCMC simulation

Formerly StartPage.PHI_ACCEPT

A class-attribute instance-attribute

A: int = field(default=0)

The newer limit for the Cal BP range to be considered considered during MCMC, denoting the most recent time.

Formerly StartPage.A

P class-attribute instance-attribute

P: int = field(default=0)

The older limit for the Cal BP range to be considered considered during MCMC

Formerly StartPage.P

all_samples_context class-attribute instance-attribute

all_samples_context: List[List[float]] = field(default_factory=list)

A list of lists containing all samples for contexts from MCMC, including rejected samples.

I.e. accept_samples_context is a subset of this.

Formerly StartPage.ALL_SAMPS_CONT

all_samples_phi class-attribute instance-attribute

all_samples_phi: List[List[float]] = field(default_factory=list)

A list of lists, containing all samples for group boundaries from MCMC, including rejected samples.

I.e. accept_samples_phi is a subset of this.

Formerly StartPage.ALL_SAMPS_PHI

accept_group_limits class-attribute instance-attribute

accept_group_limits: Dict[str, List[float]] = field(default_factory=dict)

A dictionary of group limits from accepted phi samples. Produced in phase_labels.

Formerly StartPage.resultsdict

all_group_limits class-attribute instance-attribute

all_group_limits: Dict[str, List[float]] = field(default_factory=dict)

A dictionary of all_results? returned by MCMC_func, which is used to find the phase lengths during node finding on the results page.

Formerly StartPage.all_results_dict

save_results_dataframes

save_results_dataframes(path: Path, group_df: DataFrame) -> None

Save some MCMC data to disk, separately from the serialised version of this class

Formerly part of StartPage.save_state_1

Parameters:

Name	Type	Description	Default
path	Path	The path to the directory in which files should be created	required
group_df	DataFrame	A pandas dataframe containing context group information.	required

Todo

• Check if full_results_df include the first 10000 elements or not?

to_json

to_json(pretty: bool = False) -> str

Serialise this object to JSON

Parameters:

Name	Type	Description	Default
pretty	bool	If the json should be prettified or not.	False

Return

JSON string representing the MCMCData instance

save

save(path: Path, group_df: DataFrame, verbose: bool = False) -> None

Save the current state of this file to the specified path

Parameters:

Name	Type	Description	Default
path	Path	The directory in which the files will be saved.	required
group_df	DataFrame	A pandas dataframe containing context group information	required
verbose	bool	If verbose output should be used	False

Raises:

Type	Description
RuntimeError	If the provided path is not a directory
Exception	If other exceptions are raised during saving of files to disk

load_from_disk classmethod

load_from_disk(json_path: Path) -> MCMCData

Get an instance of the MCMCData from serialised json on disk.

Parameters:

Name	Type	Description	Default
json_path	Path	Path to json file to load it from	required

Returns:

Type	Description
MCMCData	MCMCData instance

Raises:

Type	Description
RuntimeError	If the json_path is not an existing file; or the 'polychron_version' key is missing from the json file'; or the 'mcmc_data' key is missing from the json file
JSONDecodeError	If the json_path does not point to a valid json file