Skip to content

Record

Overview

Complete storage of the computational history of: - info: an object storing the basic information of the material - history: an OrderedDict object that records the calculations ran and their outputs - results: a results object storing the results for different physical properties Note: a mat_info contains only the basic information of a material which is usually pulled directly from an existing database. 

from DeWorks.schema import Record

Attributes

Core Fields

Field Type Description
id str Unique identifier,used primarily as the _id attribute when stored in mongodb
info Info Immutable metadata describing the material
history dict[str, ProcessResolved] All performed or in-progress calculations on this material
results dict[str, PropertyResolved] Output physical properties derived from calculations

Save Setting

Field Type Description
saving_mode Literal[1, 2] 1. Saves to JSON file
2. Save to cache memory
saving_dir str Target directory for saving JSON file
Defaults to current working directory
saved_json str | None File path of save JSON, populated automatically

model_config

Configuration to allow extra fields beyond the declared schema.

getattr(self, item:str)

If a requested attribute is not found on the model, this method checks the results dictionary to retrieve the value.

model_dump(self, args, *kwargs)

Serializes the model with alias preservation.

Computed fields

Field Type Description
name str Returns the name of the material prefixed with Rc_, derived from info.name

Validation and De-serialization

default_id(cls, values:dict)

Generates _id if not provided. This fixes legacy data that did not have _id.

default_properties(cls, value: dict)

Initializes core properties (ENCUT< KSPACING, SYMPREC) with default values if they are missing

load_history(self) -> Self

Sets host object on process in history and updates them accordingly.

load(cls, value: str | dict)

Loaded from file path or dictionary 

record = Record.load("path/to/record.json")

Utilities

save(self, mode: Literal[1, 2] | None = None, dump = False)

Used to save any progress or modification to the calculations being run. - mode = 1: Saves in the local work directory as a json file - mode = 2: Saves to the history attribute of a mat_record object

confirm_overwrite(self, calculation, _confirmation: Literal["yes", "no"] | None = None)

for user to confirm whether to overwrite an existing calculations in the history.

recall(self, include, exclude)

Returns a filtered list of previous calculation from history.

progress(self, type, state, include, exclude, show) -> pd.DataFrame | JobInfo

Returns a summary of a job progress. - include: the calculations to include - exclude: the calculations to exclude 

progress(type="simple")     # percentage completed
progress(type="states")     # full job states in a table
progress(type="full_info")  # detailed job object info

status(self, include, exclude, show) -> pd.DataFrame

Returns job state (e.g., COMPLETED, FAILED) for each calculation.

list_calculations(self, show: bool = True) ->pd.DataFrame

Returns a DataFrame summarizing all stored calculations.

list_results(self, show: bool = True) -> pd.DataFrame

Restuls all the results in a pd.DataFrame

calculate(self, calculation, generat_func, overwrite, start, submit, cluster, force_confirmation, **kwargs) -> process

Initializes and optionally runs a new calculation - calculation: a process subclass or the name of the subclass in str - generator_func: A user created function that takes self. Addition process() specific arguements given as **kwargs will be ignored - overwrite: option to overwrite any existing calculations of the same name in history. - start: Whether to call the start() function of the calculation automatically or requires the user to call the function explicitly on the process obejct returned. - start+kwargs: kwargs parsed to the start() function of the process - force_confirmation: used by higher level objects to force a confirmation by using the confirmation it has received 

record.calculate(calculation="Relax", submit=True)

output_collate(self, JS, JC, include, exclude)

Aggregates results for completed jobs using JS: JobStore and JC: JobController.

write_property(self, include, exclude, **kwargs) -> pd.DataFrame

Extracts a property from each relevant calculation and stores it in results.

result_collection(self, JS, JC, include, exclude, **kwargs) -> pd.DataFrame

Combines output_collate() and write_property() as a full post-processing pipeline.

fail_assess(self, show, include, exclude, **kwargs) -> pd.DataFrame

Analyzes and returns a table of failed jobs and their issues.

rerun(self, new_cluster, new_job, new_incr, include, exclude) -> dict

Resubmits failed or incomplete jobs with new settings. Returns a dict of changes.

Note

  • This is material centric, as opposed to Project which is system-centric.