Class: MatchFinder

A class which uses optimization results to find real material matches in the Materials Project databases, using the Materials Project API.
Built on the Pydantic BaseModel.

Attributes

  • optimization_params, type: OptimizationParams
  • optimized_population, type: Population
  • ga_params, type: GeneticAlgorithmParams

Class Methods

  • get_dict_of_best_designs

    • Constructs a dictionary containing the best designs found.
    • Returns: best_designs_dict (dict)

  • get_material_matches

    • Identifies materials in the MP database which match those recommended by the optimization.
    • Args:
      • overall_bounds_dict (dict, optional)
      • consolidated_dict (dict, optional)
      • threshold (float, optional)
        • Should be between 0 and 1, by default 1
    • Returns: best_designs_dict (dict)
      • final_matching_materials (dict)
        • Keys are fake materials recommended by the genetic algorithm
        • Values are mp-ids of real materials

  • get_all_possible_vol_frac_combos

    • Computes the optimal volume fractions of known materials.
    • Once real materials have been identified, we must calculate which volume fraction combinations are best for the composite.
    • Args: num_fractions (int)
    • Returns: all_vol_frac_combos (list)

  • generate_consolidated_dict

    • Main function used to generate material property dictionary depending on user request
    • Args:
      • overall_bounds_dict (dict)
          • Dictionary of bounds that encompass all the search bounds individually defined for constituent materials.
    • Returns:
      • consolidated_dict (dict)
        • Keyed by constituent materials
        • Items are lists of dictionaries keyed by mp-ids with properties which match the bounds criteria

  • get_material_match_costs

    • Evaluates the real candidate composites with the same cost function used for optimization.
    • Args:
      • matches_dict (dict)
      • consolidated_dict (dict)
    • Returns:
      • plotly.graph_objects.Figure
        • A table of the matches and their costs as evaluated by the genetic algorithm