f_c.py

import numpy as np

def f_c(components: list, components_prime: list):
    """
    This functions tells us how similar two components of the two models are. It calculates the proportion of attributes that are equal.\n
    Input:
        - components: list of components of the model A.\n
        - components_prime: list of components of the model B.\n
    Output: 
        - matrix of dimension len(components) x len(components_prime).
    """
    # Auxiliary variables
    len_comp = len(components)
    len_comp_prime = len(components_prime)

    # Initialize the matrix
    f_c_matrix = np.zeros((len_comp, len_comp_prime)) 

    # Go through every component of model A and every component of model B and calculate the proportion of equal attributes
    for c in range(len_comp):
        for c_prime in range(len_comp_prime):
            # Check if the components are of the same type and if they are, calculate the proportion of equal attributes. There are exceptions/special cases for the bearing types, stage_gear_data types and the coupling types. They may be matched with each other. 
            if (components[c].type == "concept_bearing" and components_prime[c_prime].type in ["concept_bearing", "rolling_bearing_with_catalog_geometry", "rolling_bearing_with_detailed_geometry", "slide_bearing"]) or (components[c].type == "rolling_bearing_with_catalog_geometry" and components_prime[c_prime].type in ["concept_bearing", "rolling_bearing_with_catalog_geometry", "rolling_bearing_with_detailed_geometry", "slide_bearing"]) or (components[c].type == "rolling_bearing_with_detailed_geometry" and components_prime[c_prime].type in ["concept_bearing", "rolling_bearing_with_catalog_geometry", "rolling_bearing_with_detailed_geometry"]) or (components[c].type == "slide_bearing" and components_prime[c_prime].type in ["concept_bearing", "slide_bearing"]) or (components[c].type =="stage_gear_data" and "stage_gear_data" in components_prime[c_prime].type) or ("stage_gear_data" in components[c].type and components_prime[c_prime].type == "stage_gear_data") or (components[c].type in ["coupling", "switchable_coupling" ] and components_prime[c_prime].type in ["coupling", "switchable_coupling"]) or (components[c].type == components_prime[c_prime].type):
                total_number_attributes = max(len(components[c].attributes), len(components_prime[c_prime].attributes)) # Calculate the max number of attributes 
                # Special case: both components have zero attributes
                if len(components[c].attributes) == 0 and len(components_prime[c_prime].attributes) == 0:
                    f_c_matrix[c][c_prime] = 1
                # Special case: one component has zero attributes
                elif min(len(components[c].attributes), len(components_prime[c_prime].attributes)) == 0:
                    f_c_matrix[c][c_prime] = 0.1
                # General case: calculate the proportion of equal attributes
                else:
                    number_attributes_equal = 0
                    attributes = components[c].attributes
                    attributes_prime = components_prime[c_prime].attributes
                    for attribute in attributes:
                        for attribute_prime in attributes_prime:
                            if attribute.id == attribute_prime.id:
                                if attribute.param_value == attribute_prime.param_value:
                                    number_attributes_equal += 1
                    f_c_matrix[c][c_prime] = number_attributes_equal/total_number_attributes
            # If the components are not of the same type, set the value of the matrix to -10
            else:
                f_c_matrix[c][c_prime] = -10
    return f_c_matrix