examples/aflow/scicode_full/optimized/round_4/graph.py

import evoagentx.workflow.operators as operator
import examples.aflow.scicode_full.optimized.round_4.prompt as prompt_custom
from evoagentx.models.model_configs import LLMConfig
from evoagentx.benchmark.benchmark import Benchmark
from evoagentx.models.model_utils import create_llm_instance

class Workflow:
    
    def __init__(
        self,
        name: str,
        llm_config: LLMConfig,
        benchmark: Benchmark
    ):
        self.name = name
        self.llm = create_llm_instance(llm_config)
        self.benchmark = benchmark 
        self.custom = operator.Custom(self.llm)
        self.custom_code_generate = operator.CustomCodeGenerate(self.llm)
        self.test_operator = operator.Test(self.llm)  # Initialized the test operator
        self.sc_ensemble = operator.ScEnsemble(self.llm)  # Added ensemble operator

    async def __call__(self, problem: str, entry_point: str):
        """
        Implementation of the workflow
        Custom operator to generate anything you want.
        But when you want to get standard code, you should use custom_code_generate operator.
        """
        # Generate multiple potential solutions
        solution_1 = await self.custom_code_generate(problem=problem, entry_point=entry_point, instruction=prompt_custom.GENERATE_PYTHON_CODE_PROMPT)
        solution_2 = await self.custom_code_generate(problem=problem, entry_point=entry_point, instruction=prompt_custom.GENERATE_PYTHON_CODE_PROMPT)  # Generating a second solution
        solutions = [solution_1['response'], solution_2['response']]  # Collecting both solutions
        
        # Use ScEnsemble to select the best solution from the generated solutions
        best_solution = await self.sc_ensemble.sc_ensemble(solutions=solutions, problem=problem)

        # Test the ensemble solution
        test_result = await self.test_operator.test(problem=problem, solution=best_solution['response'], entry_point=entry_point, benchmark=self.benchmark)
        
        if not test_result['result']:
            return {"success": False, "current_solution": test_result['solution']}
        return {"success": True, "final_solution": best_solution['response']}  # Return the final verified solution