feat: add normal rag into framework (#360)

* add normal rag into framework

* fix a typo

* fix a bug

rdagent/scenarios/kaggle/developer/feedback.py

@@ -161,6 +161,7 @@ class KGHypothesisExperiment2Feedback(HypothesisExperiment2Feedback):
 
         if self.scen.if_using_vector_rag:
             self.scen.vector_base.add_experience_to_vector_base(experiment_feedback)
+            self.scen.vector_base.save()
         elif self.scen.if_using_graph_rag:
             self.scen.trace.knowledge_base.load_from_documents([experiment_feedback], self.scen)
 

rdagent/scenarios/kaggle/experiment/scenario.py

@@ -1,6 +1,7 @@
 import io
 import json
 import pickle
+from datetime import datetime, timezone
 from pathlib import Path
 
 import pandas as pd
@@ -43,8 +44,9 @@ class KGScenario(Scenario):
         self.if_using_vector_rag = KAGGLE_IMPLEMENT_SETTING.if_using_vector_rag
 
         if self.if_using_vector_rag and KAGGLE_IMPLEMENT_SETTING.rag_path:
-            self.vector_base = KaggleExperienceBase()
-            self.vector_base.load(KAGGLE_IMPLEMENT_SETTING.rag_path)
+            self.vector_base = KaggleExperienceBase(KAGGLE_IMPLEMENT_SETTING.rag_path)
+            self.vector_base.path = datetime.now(timezone.utc).strftime("%Y-%m-%d-%H-%M-%S") + "_kaggle_kb.pkl"
+            self.vector_base.save()
 
         self._output_format = self.output_format
         self._interface = self.interface
@@ -197,7 +199,7 @@ The model code should follow the simulator:
     @property
     def rich_style_description(self) -> str:
         return f"""
-This is the Kaggle scenario for the competition: {self.competitionn}
+This is the Kaggle scenario for the competition: {self.competition}
 """
 
     def get_scenario_all_desc(self) -> str:

rdagent/scenarios/kaggle/knowledge_management/extract_knowledge.py

@@ -70,18 +70,16 @@ def extract_knowledge_from_feedback(feedback_response: dict) -> dict:
 def process_all_case_files(directory_path: str):
     output_file = Path(directory_path) / "kaggle_experience_results.json"
     json_output = []
-    for filename in os.listdir(directory_path):
-        if filename.endswith(".case"):
-            file_path = os.path.join(directory_path, filename)
 
-            with open(file_path, "r", encoding="utf-8") as file:
-                content = file.read()
-                knowladge = extract_knowledge_from_high_score_answers(content)
-                json_output.append(knowladge)
+    for file_path in Path(directory_path).rglob("*.case"):
+        with open(file_path, "r", encoding="utf-8") as file:
+            content = file.read()
+            knowledge = extract_knowledge_from_high_score_answers(content)
+            json_output.append(knowledge)
 
     with open(output_file, "w", encoding="utf-8") as json_file:
         json.dump(json_output, json_file, ensure_ascii=False)
 
 
 if __name__ == "__main__":
-    process_all_case_files(directory_path="git_ignore_folder/experience/tabular_cases_all")
+    process_all_case_files(directory_path="git_ignore_folder/data/kaggle")

rdagent/scenarios/kaggle/knowledge_management/vector_base.py

@@ -122,11 +122,8 @@ class KaggleExperienceBase(PDVectorBase):
         super().__init__(vector_df_path)
         self.kaggle_experience_path = kaggle_experience_path
         self.kaggle_experience_data = []
-        # if path is not None and Path(path).exists():
-        #     self.load_kaggle_experience(kaggle_experience_path)
-        #     self.path = Path(path).parent / (datetime.now(timezone.utc).strftime("%Y-%m-%d-%H-%M-%S") + "_kaggle_kb.pkl")
-        # else:
-        #     pass
+        if kaggle_experience_path:
+            self.load_kaggle_experience(kaggle_experience_path)
 
     def add(self, document: Union[KGDocument, List[KGDocument]]):
         document.split_into_trunk()

rdagent/scenarios/kaggle/proposal/proposal.py

@@ -93,7 +93,10 @@ class KGHypothesisGen(ModelHypothesisGen):
         self.confidence_parameter = 1.0
         self.initial_performance = 0.0
 
-    def generate_RAG_content(self, trace: Trace) -> str:
+    def generate_RAG_content(self, trace: Trace, hypothesis_and_feedback: str) -> str:
+        if self.scem.if_using_vector_rag:
+            rag_results, _ = self.scen.vector_base.search_experience(hypothesis_and_feedback, topk_k=5)
+            return "\n".join([doc.content for doc in rag_results])
         if self.scen.if_using_graph_rag is False or trace.knowledge_base is None:
             return None
         same_competition_node = trace.knowledge_base.get_node_by_content(trace.scen.get_competition_full_desc())
@@ -235,7 +238,7 @@ class KGHypothesisGen(ModelHypothesisGen):
 
         context_dict = {
             "hypothesis_and_feedback": hypothesis_and_feedback,
-            "RAG": self.generate_RAG_content(trace),
+            "RAG": self.generate_RAG_content(trace, hypothesis_and_feedback),
             "hypothesis_output_format": Environment(undefined=StrictUndefined)
             .from_string(prompt_dict["hypothesis_output_format"])
             .render(if_using_feature_selection=KAGGLE_IMPLEMENT_SETTING.if_using_feature_selection),