String together the entire factor process

rdagent/app/qlib_rd_loop/conf.py

@@ -25,5 +25,6 @@ class PropSetting(BaseSettings):
     evolving_n: int = 10
     
     py_bin: str = "/usr/bin/python"
+    local_qlib_folder: Path = Path("/home/rdagent/qlib")
     
 PROP_SETTING = PropSetting()

rdagent/app/qlib_rd_loop/factor.py

@@ -25,6 +25,7 @@ hypothesis_gen: HypothesisGen = import_class(PROP_SETTING.qlib_factor_hypothesis
 hypothesis2experiment: Hypothesis2Experiment = import_class(PROP_SETTING.qlib_factor_hypothesis2experiment)()
 
 qlib_factor_coder: TaskGenerator = import_class(PROP_SETTING.qlib_factor_coder)(scen)
+
 qlib_factor_runner: TaskGenerator = import_class(PROP_SETTING.qlib_factor_runner)(scen)
 
 qlib_factor_summarizer: HypothesisExperiment2Feedback = import_class(PROP_SETTING.qlib_factor_summarizer)(scen)
@@ -38,21 +39,4 @@ for _ in range(PROP_SETTING.evolving_n):
     exp = qlib_factor_runner.generate(exp)
     feedback = qlib_factor_summarizer.generateFeedback(exp, hypothesis, trace)
 
-    trace.hist.append((hypothesis, exp, feedback))
-
-
-"""
-trace = Trace(scen=scen)
-# for _ in range(PROP_SETTING.evolving_n):
-for _ in range(1):
-    hypothesis = hypothesis_gen.gen(trace)
-    exp = hypothesis2experiment.convert(hypothesis, trace)
-    # exp = qlib_factor_coder.generate(exp)
-    import pickle
-    file_path = '/home/finco/v-yuanteli/RD-Agent/git_ignore_folder/factor_data_output/exp_new.pkl'
-    with open(file_path, 'rb') as file:
-        exp = pickle.load(file)
-    exp = qlib_factor_runner.generate(exp)
-    feedback = qlib_factor_summarizer.generateFeedback(exp, hypothesis, trace)
-    # trace.hist.append((hypothesis, exp, feedback))
-"""
+    trace.hist.append((hypothesis, exp, feedback))

rdagent/components/coder/factor_coder/CoSTEER/evolving_strategy.py

@@ -149,7 +149,7 @@ class FactorEvolvingStrategy(MultiProcessEvolvingStrategy):
                     queried_former_failed_knowledge=queried_former_failed_knowledge_to_render,
                 )
             )
-            session = APIBackend(use_chat_cache=True).build_chat_session(
+            session = APIBackend(use_chat_cache=False).build_chat_session(
                 session_system_prompt=system_prompt,
             )
 
@@ -249,7 +249,7 @@ class FactorEvolvingStrategyWithGraph(MultiProcessEvolvingStrategy):
                 )
             )
 
-            session = APIBackend(use_chat_cache=True).build_chat_session(
+            session = APIBackend(use_chat_cache=False).build_chat_session(
                 session_system_prompt=system_prompt,
             )
 
@@ -276,7 +276,7 @@ class FactorEvolvingStrategyWithGraph(MultiProcessEvolvingStrategy):
                         )
                         .strip("\n")
                     )
-                    session_summary = APIBackend(use_chat_cache=True).build_chat_session(
+                    session_summary = APIBackend(use_chat_cache=False).build_chat_session(
                         session_system_prompt=error_summary_system_prompt,
                     )
                     for _ in range(10):  # max attempt to reduce the length of error_summary_user_prompt

rdagent/scenarios/qlib/prompts.yaml

@@ -67,13 +67,13 @@ data_feedback_generation:
     }
   user: |-
     Target hypothesis: 
-    {{hypothesis}}
+    {{ hypothesis_text }}
     Tasks and Factors:
-    {{task_details}}
+    {{ task_details }}
     Current Result: 
-    {{current_result}}
+    {{ current_result }}
     SOTA Result: 
-    {{sota_result}}
+    {{ sota_result }}
     Analyze the current result in the context of its ability to:
     1. Support or refute the hypothesis.
     2. Show improvement or deterioration compared to the last experiment.

rdagent/scenarios/qlib/task_generator/data.py

@@ -24,6 +24,8 @@ logger = RDAgentLog()
 #         de = DockerEnv()
 #         de.run(local_path=self.ws_path, entry="qrun conf.yaml")
 
+# TODO: supporting multiprocessing and keep previous results
+
 class QlibFactorRunner(TaskGenerator[QlibFactorExperiment]):
     """
     Docker run
@@ -32,8 +34,6 @@ class QlibFactorRunner(TaskGenerator[QlibFactorExperiment]):
     - price-volume data dumper
     - `data.py` + Adaptor to Factor implementation
     - results in `mlflow`
-
-    - TODO: implement a qlib handler
     """
     
     def FetchAlpha158ResultFromDocker(self):
@@ -86,8 +86,9 @@ class QlibFactorRunner(TaskGenerator[QlibFactorExperiment]):
         Generate the experiment by processing and combining factor data,
         then passing the combined data to Docker for backtest results.
         """
-
-        SOTA_factor = self.process_factor_data(exp.based_experiments)
+        SOTA_factor = None
+        if exp.based_experiments.__len__() != 1:
+            SOTA_factor = self.process_factor_data(exp.based_experiments)
         
         if exp.based_experiments[-1].result is None:
             exp.based_experiments[-1].result = self.FetchAlpha158ResultFromDocker()
@@ -133,6 +134,7 @@ class QlibFactorRunner(TaskGenerator[QlibFactorExperiment]):
             result = pickle.load(f)
          """
         
+        # TODO: Implement the Docker run in the following way
         # Local run
         # Clean up any previous run artifacts by deleting the mlruns directory
         mlruns_path = DIRNAME_local / 'mlruns' / '1'
@@ -147,7 +149,7 @@ class QlibFactorRunner(TaskGenerator[QlibFactorExperiment]):
         qle = LocalEnv(conf=local_conf)
         qle.prepare()
         conf_path = str(DIRNAME / "env_factor" / "conf_combined.yaml") 
-        qle.run(entry="qrun " + conf_path)
+        qle.run(entry="qrun " + conf_path, local_path=PROP_SETTING.local_qlib_folder)
 
         # Verify if the new folder is created
         mlrun_p = DIRNAME_local / 'mlruns' / '1' 

rdagent/scenarios/qlib/task_generator/feedback.py

@@ -39,20 +39,15 @@ class QlibFactorHypothesisExperiment2Feedback(HypothesisExperiment2Feedback):
         hypothesis_text = hypothesis.hypothesis
         current_result = exp.result
         tasks_factors = [task.get_factor_information() for task in exp.sub_tasks]
-
         sota_result = exp.based_experiments[-1].result
 
         # Generate the system prompt
         sys_prompt = Environment(undefined=StrictUndefined).from_string(feedback_prompts["data_feedback_generation"]["system"]).render(scenario=self.scen.get_scenario_all_desc())
-        
-
-        # Prepare task details
-        task_details = "\n".join([f"Task: {factor_name}, Factor: {factor_description}" for factor_name, factor_description in tasks_factors])
 
         # Generate the user prompt
-        usr_prompt = Environment(undefined=StrictUndefined).from_string(feedback_prompts["data_feedback_generation"]["user"]).format(
+        usr_prompt = Environment(undefined=StrictUndefined).from_string(feedback_prompts["data_feedback_generation"]["user"]).render(
             hypothesis_text=hypothesis_text,
-            task_details=task_details,
+            task_details=tasks_factors,
             current_result=current_result,
             sota_result=sota_result
         )