Skip to content

Reflection

Reflector

Bases: BaseModel


              flowchart TD
              src.tradingagents.graph.reflection.Reflector[Reflector]

              

              click src.tradingagents.graph.reflection.Reflector href "" "src.tradingagents.graph.reflection.Reflector"

Handles reflection on decisions and updating memory.

Methods:

Name Description
reflect_bull_researcher

Reflect on bull researcher's analysis and update memory.
reflect_bear_researcher

Reflect on bear researcher's analysis and update memory.
reflect_trader

Reflect on trader's decision and update memory.
reflect_invest_judge

Reflect on investment judge's decision and update memory.
reflect_risk_manager

Reflect on risk manager's decision and update memory.

quick_thinking_llm 

quick_thinking_llm: SkipValidation[ChatModel] = Field(
    ...,
    title="Quick Thinking LLM",
    description="LLM instance used for generating reflection analysis",
)

reflect_bull_researcher 

reflect_bull_researcher(
    current_state: AgentState, returns_losses: float, bull_memory: FinancialSituationMemory
) -> None

Reflect on bull researcher's analysis and update memory.

Parameters:

Name Type Description Default

current_state

AgentState

The current state of the agent graph.

 required

returns_losses

float

The actual returns or losses observed.

 required

bull_memory

FinancialSituationMemory

The memory component to update.

 required
Source code in src/tradingagents/graph/reflection.py 
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
def reflect_bull_researcher(
    self,
    current_state: AgentState,
    returns_losses: float,
    bull_memory: FinancialSituationMemory,
) -> None:
    """Reflect on bull researcher's analysis and update memory.

    Args:
        current_state (AgentState): The current state of the agent graph.
        returns_losses (float): The actual returns or losses observed.
        bull_memory (FinancialSituationMemory): The memory component to update.
    """
    situation = self._extract_current_situation(current_state)
    result = self._reflect_on_component(
        "BULL", current_state.investment_debate_state.bull_history, situation, returns_losses
    )
    bull_memory.add_situations([(situation, result)])

reflect_bear_researcher 

reflect_bear_researcher(
    current_state: AgentState, returns_losses: float, bear_memory: FinancialSituationMemory
) -> None

Reflect on bear researcher's analysis and update memory.

Parameters:

Name Type Description Default

current_state

AgentState

The current state of the agent graph.

 required

returns_losses

float

The actual returns or losses observed.

 required

bear_memory

FinancialSituationMemory

The memory component to update.

 required
Source code in src/tradingagents/graph/reflection.py 
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
def reflect_bear_researcher(
    self,
    current_state: AgentState,
    returns_losses: float,
    bear_memory: FinancialSituationMemory,
) -> None:
    """Reflect on bear researcher's analysis and update memory.

    Args:
        current_state (AgentState): The current state of the agent graph.
        returns_losses (float): The actual returns or losses observed.
        bear_memory (FinancialSituationMemory): The memory component to update.
    """
    situation = self._extract_current_situation(current_state)
    result = self._reflect_on_component(
        "BEAR", current_state.investment_debate_state.bear_history, situation, returns_losses
    )
    bear_memory.add_situations([(situation, result)])

reflect_trader 

reflect_trader(
    current_state: AgentState, returns_losses: float, trader_memory: FinancialSituationMemory
) -> None

Reflect on trader's decision and update memory.

Parameters:

Name Type Description Default

current_state

AgentState

The current state of the agent graph.

 required

returns_losses

float

The actual returns or losses observed.

 required

trader_memory

FinancialSituationMemory

The memory component to update.

 required
Source code in src/tradingagents/graph/reflection.py 
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
def reflect_trader(
    self,
    current_state: AgentState,
    returns_losses: float,
    trader_memory: FinancialSituationMemory,
) -> None:
    """Reflect on trader's decision and update memory.

    Args:
        current_state (AgentState): The current state of the agent graph.
        returns_losses (float): The actual returns or losses observed.
        trader_memory (FinancialSituationMemory): The memory component to update.
    """
    situation = self._extract_current_situation(current_state)
    result = self._reflect_on_component(
        "TRADER", current_state.trader_investment_plan, situation, returns_losses
    )
    trader_memory.add_situations([(situation, result)])

reflect_invest_judge 

reflect_invest_judge(
    current_state: AgentState, returns_losses: float, invest_judge_memory: FinancialSituationMemory
) -> None

Reflect on investment judge's decision and update memory.

Parameters:

Name Type Description Default

current_state

AgentState

The current state of the agent graph.

 required

returns_losses

float

The actual returns or losses observed.

 required

invest_judge_memory

FinancialSituationMemory

The memory component to update.

 required
Source code in src/tradingagents/graph/reflection.py 
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
def reflect_invest_judge(
    self,
    current_state: AgentState,
    returns_losses: float,
    invest_judge_memory: FinancialSituationMemory,
) -> None:
    """Reflect on investment judge's decision and update memory.

    Args:
        current_state (AgentState): The current state of the agent graph.
        returns_losses (float): The actual returns or losses observed.
        invest_judge_memory (FinancialSituationMemory): The memory component to update.
    """
    situation = self._extract_current_situation(current_state)
    result = self._reflect_on_component(
        "INVEST JUDGE",
        current_state.investment_debate_state.judge_decision,
        situation,
        returns_losses,
    )
    invest_judge_memory.add_situations([(situation, result)])

reflect_risk_manager 

reflect_risk_manager(
    current_state: AgentState, returns_losses: float, risk_manager_memory: FinancialSituationMemory
) -> None

Reflect on risk manager's decision and update memory.

Parameters:

Name Type Description Default

current_state

AgentState

The current state of the agent graph.

 required

returns_losses

float

The actual returns or losses observed.

 required

risk_manager_memory

FinancialSituationMemory

The memory component to update.

 required
Source code in src/tradingagents/graph/reflection.py 
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
def reflect_risk_manager(
    self,
    current_state: AgentState,
    returns_losses: float,
    risk_manager_memory: FinancialSituationMemory,
) -> None:
    """Reflect on risk manager's decision and update memory.

    Args:
        current_state (AgentState): The current state of the agent graph.
        returns_losses (float): The actual returns or losses observed.
        risk_manager_memory (FinancialSituationMemory): The memory component to update.
    """
    situation = self._extract_current_situation(current_state)
    result = self._reflect_on_component(
        "RISK JUDGE", current_state.risk_debate_state.judge_decision, situation, returns_losses
    )
    risk_manager_memory.add_situations([(situation, result)])