ai-data-science-team 0.0.0.9010__py3-none-any.whl → 0.0.0.9012__py3-none-any.whl

ai_data_science_team/_version.py

@@ -1 +1 @@
-__version__ = "0.0.0.9010"
+__version__ = "0.0.0.9012"

ai_data_science_team/agents/__init__.py

@@ -3,3 +3,4 @@ from ai_data_science_team.agents.feature_engineering_agent import make_feature_e
 from ai_data_science_team.agents.data_wrangling_agent import make_data_wrangling_agent, DataWranglingAgent
 from ai_data_science_team.agents.sql_database_agent import make_sql_database_agent, SQLDatabaseAgent
 from ai_data_science_team.agents.data_visualization_agent import make_data_visualization_agent, DataVisualizationAgent
+from ai_data_science_team.agents.data_loader_tools_agent import make_data_loader_tools_agent, DataLoaderToolsAgent

ai_data_science_team/agents/data_loader_tools_agent.py

@@ -37,11 +37,150 @@ tools = [
     search_files_by_pattern,
 ]
 
+class DataLoaderToolsAgent(BaseAgent):
+    """
+    A Data Loader Agent that can interact with data loading tools and search for files in your file system.
+    
+    Parameters:
+    ----------
+    model : langchain.llms.base.LLM
+        The language model used to generate the tool calling agent.
+    react_agent_kwargs : dict
+        Additional keyword arguments to pass to the create_react_agent function.
+    invoke_react_agent_kwargs : dict
+        Additional keyword arguments to pass to the invoke method of the react agent.
+        
+    Methods:
+    --------
+    update_params(**kwargs)
+        Updates the agent's parameters and rebuilds the compiled graph.
+    ainvoke_agent(user_instructions: str=None, **kwargs)
+        Runs the agent with the given user instructions asynchronously.
+    invoke_agent(user_instructions: str=None, **kwargs)
+        Runs the agent with the given user instructions.
+    get_internal_messages(markdown: bool=False)
+        Returns the internal messages from the agent's response.
+    get_artifacts(as_dataframe: bool=False)
+        Returns the MLflow artifacts from the agent's response.
+    get_ai_message(markdown: bool=False)
+        Returns the AI message from the agent's response.
+    
+    """
+    
+    def __init__(
+        self, 
+        model: Any,
+        create_react_agent_kwargs: Optional[Dict]={},
+        invoke_react_agent_kwargs: Optional[Dict]={},
+    ):
+        self._params = {
+            "model": model,
+            "create_react_agent_kwargs": create_react_agent_kwargs,
+            "invoke_react_agent_kwargs": invoke_react_agent_kwargs,
+        }
+        self._compiled_graph = self._make_compiled_graph()
+        self.response = None
+        
+    def _make_compiled_graph(self):
+        """
+        Creates the compiled graph for the agent.
+        """
+        self.response = None
+        return make_data_loader_tools_agent(**self._params)
+    
+    
+    def update_params(self, **kwargs):
+        """
+        Updates the agent's parameters and rebuilds the compiled graph.
+        """
+        for k, v in kwargs.items():
+            self._params[k] = v
+        self._compiled_graph = self._make_compiled_graph()
+        
+    async def ainvoke_agent(
+        self, 
+        user_instructions: str=None, 
+        **kwargs
+    ):
+        """
+        Runs the agent with the given user instructions.
+        
+        Parameters:
+        ----------
+        user_instructions : str, optional
+            The user instructions to pass to the agent.
+        kwargs : dict, optional
+            Additional keyword arguments to pass to the agents ainvoke method.
+        
+        """
+        response = await self._compiled_graph.ainvoke(
+            {
+                "user_instructions": user_instructions,
+            }, 
+            **kwargs
+        )
+        self.response = response
+        return None
+    
+    def invoke_agent(
+        self, 
+        user_instructions: str=None, 
+        **kwargs
+    ):
+        """
+        Runs the agent with the given user instructions.
+        
+        Parameters:
+        ----------
+        user_instructions : str, optional
+            The user instructions to pass to the agent.
+        kwargs : dict, optional
+            Additional keyword arguments to pass to the agents invoke method.
+        
+        """
+        response = self._compiled_graph.invoke(
+            {
+                "user_instructions": user_instructions,
+            },
+            **kwargs
+        )
+        self.response = response
+        return None
+    
+    def get_internal_messages(self, markdown: bool=False):
+        """
+        Returns the internal messages from the agent's response.
+        """
+        pretty_print = "\n\n".join([f"### {msg.type.upper()}\n\nID: {msg.id}\n\nContent:\n\n{msg.content}" for msg in self.response["internal_messages"]])       
+        if markdown:
+            return Markdown(pretty_print)
+        else:
+            return self.response["internal_messages"]
+    
+    def get_artifacts(self, as_dataframe: bool=False):
+        """
+        Returns the MLflow artifacts from the agent's response.
+        """
+        if as_dataframe:
+            return pd.DataFrame(self.response["data_loader_artifacts"])
+        else:
+            return self.response["data_loader_artifacts"]
+    
+    def get_ai_message(self, markdown: bool=False):
+        """
+        Returns the AI message from the agent's response.
+        """
+        if markdown:
+            return Markdown(self.response["messages"][0].content)
+        else:
+            return self.response["messages"][0].content
 
+    
 
 def make_data_loader_tools_agent(
     model: Any,
-    directory: Optional[str] = os.getcwd(),
+    create_react_agent_kwargs: Optional[Dict]={},
+    invoke_react_agent_kwargs: Optional[Dict]={},
 ):
     """
     Creates a Data Loader Agent that can interact with data loading tools.
@@ -50,20 +189,84 @@ def make_data_loader_tools_agent(
     ----------
     model : langchain.llms.base.LLM
         The language model used to generate the tool calling agent.
-    directory : str, optional
-        The directory to search for files. Defaults to the current working directory.
+    react_agent_kwargs : dict
+        Additional keyword arguments to pass to the create_react_agent function.
+    invoke_react_agent_kwargs : dict
+        Additional keyword arguments to pass to the invoke method of the react agent.
     
     Returns:
     --------
-    Data Loader Agent
+    app : langchain.graphs.CompiledStateGraph
         An agent that can interact with data loading tools.
     """
     
     class GraphState(AgentState):
         internal_messages: Annotated[Sequence[BaseMessage], operator.add]
-        directory: str
         user_instructions: str
-        data_artifacts: dict
+        data_loader_artifacts: dict
+        
+    def data_loader_agent(state):
+        
+        print(format_agent_name(AGENT_NAME))
+        print("    ")
+        
+        print("    * RUN REACT TOOL-CALLING AGENT")
+        
+        tool_node = ToolNode(
+            tools=tools
+        )
+        
+        data_loader_agent = create_react_agent(
+            model, 
+            tools=tool_node, 
+            state_schema=GraphState,
+            **create_react_agent_kwargs,
+        )
+        
+        response = data_loader_agent.invoke(
+            {
+                "messages": [("user", state["user_instructions"])],
+            },
+            invoke_react_agent_kwargs,
+        )
+        
+        print("    * POST-PROCESS RESULTS")
+        
+        internal_messages = response['messages']
+
+        # Ensure there is at least one AI message
+        if not internal_messages:
+            return {
+                "internal_messages": [],
+                "mlflow_artifacts": None,
+            }
+
+        # Get the last AI message
+        last_ai_message = AIMessage(internal_messages[-1].content, role = AGENT_NAME)
+
+        # Get the last tool artifact safely
+        last_tool_artifact = None
+        if len(internal_messages) > 1:
+            last_message = internal_messages[-2]  # Get second-to-last message
+            if hasattr(last_message, "artifact"):  # Check if it has an "artifact"
+                last_tool_artifact = last_message.artifact
+            elif isinstance(last_message, dict) and "artifact" in last_message:
+                last_tool_artifact = last_message["artifact"]
+
+        return {
+            "messages": [last_ai_message], 
+            "internal_messages": internal_messages,
+            "data_loader_artifacts": last_tool_artifact,
+        }
+        
+    workflow = StateGraph(GraphState)
     
-    pass
+    workflow.add_node("data_loader_agent", data_loader_agent)
+    
+    workflow.add_edge(START, "data_loader_agent")
+    workflow.add_edge("data_loader_agent", END)
+    
+    app = workflow.compile()
+
+    return app
 

ai_data_science_team/ds_agents/__init__.py

@@ -0,0 +1 @@
+from ai_data_science_team.ds_agents.eda_tools_agent import EDAToolsAgent, make_eda_tools_agent

ai_data_science_team/ds_agents/eda_tools_agent.py

@@ -0,0 +1,245 @@
+
+
+from typing import Any, Optional, Annotated, Sequence, List, Dict, Tuple
+import operator
+import pandas as pd
+import os
+from io import StringIO, BytesIO
+import base64
+import matplotlib.pyplot as plt
+
+from IPython.display import Markdown
+
+from langchain_core.messages import BaseMessage, AIMessage
+from langgraph.prebuilt import create_react_agent, ToolNode
+from langgraph.prebuilt.chat_agent_executor import AgentState
+from langgraph.graph import START, END, StateGraph
+
+from ai_data_science_team.templates import BaseAgent
+from ai_data_science_team.utils.regex import format_agent_name
+
+from ai_data_science_team.tools.eda import (
+    describe_dataset, 
+    visualize_missing, 
+    correlation_funnel,
+    generate_sweetviz_report,
+)
+
+
+AGENT_NAME = "exploratory_data_analyst_agent"
+
+# Updated tool list for EDA
+EDA_TOOLS = [
+    describe_dataset,
+    visualize_missing,
+    correlation_funnel,
+    generate_sweetviz_report,
+]
+
+class EDAToolsAgent(BaseAgent):
+    """
+    An Exploratory Data Analysis Tools Agent that interacts with EDA tools to generate summary statistics,
+    missing data visualizations, correlation funnels, EDA reports, etc.
+    
+    Parameters:
+    ----------
+    model : langchain.llms.base.LLM
+        The language model for generating the tool-calling agent.
+    create_react_agent_kwargs : dict
+        Additional kwargs for create_react_agent.
+    invoke_react_agent_kwargs : dict
+        Additional kwargs for agent invocation.
+    """
+    
+    def __init__(
+        self, 
+        model: Any,
+        create_react_agent_kwargs: Optional[Dict] = {},
+        invoke_react_agent_kwargs: Optional[Dict] = {},
+    ):
+        self._params = {
+            "model": model,
+            "create_react_agent_kwargs": create_react_agent_kwargs,
+            "invoke_react_agent_kwargs": invoke_react_agent_kwargs,
+        }
+        self._compiled_graph = self._make_compiled_graph()
+        self.response = None
+        
+    def _make_compiled_graph(self):
+        """
+        Creates the compiled state graph for the EDA agent.
+        """
+        self.response = None
+        return make_eda_tools_agent(**self._params)
+    
+    def update_params(self, **kwargs):
+        """
+        Updates the agent's parameters and rebuilds the compiled graph.
+        """
+        for k, v in kwargs.items():
+            self._params[k] = v
+        self._compiled_graph = self._make_compiled_graph()
+        
+    async def ainvoke_agent(
+        self, 
+        user_instructions: str = None, 
+        data_raw: pd.DataFrame = None, 
+        **kwargs
+    ):
+        """
+        Asynchronously runs the agent with user instructions and data.
+        
+        Parameters:
+        ----------
+        user_instructions : str, optional
+            The instructions for the agent.
+        data_raw : pd.DataFrame, optional
+            The input data as a DataFrame.
+        """
+        response = await self._compiled_graph.ainvoke(
+            {
+                "user_instructions": user_instructions,
+                "data_raw": data_raw.to_dict() if data_raw is not None else None,
+            },
+            **kwargs
+        )
+        self.response = response
+        return None
+    
+    def invoke_agent(
+        self, 
+        user_instructions: str = None, 
+        data_raw: pd.DataFrame = None, 
+        **kwargs
+    ):
+        """
+        Synchronously runs the agent with user instructions and data.
+        
+        Parameters:
+        ----------
+        user_instructions : str, optional
+            The instructions for the agent.
+        data_raw : pd.DataFrame, optional
+            The input data as a DataFrame.
+        """
+        response = self._compiled_graph.invoke(
+            {
+                "user_instructions": user_instructions,
+                "data_raw": data_raw.to_dict() if data_raw is not None else None,
+            },
+            **kwargs
+        )
+        self.response = response
+        return None
+    
+    def get_internal_messages(self, markdown: bool = False):
+        """
+        Returns internal messages from the agent response.
+        """
+        pretty_print = "\n\n".join(
+            [f"### {msg.type.upper()}\n\nID: {msg.id}\n\nContent:\n\n{msg.content}" 
+             for msg in self.response["internal_messages"]]
+        )
+        if markdown:
+            return Markdown(pretty_print)
+        else:
+            return self.response["internal_messages"]
+    
+    def get_artifacts(self, as_dataframe: bool = False):
+        """
+        Returns the EDA artifacts from the agent response.
+        """
+        if as_dataframe:
+            return pd.DataFrame(self.response["eda_artifacts"])
+        else:
+            return self.response["eda_artifacts"]
+    
+    def get_ai_message(self, markdown: bool = False):
+        """
+        Returns the AI message from the agent response.
+        """
+        if markdown:
+            return Markdown(self.response["messages"][0].content)
+        else:
+            return self.response["messages"][0].content
+
+def make_eda_tools_agent(
+    model: Any,
+    create_react_agent_kwargs: Optional[Dict] = {},
+    invoke_react_agent_kwargs: Optional[Dict] = {},
+):
+    """
+    Creates an Exploratory Data Analyst Agent that can interact with EDA tools.
+    
+    Parameters:
+    ----------
+    model : Any
+        The language model used for tool-calling.
+    create_react_agent_kwargs : dict
+        Additional kwargs for create_react_agent.
+    invoke_react_agent_kwargs : dict
+        Additional kwargs for agent invocation.
+    
+    Returns:
+    -------
+    app : langgraph.graph.CompiledStateGraph
+        The compiled state graph for the EDA agent.
+    """
+    
+    class GraphState(AgentState):
+        internal_messages: Annotated[Sequence[BaseMessage], operator.add]
+        user_instructions: str
+        data_raw: dict
+        eda_artifacts: dict
+
+    def exploratory_agent(state):
+        print(format_agent_name(AGENT_NAME))
+        print("    * RUN REACT TOOL-CALLING AGENT FOR EDA")
+        
+        tool_node = ToolNode(
+            tools=EDA_TOOLS
+        )
+        
+        eda_agent = create_react_agent(
+            model,
+            tools=tool_node,
+            state_schema=GraphState,
+            **create_react_agent_kwargs,
+        )
+        
+        response = eda_agent.invoke(
+            {
+                "messages": [("user", state["user_instructions"])],
+                "data_raw": state["data_raw"],
+            },
+            invoke_react_agent_kwargs,
+        )
+        
+        print("    * POST-PROCESSING EDA RESULTS")
+        
+        internal_messages = response['messages']
+        if not internal_messages:
+            return {"internal_messages": [], "eda_artifacts": None}
+        
+        last_ai_message = AIMessage(internal_messages[-1].content, role=AGENT_NAME)
+        last_tool_artifact = None
+        if len(internal_messages) > 1:
+            last_message = internal_messages[-2]
+            if hasattr(last_message, "artifact"):
+                last_tool_artifact = last_message.artifact
+            elif isinstance(last_message, dict) and "artifact" in last_message:
+                last_tool_artifact = last_message["artifact"]
+        
+        return {
+            "messages": [last_ai_message],
+            "internal_messages": internal_messages,
+            "eda_artifacts": last_tool_artifact,
+        }
+    
+    workflow = StateGraph(GraphState)
+    workflow.add_node("exploratory_agent", exploratory_agent)
+    workflow.add_edge(START, "exploratory_agent")
+    workflow.add_edge("exploratory_agent", END)
+    
+    app = workflow.compile()
+    return app

ai_data_science_team/ml_agents/h2o_ml_agent.py

@@ -506,6 +506,7 @@ def make_h2o_ml_agent(
             while remaining flexible to user instructions.
             - Return a dict with keys: leaderboard, best_model_id, model_path, and model_results.
             - If enable_mlfow is True, log the top metrics and save the model as an artifact. (See example function)
+            - IMPORTANT: if enable_mlflow is True, make sure to set enable_mlflow to True in the function definition.
             
             Initial User Instructions (Disregard any instructions that are unrelated to modeling):
                 {user_instructions}
@@ -533,7 +534,7 @@ def make_h2o_ml_agent(
                 sort_metric: str ,
                 model_directory: Optional[str] = None,
                 log_path: Optional[str] = None,
-                enable_mlflow: bool,               
+                enable_mlflow: bool, # If use has specified to enable MLflow, make sure to make this True              
                 mlflow_tracking_uri: Optional[str], 
                 mlflow_experiment_name: str,
                 mlflow_run_name: str,

ai_data_science_team/ml_agents/mlflow_tools_agent.py

@@ -1,5 +1,5 @@
 
-from typing import Any, Optional, Annotated, Sequence
+from typing import Any, Optional, Annotated, Sequence, Dict
 import operator
 
 import pandas as pd
@@ -63,8 +63,10 @@ class MLflowToolsAgent(BaseAgent):
         The tracking URI for MLflow. Defaults to None.
     mlflow_registry_uri : str, optional
         The registry URI for MLflow. Defaults to None.
-    **react_agent_kwargs : dict, optional
-        Additional keyword arguments to pass to the agent's react agent.
+    react_agent_kwargs : dict
+        Additional keyword arguments to pass to the create_react_agent function.
+    invoke_react_agent_kwargs : dict
+        Additional keyword arguments to pass to the invoke method of the react agent.
     
     Methods:
     --------
@@ -114,13 +116,15 @@ class MLflowToolsAgent(BaseAgent):
         model: Any,
         mlflow_tracking_uri: Optional[str]=None,
         mlflow_registry_uri: Optional[str]=None,
-        **react_agent_kwargs,
+        create_react_agent_kwargs: Optional[Dict]={},
+        invoke_react_agent_kwargs: Optional[Dict]={},
     ):
         self._params = {
             "model": model,
             "mlflow_tracking_uri": mlflow_tracking_uri,
             "mlflow_registry_uri": mlflow_registry_uri,
-            **react_agent_kwargs,
+            "create_react_agent_kwargs": create_react_agent_kwargs,
+            "invoke_react_agent_kwargs": invoke_react_agent_kwargs,
         }
         self._compiled_graph = self._make_compiled_graph()
         self.response = None
@@ -185,8 +189,6 @@ class MLflowToolsAgent(BaseAgent):
             The user instructions to pass to the agent.
         data_raw : pd.DataFrame, optional
             The raw data to pass to the agent. Used for prediction and tool calls where data is required.
-        kwargs : dict, optional
-            Additional keyword arguments to pass to the agents invoke method.
         
         """
         response = self._compiled_graph.invoke(
@@ -234,10 +236,30 @@ def make_mlflow_tools_agent(
     model: Any,
     mlflow_tracking_uri: str=None,
     mlflow_registry_uri: str=None,
-    **react_agent_kwargs,
+    create_react_agent_kwargs: Optional[Dict]={},
+    invoke_react_agent_kwargs: Optional[Dict]={},
 ):
     """
     MLflow Tool Calling Agent
+    
+    Parameters:
+    ----------
+    model : Any
+        The language model used to generate the agent.
+    mlflow_tracking_uri : str, optional
+        The tracking URI for MLflow. Defaults to None.
+    mlflow_registry_uri : str, optional
+        The registry URI for MLflow. Defaults to None.
+    create_react_agent_kwargs : dict, optional
+        Additional keyword arguments to pass to the agent's create_react_agent method.
+    invoke_react_agent_kwargs : dict, optional
+        Additional keyword arguments to pass to the agent's invoke method.
+        
+    Returns
+    -------
+    app : langchain.graphs.CompiledStateGraph
+        A compiled state graph for the MLflow Tool Calling Agent.
+    
     """
     
     try:
@@ -274,7 +296,7 @@ def make_mlflow_tools_agent(
             model, 
             tools=tool_node, 
             state_schema=GraphState,
-            **react_agent_kwargs,
+            **create_react_agent_kwargs,
         )
         
         response = mlflow_agent.invoke(
@@ -282,6 +304,7 @@ def make_mlflow_tools_agent(
                 "messages": [("user", state["user_instructions"])],
                 "data_raw": state["data_raw"],
             },
+            invoke_react_agent_kwargs,
         )
         
         print("    * POST-PROCESS RESULTS")