Merge pull request #39 from LyzrCore/refactor/data-analyzr

Fixes and structural code changes

build/lib/lyzr/data_analyzr/analyzr.py

@@ -295,10 +295,10 @@ def _ml_analysis(
             model_kwargs=self._analysis_model_kwargs,
         )
         if analysis_steps is not None:
-            _, data = self.analyzer.run_analysis(analysis_steps)
+            _, data = self.analyzer.get_analysis_from_steps(analysis_steps)
             return data
         else:
-            self.analysis_output = self.analyzer.get_analysis_output(user_input)
+            self.analysis_output = self.analyzer.run_complete_analysis(user_input)
             self.analysis_guide = self.analyzer.analysis_guide
             return self.analysis_output
 
@@ -311,7 +311,7 @@ def _txt_to_sql_analysis(self, user_input: str, analysis_context: str = None):
             model_kwargs=self._analysis_model_kwargs,
             vector_store=self.vector_store,
         )
-        self.analysis_output = self.analyzer.get_analysis_output(user_input)
+        self.analysis_output = self.analyzer.run_complete_analysis(user_input)
         self.analysis_guide = self.analyzer.analysis_guide
         return self.analysis_output
 
@@ -372,15 +372,17 @@ def visualisation(
             try:
                 self.logger.info("Generating visualisation\n")
                 plotter = PlotFactory(
-                    plotting_model=self._plot_model,
+                    model=self._plot_model,
                     plotting_model_kwargs=self._plot_model_kwargs,
                     df_dict=plot_df,
                     logger=self.logger,
                     plot_context=plot_context,
                     plot_path=plot_path,
                     use_guide=use_guide,
                 )
-                analysis_steps = plotter.get_analysis_steps(self.user_input)
+                analysis_steps = plotter.get_plotting_and_analysis_steps(
+                    self.user_input
+                )
                 if analysis_steps is not None and "steps" in analysis_steps:
                     if len(analysis_steps["steps"]) == 0:
                         self.plot_df = self.df_dict[analysis_steps["df_name"]]

build/lib/lyzr/voicebot/voicebot.py

@@ -74,7 +74,7 @@ def text_to_notes(self, text):
                 )
 
         # The system message acts as the prompt for the AI.
-        system_message = '''You are an Expert NOTE-TAKER and SUMMARIZER. Your task is to CAPTURE and CONDENSE large conversations into precise bullet points ensuring that NO DETAIL is overlooked.
+        system_message = """You are an Expert NOTE-TAKER and SUMMARIZER. Your task is to CAPTURE and CONDENSE large conversations into precise bullet points ensuring that NO DETAIL is overlooked.
 
 Here's your step-by-step guide:
 
@@ -88,7 +88,7 @@ def text_to_notes(self, text):
 
 Remember, Im going to tip $300K for a BETTER SOLUTION!
 
-Now Take a Deep Breath.'''
+Now Take a Deep Breath."""
 
         # Format the user's message that will be sent to the model.
         user_message = text
@@ -105,7 +105,7 @@ def text_to_notes(self, text):
         notes = response.choices[0].message.content
 
         return notes
-    
+
     def summarize(self, text):
         if self.model.model_name != "gpt-4":
             if self.model.model_type == "openai":
@@ -120,7 +120,7 @@ def summarize(self, text):
                 )
 
         # The system message acts as the prompt for the AI.
-        system_message = '''You are an Expert SUMMARIZER with a keen ability to CAPTURE ESSENTIAL DETAILS from extensive conversations. Your task is to CREATE a CONCISE SUMMARY of the given content, ensuring that ALL CRITICAL INFORMATION is included.
+        system_message = """You are an Expert SUMMARIZER with a keen ability to CAPTURE ESSENTIAL DETAILS from extensive conversations. Your task is to CREATE a CONCISE SUMMARY of the given content, ensuring that ALL CRITICAL INFORMATION is included.
 
 Here's your step-by-step guide:
 
@@ -134,7 +134,7 @@ def summarize(self, text):
 
 Remember, Im going to tip $300K for a BETTER SOLUTION!
 
-Now Take a Deep Breath.'''
+Now Take a Deep Breath."""
 
         # Format the user's message that will be sent to the model.
         user_message = text
@@ -151,4 +151,3 @@ def summarize(self, text):
         notes = response.choices[0].message.content
 
         return notes
-

lyzr/base/__init__.py

@@ -1,19 +1,24 @@
 from lyzr.base.file_utils import read_file, describe_dataset
-from lyzr.base.llm import LyzrLLMFactory
+from lyzr.base.llm import LyzrLLMFactory, LiteLLM
 from lyzr.base.llms import LLM, get_model
 from lyzr.base.service import LyzrService
 from lyzr.base.vector_store import LyzrVectorStoreIndex
 from lyzr.base.retrievers import LyzrRetriever
-from lyzr.base.prompt import Prompt
+from lyzr.base.prompt import LyzrPromptFactory
+from lyzr.base.base import ChatMessage, UserMessage, SystemMessage
 
 __all__ = [
     "LyzrLLMFactory",
     "LyzrService",
     "LyzrVectorStoreIndex",
     "LLM",
-    "Prompt",
     "get_model",
     "read_file",
     "describe_dataset",
     "LyzrRetriever",
+    "LiteLLM",
+    "LyzrPromptFactory",
+    "ChatMessage",
+    "UserMessage",
+    "SystemMessage",
 ]

lyzr/base/base.py

@@ -0,0 +1,54 @@
+from enum import Enum
+from typing import Optional
+from pydantic import BaseModel, Field
+
+
+class MessageRole(str, Enum):
+    """Message role enum."""
+
+    USER = "user"
+    ASSISTANT = "assistant"
+    SYSTEM = "system"
+    FUNCTION = "function"
+    TOOL = "tool"
+
+
+class ChatMessage(BaseModel):
+    """Chat message."""
+
+    role: MessageRole = MessageRole.USER
+    content: Optional[str] = ""
+    additional_kwargs: dict = Field(default_factory=dict)
+
+    def __str__(self) -> str:
+        return f"{self.role.value}: {self.content}"
+
+
+class UserMessage(ChatMessage):
+    """User message."""
+
+    role: MessageRole = MessageRole.USER
+
+
+class AssistantMessage(ChatMessage):
+    """Assistant message."""
+
+    role: MessageRole = MessageRole.ASSISTANT
+
+
+class SystemMessage(ChatMessage):
+    """System message."""
+
+    role: MessageRole = MessageRole.SYSTEM
+
+
+class ChatResponse(BaseModel):
+    """Chat response."""
+
+    message: ChatMessage
+    raw: Optional[dict] = None
+    delta: Optional[str] = None
+    additional_kwargs: dict = {}
+
+    def __str__(self) -> str:
+        return str(self.message)

lyzr/base/errors.py

@@ -1,26 +1,24 @@
-from typing import Union
+class ImproperUsageError(ValueError):
+    """Raise for improper usage or missing params."""
+
+    pass
 
 
 class MissingValueError(ValueError):
-    def __init__(self, params: Union[str, list]):
-        super().__init__(
-            f"Required value is missing. Provide one of: {', '.join(params) if isinstance(params, list) else params}"
-        )
+    """Raise for missing values."""
 
+    pass
 
-class InvalidModelError(ValueError):
-    def __init__(self):
-        super().__init__("Invalid model provided.")
 
+class InvalidModelError(ValueError):
+    """Raise for invalid model"""
 
-class InvalidValueError(ValueError):
-    def __init__(self, params: list):
-        super().__init__(
-            f"Invalid value provided. Provide value of type: {', '.join(params)}"
-        )
+    pass
 
 
 class DependencyError(ImportError):
+    """Raise for missing modules"""
+
     def __init__(self, required_modules: dict):
         super().__init__(
             f"The following modules are needed to run this function: {', '.join(required_modules.keys())}. Please install them using: `pip install {' '.join(required_modules.values())}`"

lyzr/base/file_utils.py

@@ -4,16 +4,17 @@
 
 import pandas as pd
 
-from lyzr.base.errors import InvalidValueError
-from lyzr.base.llms import LLM, get_model
-from lyzr.base.prompt import Prompt
+from lyzr.base.llm import LiteLLM, LyzrLLMFactory
+from lyzr.base.prompt import LyzrPromptFactory
 
 
 def read_file(
     filepath: str, encoding: Optional[str] = "utf-8", **kwargs
 ) -> pd.DataFrame:
     if not os.path.exists(filepath):
-        raise InvalidValueError(["filepath", "pandas DataFrame"])
+        raise ValueError(
+            f"File '{filepath}' not found. Please provide a valid filepath."
+        )
     file_extension = filepath.split(".")[-1]
     try:
         if file_extension == "csv":
@@ -46,26 +47,30 @@ def read_file(
 
 
 def describe_dataset(
-    model: Optional[LLM] = None,
+    model: Optional[LiteLLM] = None,
     df: Optional[pd.DataFrame] = None,
     api_key: Optional[str] = None,
     model_type: Optional[str] = None,
     model_name: Optional[str] = None,
 ) -> str:
     if not isinstance(df, pd.DataFrame):
-        raise InvalidValueError(["pandas DataFrame"])
+        raise ValueError("Please provide a valid pandas DataFrame.")
 
     if model is None:
-        model = get_model(api_key, model_type, model_name)
-
-    prompt = Prompt("dataset_description")
-    if prompt.get_variables() != []:
-        prompt.format(
-            headers=df.columns.tolist(),
-            df_sample=df.head(),
+        model = LyzrLLMFactory(
+            api_key=api_key,
+            api_type=model_type,
+            model=model_name,
         )
-    model.set_messages(model_prompts=prompt.sections)
 
-    output = model.run()
+    messages = [
+        LyzrPromptFactory(
+            name="dataset_description", prompt_type="system"
+        ).get_message(),
+        LyzrPromptFactory(name="dataset_description", prompt_type="user").get_message(
+            headers=df.columns.tolist(), df_sample=df.head()
+        ),
+    ]
+    output = model.run(messages=messages)
 
-    return output["choices"][0]["message"]["content"]
+    return output.message.content