Upload 7 files

app.py

@@ -0,0 +1,72 @@
+import pandas as pd
+import json
+from document_preprocessor import generate_document
+from llm import LLM
+from prompt import stock_analysis_prompt
+import streamlit as st
+from streamlit_searchbox import st_searchbox
+
+st.set_page_config(
+    page_title="Stock_Picker",
+    page_icon="💰",
+    layout="wide",
+    initial_sidebar_state="expanded",
+)
+
+st.markdown("### 📈 Stock Picker")
+
+left_co, cent_co,last_co = st.columns(3)
+with cent_co:
+    st.image(image=".streamlit/stock-market.png", width=300)
+
+st.markdown("---")
+
+
+stocks = pd.read_excel("MCAP31122023.xlsx").set_index('Company Name')
+
+url = "https://ticker.finology.in/company/"
+
+model = LLM(model_name="Gemini")
+
+# function with list of labels
+def search_stocks(searchterm: str):
+    if not searchterm:
+        return []
+    matching_stocks = stocks[stocks.index.str.contains(searchterm, case=False, na=False)]
+    return matching_stocks['Symbol'].tolist()
+
+
+selected_value = st_searchbox(
+    search_stocks,
+    key="wiki_searchbox",
+)
+
+if selected_value:
+    stock_url = f"https://ticker.finology.in/company/{selected_value}"
+    stock_fundamentals = generate_document(stock_url)
+    prompt = stock_analysis_prompt.replace(
+        "{stock_name}",selected_value).replace("{context}",stock_fundamentals.page_content)
+    result = model(prompt=prompt).replace('```',"")
+    
+    try:
+        res = json.loads(result)
+        confidence_score = res['buy']
+        analysis = res["detailed_analysis"]
+
+        if confidence_score >= 75:
+            st.success("High Confidence Score!")
+        elif confidence_score > 40:
+            st.warning("Moderate Confidence Score.")
+        else:
+            st.error("Low Confidence Score.")
+
+
+        col1, col2 = st.columns(2)
+        col1.write(f'**Buy Confidence Score:** {str(confidence_score)}')
+
+        with st.expander("See explanation"):
+            st.write(f"**Detailed Analysis:** {analysis}")
+        st.markdown(f"[Learn more about {selected_value}]({stock_url})")
+
+    except:
+        st.write(result)

document_preprocessor.py

@@ -0,0 +1,13 @@
+from langchain_community.document_loaders import UnstructuredURLLoader
+from langchain.docstore.document import Document
+from unstructured.cleaners.core import remove_punctuation,clean,clean_extra_whitespace
+
+def generate_document(url):
+    "Given an URL, return a langchain Document to futher processing"
+    loader = UnstructuredURLLoader(urls=[url],
+    mode="elements",
+    post_processors=[clean,remove_punctuation,clean_extra_whitespace])
+    elements = loader.load()
+    selected_elements = [e for e in elements]
+    full_clean = " ".join([e.page_content for e in selected_elements])
+    return Document(page_content=full_clean, metadata={"source":url})

gitignore

@@ -0,0 +1,5 @@
+venv/
+*.ipynb
+*.csv
+__pycache__
+*.json

llm.py

@@ -0,0 +1,54 @@
+import os
+import google.generativeai as genai
+from dotenv import load_dotenv
+
+load_dotenv()
+
+generation_config = {
+  "temperature": 0,
+  "top_k": 1,
+  "max_output_tokens": 4000,
+}
+
+
+class LLM:
+    def __init__(self, model_name) -> None:
+        self.model_name = model_name
+        self.model = self.create_model(model_name)
+
+    def create_model(self, model_name):
+        match model_name:
+            case "GeminiVision":
+                genai.configure(api_key=os.environ.get("GOOGLE_API_KEY"))
+                return genai.GenerativeModel('gemini-pro-vision')
+            case "Gemini":
+                genai.configure(api_key=os.environ.get("GOOGLE_API_KEY"))
+                return genai.GenerativeModel(
+                    'gemini-pro',generation_config=generation_config)
+  
+            case _:
+                print("Not Implemented")
+          
+    def __call__(self, prompt, image=None):
+        if self.model_name == 'GeminiVision':
+            response = self.model.generate_content(
+                [image, prompt]
+            )
+        elif self.model_name == "Gemini":
+            response = self.model.generate_content(
+                prompt)
+            # print(response.text)
+            return response.text
+        
+        elif self.model_name == 'openai':
+            res = self.model.chat.completions.create(
+                model="gpt-3.5-turbo-1106",
+                response_format={"type": "json_object"},
+                messages=[
+                    # {"role": "system", "content": "You are a helpful assistant."},
+                    {"role": "user", "content": f"{prompt}"},
+                ],
+                # seed=10,
+                temperature=0
+            )
+            return res.choices[0].message.content

prompt.py

@@ -0,0 +1,10 @@
+stock_analysis_prompt = """
+You are a financial advisor which help ascertain whether the fundamentals of the
+stock is good enough for me to buy
+Stock name : {stock_name}
+The stock details is given below:
+{context}
+You should carry weight to latest news from the context for analysis
+Your output should be in JSON format with keys buy whose values should be a confidence score between 0-100
+and detailed analysis of your decision to buy or not
+"""

requirements.txt

@@ -0,0 +1,9 @@
+streamlit==1.30.0
+streamlit-searchbox==0.1.6
+google-generativeai==0.3.2
+langchain==0.1.4
+langchain-community==0.0.16
+# unstructured[docx,pptx]
+unstructured[all-docs]
+python-dotenv==1.0.1
+openpyxl==3.1.2