app.py

import streamlit as st
import pandas as pd
import requests
import plotly.express as px
from datetime import datetime, timedelta
import textwrap
import os

API_KEY = os.getenv("FMP_API_KEY")

# Set wide page layout
st.set_page_config(page_title="Analyst Recommendations", layout="wide")

# Sidebar: Global ticker and page navigation
st.sidebar.header("Inputs")
with st.sidebar.expander("Ticker and Page Settings", expanded=True):
    ticker = st.text_input(
        "Enter Ticker Symbol",
        value="AAPL",
        help="Input the stock ticker symbol (e.g., AAPL, MSFT)."
    )
    page = st.radio(
        "Select Page",
        ["Historical Ratings", "Recommendations"],
        help="Choose which analysis page to view."
    )

# Reset stored results if ticker changes
if "ticker" not in st.session_state or st.session_state.ticker != ticker:
    st.session_state.ticker = ticker
    st.session_state.historical_data = None
    st.session_state.analyst_data = None
    st.session_state.run_pressed_hist = False
    st.session_state.run_pressed_analyst = False

# Cached function for historical data
@st.cache_data(show_spinner=False)
def load_historical_data(ticker):
    try:
        url = f"https://financialmodelingprep.com/api/v3/historical-rating/{ticker}?apikey={API_KEY}"
        response = requests.get(url)
        if response.status_code != 200:
            st.error("Error retrieving historical data.")
            return None
        data = response.json()
        df = pd.DataFrame(data)
        # Define required columns.
        req_cols = [
            'date', 'rating', 'ratingScore', 'ratingRecommendation', 'ratingDetailsDCFScore',
            'ratingDetailsROEScore', 'ratingDetailsROERecommendation',
            'ratingDetailsROAScore', 'ratingDetailsROARecommendation',
            'ratingDetailsDEScore', 'ratingDetailsDERecommendation',
            'ratingDetailsPEScore', 'ratingDetailsPERecommendation',
            'ratingDetailsPBScore', 'ratingDetailsPBRecommendation'
        ]
        # Use alternative column if available.
        if 'ratingDetailsROCFRecommendation' in df.columns:
            req_cols.insert(4, 'ratingDetailsROCFRecommendation')
        else:
            req_cols.insert(4, 'ratingDetailsDCFRecommendation')
        df = df[req_cols]
        df['date'] = pd.to_datetime(df['date'])
        df = df.sort_values('date')
        return df
    except Exception:
        st.error("An error occurred while loading historical data.")
        return None

# Cached function for analyst recommendations data
@st.cache_data(show_spinner=False)
def load_analyst_data(ticker):
    try:
        url = f"https://financialmodelingprep.com/api/v3/analyst-stock-recommendations/{ticker}?apikey={API_KEY}"
        response = requests.get(url)
        if response.status_code != 200:
            st.error("Error retrieving analyst data.")
            return None
        data = response.json()
        if isinstance(data, dict):
            data = [data]
        df = pd.DataFrame(data)
        df['date'] = pd.to_datetime(df['date'])
        df = df.sort_values('date')
        return df
    except Exception:
        st.error("An error occurred while loading analyst data.")
        return None

# Main area explanation
#st.title("Analysts Recommendations")
#st.write("This app displays historical rating scores and analyst recommendations for the specified ticker.")


# PAGE: Historical Ratings
if page == "Historical Ratings":
    st.header("Historical Ratings")
    # Sidebar inputs for this page
    st.write("Below are a series of bar charts that show historical trends in key financial metrics.")
    with st.expander("Category Description", expanded=False):
        description = textwrap.dedent("""
            - **Overall Rating Score**: Reflects the general analyst rating, summarizing overall performance.
            - **DCF Score**: Represents the discounted cash flow valuation, which estimates a stock's intrinsic value.
            - **ROE Score**: Measures return on equity to assess how efficiently a company uses shareholder funds.
            - **ROA Score**: Indicates return on assets to gauge the effectiveness of asset use in generating profits.
            - **DE Score**: Shows the debt-to-equity ratio, highlighting the financial leverage of the company.
            - **PE Score**: Provides the price-to-earnings ratio, indicating market valuation relative to earnings.
            - **PB Score**: Measures the price-to-book ratio to assess if the stock is undervalued compared to its book value.
            Each chart displays the numerical score with a text recommendation. Colors denote recommendations like "Strong Buy", "Buy", "Neutral", "Sell", and "Strong Sell".
        """)
        st.markdown(description)
    default_date = datetime.today() - timedelta(days=365)
    with st.sidebar.expander("Date Settings", expanded=True):
        start_date = st.date_input(
            "Start Date",
            value=default_date,
            help="Select a start date for filtering historical data."
        )
    # Place run button below the start date input
    if st.sidebar.button("Run Analysis", key="hist_run_button"):
        st.session_state.run_pressed_hist = True

    if st.session_state.run_pressed_hist:
        # Load data if not already loaded
        if st.session_state.historical_data is None:
            st.session_state.historical_data = load_historical_data(ticker)
        df_hist = st.session_state.historical_data
        if df_hist is not None:
            # Filter data based on start date.
            df_filtered = df_hist[df_hist['date'] >= pd.to_datetime(start_date)]
            #st.subheader(f"Historical Ratings for {ticker}")
            #st.write("Bar charts below show various score metrics over time.")
            recommendation_colors = {
                "Strong Buy": "green",
                "Buy": "lightgreen",
                "Neutral": "orange",
                "Sell": "lightcoral",
                "Strong Sell": "red"
            }
            categories = [
                'ratingScore', 'ratingDetailsDCFScore', 'ratingDetailsROEScore',
                'ratingDetailsROAScore', 'ratingDetailsDEScore', 'ratingDetailsPEScore',
                'ratingDetailsPBScore'
            ]
            titles = [
                'Overall Rating Score', 'DCF Score', 'ROE Score',
                'ROA Score', 'DE Score', 'PE Score', 'PB Score'
            ]
            for category, title in zip(categories, titles):
                recommendation_col = category.replace("Score", "Recommendation")
                st.markdown(f"**{title} Chart**")
                #st.write("This chart shows the score over time with its associated recommendation.")
                try:
                    fig = px.bar(
                        df_filtered,
                        x='date',
                        y=category,
                        text=category,
                        labels={'date': 'Date', category: 'Score'},
                        title=title,
                        color=recommendation_col,
                        color_discrete_map=recommendation_colors,
                        custom_data=['rating', recommendation_col]
                    )
                    fig.update_traces(
                        texttemplate="%{text}<br>%{customdata[0]} (%{customdata[1]})",
                        textposition='outside',
                        hovertemplate="<b>Date</b>: %{x}<br><b>Score</b>: %{y}<br><b>Rating</b>: %{customdata[0]}<br><b>Recommendation</b>: %{customdata[1]}<extra></extra>"
                    )
                    st.plotly_chart(fig, use_container_width=True)
                except Exception:
                    st.error("Error displaying the chart. Please check the data and inputs.")
            with st.expander("Show Detailed Historical Data"):
                st.dataframe(df_filtered)
    else:
        st.info("Press 'Run Analysis' in the sidebar to load historical ratings data.")

# PAGE: Analyst Recommendations
elif page == "Recommendations":
    st.header("Analyst Recommendations")
    st.write("This section presents the monthly analyst recommendations for the selected ticker. The stacked bar chart aggregates various recommendation types over time.")
    with st.expander("Category Description", expanded=False):
        description = textwrap.dedent("""
            - **Strong Buy**: Indicates that analysts are very confident the stock will perform strongly.
            - **Buy**: Reflects a positive outlook from analysts regarding future performance.
            - **Hold**: Suggests that analysts expect the stock to maintain its current performance.
            - **Sell**: Signals a negative outlook, indicating the stock may underperform.
            - **Strong Sell**: Represents a very bearish sentiment, with analysts expecting significant underperformance.
            The chart lets you observe shifts in market sentiment over time and compare the prevalence of each recommendation type.
        """)
        st.markdown(description)
    # No additional page-specific inputs; thus, no expander is shown.
    if st.sidebar.button("Run Analysis", key="analyst_run_button"):
        st.session_state.run_pressed_analyst = True

    if st.session_state.run_pressed_analyst:
        if st.session_state.analyst_data is None:
            st.session_state.analyst_data = load_analyst_data(ticker)
        df_analyst = st.session_state.analyst_data
        if df_analyst is not None:
            st.subheader(f"Analyst Recommendations for {ticker}")
            st.write("The stacked bar chart below shows monthly analyst recommendations.")
            rating_cols = [
                "analystRatingsStrongBuy",
                "analystRatingsbuy",
                "analystRatingsHold",
                "analystRatingsSell",
                "analystRatingsStrongSell"
            ]
            df_melted = pd.melt(
                df_analyst,
                id_vars=["date"],
                value_vars=rating_cols,
                var_name="RatingType",
                value_name="Count"
            )
            color_map = {
                "analystRatingsStrongBuy": "green",
                "analystRatingsbuy": "lightgreen",
                "analystRatingsHold": "orange",
                "analystRatingsSell": "lightcoral",
                "analystRatingsStrongSell": "red"
            }
            rating_order = [
                "analystRatingsStrongBuy",
                "analystRatingsbuy",
                "analystRatingsHold",
                "analystRatingsSell",
                "analystRatingsStrongSell"
            ]
            try:
                fig2 = px.bar(
                    df_melted,
                    x="date",
                    y="Count",
                    color="RatingType",
                    text="Count",
                    title=f"Monthly Analyst Recommendations for {ticker.upper()}",
                    color_discrete_map=color_map,
                    category_orders={"RatingType": rating_order},
                    custom_data=["RatingType"],
                    labels={"date": "Date", "Count": "Number of Recommendations", "RatingType": "Recommendation Type"}
                )
                fig2.update_layout(barmode="stack")
                fig2.update_traces(
                    texttemplate="%{text}",
                    textposition="inside",
                    hovertemplate="<b>Date</b>: %{x}<br><b>Count</b>: %{y}<br><b>Rating Type</b>: %{customdata[0]}<extra></extra>"
                )
                st.plotly_chart(fig2, use_container_width=True)
            except Exception:
                st.error("Error displaying the chart. Please check the data and inputs.")
            with st.expander("Show Detailed Analyst Data"):
                st.dataframe(df_analyst)
    else:
        st.info("Press 'Run Analysis' in the sidebar to load analyst recommendations data.")


# Hide default Streamlit style
st.markdown(
    """
    <style>
    #MainMenu {visibility: hidden;}
    footer {visibility: hidden;}
    </style>
    """,
    unsafe_allow_html=True
)