app.py

import streamlit as st
import pandas as pd
import plotly.express as px

st.set_page_config(page_title="Trader Behavior vs Sentiment", layout="wide")

st.title("Primetrade.ai Data Science Intern Assignment")
st.markdown("### Analyzing Trader Performance vs Market Sentiment on Hyperliquid")

@st.cache_data
def load_data():
    try:
        df = pd.read_csv('aggregated_trader_metrics.csv')
        df['date'] = pd.to_datetime(df['date'])
        return df
    except FileNotFoundError:
        return pd.DataFrame()

df = load_data()

if df.empty:
    st.warning("Data not found. Please run analysis.py first to generate the aggregated metrics.")
else:
    st.sidebar.header("Filters")
    sentiment_filter = st.sidebar.multiselect("Sentiment", options=df['Sentiment_Broad'].unique(), default=df['Sentiment_Broad'].unique())
    
    filtered_df = df[df['Sentiment_Broad'].isin(sentiment_filter)]
    
    col1, col2 = st.columns(2)
    
    with col1:
        st.subheader("Average PnL by Sentiment")
        fig1 = px.box(filtered_df, x="Sentiment_Broad", y="Closed PnL", color="Sentiment_Broad", title="PnL Distribution")
        st.plotly_chart(fig1, use_container_width=True)
        
    with col2:
        st.subheader("Win Rate Distribution")
        fig2 = px.box(filtered_df, x="Sentiment_Broad", y="Win Rate", color="Sentiment_Broad", title="Win Rate")
        st.plotly_chart(fig2, use_container_width=True)
        
    st.subheader("Behavioral Changes")
    col3, col4, col5 = st.columns(3)
    
    with col3:
        fig3 = px.histogram(filtered_df, x="Trade Count", color="Sentiment_Broad", barmode="overlay", title="Trade Frequency")
        st.plotly_chart(fig3, use_container_width=True)
        
    with col4:
        fig4 = px.box(filtered_df, x="Sentiment_Broad", y="Avg Trade Size USD", color="Sentiment_Broad", title="Trade Size (USD)")
        # filter out extreme outliers for better visualization
        fig4.update_yaxes(range=[0, filtered_df['Avg Trade Size USD'].quantile(0.95)])
        st.plotly_chart(fig4, use_container_width=True)
        
    with col5:
        fig5 = px.box(filtered_df, x="Sentiment_Broad", y="Long/Short Ratio", color="Sentiment_Broad", title="Long/Short Bias")
        fig5.update_yaxes(range=[0, 5])
        st.plotly_chart(fig5, use_container_width=True)
        
    st.markdown("### Actionable Output Segments")
    st.markdown("""
    - **High Leverage vs Low Leverage**: Use Total Volume USD as a proxy.
    - **Frequent vs Infrequent**: Based on median daily Trade Count.
    - **Consistent Winners vs Inconsistent**: Based on Win Rate.
    """)
    
    if st.checkbox("Show Raw Data"):
        st.write(filtered_df.head(100))