pages/2_player_comparison_images.py

import streamlit as st
import pandas as pd
import numpy as np
import joblib
import cv2
from PIL import Image
import plotly.express as px
import plotly.graph_objects as go

# -------------------------------
# 1. Setup
# -------------------------------
st.set_page_config(page_title="Cricket Player Comparision", layout="centered")
st.title("🏏 Cricket Player Comparision Tool")

# -------------------------------
# 2. Load Data & Models
# -------------------------------
@st.cache_data
def load_dataset():
    return pd.read_csv("cric_final.csv")

@st.cache_resource
def load_assets():
    model = joblib.load("svc_face_classifier.pkl")
    encoder = joblib.load("label_encoder.pkl")
    detector = cv2.CascadeClassifier(cv2.data.haarcascades + "haarcascade_frontalface_default.xml")
    return model, encoder, detector

df = load_dataset()
model, encoder, detector = load_assets()
formats = ["Test", "ODI", "T20", "IPL"]

# -------------------------------
# 3. Player Detection from Image
# -------------------------------
def detect_player(image_file, model, encoder, detector):
    try:
        img = Image.open(image_file).convert("RGB")
        img_np = np.array(img)
        gray = cv2.cvtColor(img_np, cv2.COLOR_RGB2GRAY)
        faces = detector.detectMultiScale(gray, 1.3, 5)
        if not len(faces):
            return None, "No face detected"
        x, y, w, h = faces[0]
        face = cv2.resize(gray[y:y+h, x:x+w], (64, 64)).flatten().reshape(1, -1)
        label = model.predict(face)[0]
        name = encoder.inverse_transform([label])[0]
        return name, None
    except Exception as e:
        return None, str(e)

# -------------------------------
# 4. Player Stats Summary
# -------------------------------
def get_summary(player_row, formats):
    return {
        "Total Runs": sum(player_row.get(f'batting_Runs_{f}', 0) for f in formats),
        "Total Wickets": sum(player_row.get(f'bowling_{f}_Wickets', 0) for f in formats),
        "Best Strike Rate": max(player_row.get(f'batting_SR_{f}', 0) for f in formats)
    }

# -------------------------------
# 5. Upload Player Images
# -------------------------------
col1, col2 = st.columns(2)
with col1:
    img1 = st.file_uploader("Upload Player 1 Image", type=["jpg", "png", "jpeg"], key="img1")
with col2:
    img2 = st.file_uploader("Upload Player 2 Image", type=["jpg", "png", "jpeg"], key="img2")

p1_name, p2_name = None, None

if img1:
    p1_name, err1 = detect_player(img1, model, encoder, detector)
    if err1:
        st.error(f"Player 1 Error: {err1}")
    else:
        col1.image(img1, caption=f"Player 1: {p1_name}", width=200)

if img2:
    p2_name, err2 = detect_player(img2, model, encoder, detector)
    if err2:
        st.error(f"Player 2 Error: {err2}")
    else:
        col2.image(img2, caption=f"Player 2: {p2_name}", width=200)

if not (p1_name and p2_name):
    st.warning("Upload images for both players to load the data.")
    st.stop()

if p1_name not in df["Player"].values or p2_name not in df["Player"].values:
    st.error("One or both players are not in the dataset.")
    st.stop()

# -------------------------------
# 6. Stats Extraction
# -------------------------------
player1 = df[df["Player"] == p1_name].iloc[0]
player2 = df[df["Player"] == p2_name].iloc[0]
stats1 = get_summary(player1, formats)
stats2 = get_summary(player2, formats)

# -------------------------------
# 7. Display Summary Stats
# -------------------------------
st.subheader("📊 Player Summary")
col1, col2 = st.columns(2)
for col, stats in zip([col1, col2], [stats1, stats2]):
    col.metric("Total Runs", stats["Total Runs"])
    col.metric("Total Wickets", stats["Total Wickets"])
    col.metric("Best SR", round(stats["Best Strike Rate"], 2))

# -------------------------------
# 8. Visual Comparisons
# -------------------------------
st.markdown("## 📈 Visual Comparison")

# Batting
bat_df = pd.DataFrame({
    "Format": formats,
    p1_name: [player1.get(f'batting_Runs_{f}', 0) for f in formats],
    p2_name: [player2.get(f'batting_Runs_{f}', 0) for f in formats]
})
st.plotly_chart(px.bar(bat_df, x="Format", y=[p1_name, p2_name], barmode="group", title="Batting Runs"))

# Bowling
bowl_df = pd.DataFrame({
    "Format": formats,
    p1_name: [player1.get(f'bowling_{f}_Wickets', 0) for f in formats],
    p2_name: [player2.get(f'bowling_{f}_Wickets', 0) for f in formats]
})
st.plotly_chart(px.bar(bowl_df, x="Format", y=[p1_name, p2_name], barmode="group", title="Bowling Wickets"))

# -------------------------------

# Strike Rate Comparison
sr_df = pd.DataFrame({
    "Format": formats,
    p1_name: [player1.get(f'batting_SR_{f}', 0) for f in formats],
    p2_name: [player2.get(f'batting_SR_{f}', 0) for f in formats]
})
st.plotly_chart(px.bar(sr_df, x="Format", y=[p1_name, p2_name], barmode="group", title="Strike Rate Comparison"))

#---------------------------------
# 9. Match Distribution
# -------------------------------
for name, row, col, key_suffix in zip(
    [p1_name, p2_name], 
    [player1, player2], 
    [col1, col2], 
    ["player1", "player2"]
):
    values = [row.get(f"Matches_{f}", 0) for f in formats]
    fig = px.pie(
        values=values, 
        names=formats, 
        title=f"{name}'s Match Distribution"
    )
    col.plotly_chart(fig, key=f"pie_chart_{key_suffix}")  # 👈 Unique key

# 10. Milestones
# -------------------------------
st.subheader("🏆 Milestones")
for fmt in formats:
    st.markdown(f"### {fmt}")
    col1, col2 = st.columns(2)
    for m in ["50s", "100s", "200s"]:
        col1.metric(f"{p1_name} {m}", player1.get(f"batting_{m}_{fmt}", 0))
        col2.metric(f"{p2_name} {m}", player2.get(f"batting_{m}_{fmt}", 0))