plot.py

import csv
import sys
import matplotlib
matplotlib.use('Agg')# no display needed — saves to file
import matplotlib.pyplot as plt
import matplotlib.ticker as mticker
from collections import defaultdict

#Color per policy
COLOR = {
    'FIFO'   : '#4C72B0',   # blue
    'LRU'    : '#55A868',   # green
    'Random' : '#C44E52',   # red
    'Belady' : '#8172B2',   # purple  ← optimal upper bound
    'SA-LRU' : '#DD8452',   # orange  ← our proposed policy
}
ORDER = ['FIFO', 'LRU', 'Random', 'Belady', 'SA-LRU']


# Read results_fixed.csv
def read_fixed(path='results_fixed.csv'):
    """
    Returns { policy -> { hit_rate: %, miss_rate: % } }
    """
    data = {}
    try:
        with open(path) as f:
            for row in csv.DictReader(f):
                data[row['policy']] = {
                    'hit_rate' : float(row['hit_rate'])  * 100,
                    'miss_rate': float(row['miss_rate']) * 100,
                }
    except FileNotFoundError:
        print(f"ERROR: {path} not found. Run ./cache_sim first.")
        sys.exit(1)
    return data


# ── Read results.csv (cache size sweep) ──────────────────────
def read_sweep(path='results.csv'):
    """
    Returns { policy -> { cache_kb -> hit_rate% } }
    """
    data = defaultdict(dict)
    try:
        with open(path) as f:
            for row in csv.DictReader(f):
                data[row['policy']][int(row['cache_kb'])] = float(row['hit_rate']) * 100
    except FileNotFoundError:
        print(f"ERROR: {path} not found. Run ./cache_sim first.")
        sys.exit(1)
    return data


# Plotting
def plot(fixed, sweep, out='cs204_results.png'):
    fig, axes = plt.subplots(2, 2, figsize=(14, 9))
    fig.suptitle('CS204 — Cache Replacement Policy Analysis',
                 fontsize=14, fontweight='bold', y=1.01)
    plt.tight_layout(pad=3.5)

    policies   = [p for p in ORDER if p in fixed]
    hit_rates  = [fixed[p]['hit_rate']  for p in policies]
    miss_rates = [fixed[p]['miss_rate'] for p in policies]
    colors     = [COLOR[p] for p in policies]
    belady_hr  = fixed.get('Belady', {}).get('hit_rate', 0)

    #Graph 1: Hit Rate Bar
    ax = axes[0][0]
    bars = ax.bar(policies, hit_rates, color=colors, edgecolor='white',
                  linewidth=0.8, width=0.55)
    ax.set_title('Hit Rate by Policy\n(4 KB cache, 64 B block, 4-way)', fontsize=11)
    ax.set_ylabel('Hit Rate (%)')
    ax.set_ylim(max(0, min(hit_rates) - 5), 100)
    ax.yaxis.set_major_formatter(mticker.PercentFormatter(xmax=100))
    for bar, val in zip(bars, hit_rates):
        ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 0.2,
                f'{val:.2f}%', ha='center', va='bottom', fontsize=8)
    if belady_hr:
        ax.axhline(y=belady_hr, color=COLOR['Belady'], linestyle='--',
                   linewidth=1.3, label=f'Belady bound ({belady_hr:.2f}%)')
        ax.legend(fontsize=8)

    # ── Graph 2: Miss Rate Bar ────────────────────────────
    ax = axes[0][1]
    bars = ax.bar(policies, miss_rates, color=colors, edgecolor='white',
                  linewidth=0.8, width=0.55)
    ax.set_title('Miss Rate by Policy', fontsize=11)
    ax.set_ylabel('Miss Rate (%)')
    ax.set_ylim(0, max(miss_rates) * 1.35 if miss_rates else 30)
    ax.yaxis.set_major_formatter(mticker.PercentFormatter(xmax=100))
    for bar, val in zip(bars, miss_rates):
        ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 0.05,
                f'{val:.2f}%', ha='center', va='bottom', fontsize=8)

    # Graph 3: Hit Rate vs Cache Size 
    ax = axes[1][0]
    cache_sizes = sorted(next(iter(sweep.values())).keys()) if sweep else []
    for pol in ORDER:
        if pol not in sweep: continue
        vals = [sweep[pol].get(sz, 0) for sz in cache_sizes]
        ax.plot(cache_sizes, vals, marker='o', label=pol,
                color=COLOR[pol],
                linewidth=2.0 if pol in ('Belady', 'SA-LRU') else 1.3,
                linestyle='--' if pol == 'Belady' else '-',
                zorder=3 if pol in ('Belady', 'SA-LRU') else 2)
    ax.set_title('Hit Rate vs Cache Size', fontsize=11)
    ax.set_xlabel('Cache Size (KB)')
    ax.set_ylabel('Hit Rate (%)')
    ax.set_xticks(cache_sizes)
    ax.yaxis.set_major_formatter(mticker.PercentFormatter(xmax=100))
    ax.legend(fontsize=8, loc='lower right')
    ax.grid(True, linestyle=':', alpha=0.5)

    #Graph 4: SA-LRU vs LRU vs Belady
    ax = axes[1][1]
    if all(p in sweep for p in ('LRU', 'SA-LRU', 'Belady')):
        lru_v   = [sweep['LRU'].get(sz, 0)    for sz in cache_sizes]
        salru_v = [sweep['SA-LRU'].get(sz, 0) for sz in cache_sizes]
        bel_v   = [sweep['Belady'].get(sz, 0) for sz in cache_sizes]

        ax.plot(cache_sizes, lru_v,   marker='s', label='LRU',
                color=COLOR['LRU'],    linewidth=2)
        ax.plot(cache_sizes, salru_v, marker='D', label='SA-LRU (proposed)',
                color=COLOR['SA-LRU'], linewidth=2)
        ax.plot(cache_sizes, bel_v,   marker='^', label='Belady (optimal)',
                color=COLOR['Belady'], linewidth=1.5, linestyle='--')
        ax.fill_between(cache_sizes, lru_v, salru_v,
                        alpha=0.18, color=COLOR['SA-LRU'], label='SA-LRU gain')
        ax.set_title('SA-LRU vs LRU vs Belady\n(proposed policy focus)', fontsize=11)
        ax.set_xlabel('Cache Size (KB)')
        ax.set_ylabel('Hit Rate (%)')
        ax.set_xticks(cache_sizes)
        ax.yaxis.set_major_formatter(mticker.PercentFormatter(xmax=100))
        ax.legend(fontsize=8)
        ax.grid(True, linestyle=':', alpha=0.5)

    plt.savefig(out, dpi=150, bbox_inches='tight')
    print(f"Graph saved → {out}")
    plt.close()


#Main
if __name__ == '__main__':
    print("Reading simulator output...")
    fixed = read_fixed('results_fixed.csv')
    sweep = read_sweep('results.csv')

    belady_hr = fixed.get('Belady', {}).get('hit_rate', 0)

    print(f"\n{'Policy':<10} {'Hit%':>8} {'Miss%':>8}  vs Belady")
    print("-" * 42)
    for p in ORDER:
        if p not in fixed: continue
        gap = fixed[p]['hit_rate'] - belady_hr
        print(f"{p:<10} {fixed[p]['hit_rate']:>7.2f}%  {fixed[p]['miss_rate']:>7.2f}%  {gap:+.2f}%")

    if 'SA-LRU' in fixed and 'LRU' in fixed:
        delta = fixed['SA-LRU']['hit_rate'] - fixed['LRU']['hit_rate']
        print(f"\nSA-LRU improvement over LRU: {delta:+.2f}%")

    plot(fixed, sweep)