map2fig 0.7.7

Fast, publication-quality HEALPix sky map visualization in Rust
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222

#!/usr/bin/env python3
"""
Column Data I/O Profiler for HEALPix Plotter

Profiles the column data reading performance to identify I/O bottlenecks.
Measures the time spent reading FITS column data vs. other operations.

Usage:
    export MAP2FIG_PROFILE=1
    python3 profile_columns.py <fits_file> [--runs=N] [--col=INDEX]
"""

import subprocess
import sys
import time
import re
from pathlib import Path
from statistics import mean, stdev

def extract_timing_from_run(fits_file: str, col_idx: int = 0, output_res: int = 512) -> dict:
    """
    Run a single render and extract performance timing.
    """
    try:
        # Run the program with time prefix to capture timing
        result = subprocess.run(
            ['time', '-v', 'cargo', 'run', '--release', '--',
             '-f', fits_file,
             '-o', f'/tmp/profile_col_out_{int(time.time())}.pdf',
             '--width', str(output_res),
             '-i', str(col_idx)],
            capture_output=True,
            text=True,
            timeout=300,
            cwd='/home/dwatts/projects/healpix_plotter',
            env={**dict(__import__('os').environ), 'MAP2FIG_PROFILE': '1'}
        )
        
        output_text = result.stderr + result.stdout
        
        # Extract relevant timings
        data = {
            'returncode': result.returncode,
            'stdout': result.stdout,
            'stderr': result.stderr,
            'user_time': 0.0,
            'wall_time': 0.0,
            'max_rss': 0,
            'cache_ops': {'hits': 0, 'misses': 0},
            'fits_parse_ms': 0.0,
        }
        
        # Try to parse timing information
        for line in output_text.split('\n'):
            # Parse time -v output
            if 'User time (seconds):' in line:
                try:
                    data['user_time'] = float(line.split(':')[1].strip())
                except:
                    pass
            elif 'Elapsed (wall clock) time' in line:
                parts = line.split(':')[1].strip().split(':')
                try:
                    if len(parts) == 3:  # MM:SS.SS
                        m, s = int(parts[0]), float(parts[1] + '.' + parts[2])
                        data['wall_time'] = m * 60 + s
                    elif len(parts) == 2:  # SS.SS
                        data['wall_time'] = float(parts[1])
                except:
                    pass
            elif 'Maximum resident set size' in line:
                try:
                    data['max_rss'] = int(line.split(':')[1].strip())
                except:
                    pass
            
            # Parse diagnostics
            if '[I/O DIAG] Cache HIT' in line:
                data['cache_ops']['hits'] += 1
            elif '[I/O DIAG] Cache MISS' in line:
                data['cache_ops']['misses'] += 1
            elif 'FITS parse took' in line:
                match = re.search(r'took\s+([\d.]+)ms', line)
                if match:
                    data['fits_parse_ms'] = float(match.group(1))
        
        return data
    except subprocess.TimeoutExpired:
        print(f"ERROR: Timeout (>300s)", file=sys.stderr)
        return None
    except Exception as e:
        print(f"ERROR: {e}", file=sys.stderr)
        return None

def main():
    if len(sys.argv) < 2:
        print("Usage: python3 profile_columns.py <fits_file> [--runs=N] [--col=INDEX]")
        print("\nExample:")
        print("  python3 profile_columns.py combined_map_95GHz_nside8192_ptsrcmasked_50mJy.fits --runs=2")
        sys.exit(1)
    
    fits_file = sys.argv[1]
    num_runs = 1
    col_idx = 0
    
    for arg in sys.argv[2:]:
        if arg.startswith('--runs='):
            num_runs = int(arg.split('=')[1])
        elif arg.startswith('--col='):
            col_idx = int(arg.split('=')[1])
    
    # Resolve path
    fits_path = Path(fits_file)
    if not fits_path.exists():
        fits_path = Path('/home/dwatts/projects/healpix_plotter') / fits_file
        if not fits_path.exists():
            print(f"ERROR: FITS file not found: {fits_file}", file=sys.stderr)
            sys.exit(1)
    
    print(f"\n=== Column Data I/O Profile ===")
    print(f"File:      {fits_path}")
    print(f"Column:    {col_idx}")
    print(f"Runs:      {num_runs}")
    print(f"\nRunning profiling (includes compilation on first run)...\n")
    
    results = []
    for i in range(num_runs):
        print(f"[{i+1}/{num_runs}] Profiling...", end='', flush=True)
        start = time.time()
        result = extract_timing_from_run(str(fits_path), col_idx, 512)
        elapsed = time.time() - start
        
        if result is None:
            print(" FAILED")
            continue
        
        if result['returncode'] != 0:
            print(f" FAILED")
            continue
        
        results.append(result)
        print(f" OK ({elapsed:.1f}s wall, {result['user_time']:.1f}s CPU, RSS: {result['max_rss']//1024}MB)")
        
        # Show diagnostics
        cache_total = result['cache_ops']['hits'] + result['cache_ops']['misses']
        if cache_total > 0:
            hit_pct = 100.0 * result['cache_ops']['hits'] / cache_total
            print(f"     Cache: {result['cache_ops']['hits']} hits, {result['cache_ops']['misses']} misses ({hit_pct:.0f}%)")
        if result['fits_parse_ms'] > 0:
            print(f"     FITS parse: {result['fits_parse_ms']:.2f}ms")
    
    if not results:
        print("ERROR: No successful profiles", file=sys.stderr)
        sys.exit(1)
    
    # Analysis
    print(f"\n=== Analysis ({len(results)} runs) ===\n")
    
    wall_times = [r['wall_time'] for r in results if r['wall_time'] > 0]
    user_times = [r['user_time'] for r in results if r['user_time'] > 0]
    
    if wall_times:
        print(f"Wall Clock Time (seconds):")
        print(f"  Min:    {min(wall_times):8.2f}s")
        print(f"  Max:    {max(wall_times):8.2f}s")
        print(f"  Mean:   {mean(wall_times):8.2f}s")
        if len(wall_times) > 1:
            print(f"  StdDev: {stdev(wall_times):8.2f}s")
    
    if user_times:
        print(f"\nCPU Time (seconds):")
        print(f"  Min:    {min(user_times):8.2f}s")
        print(f"  Max:    {max(user_times):8.2f}s")
        print(f"  Mean:   {mean(user_times):8.2f}s")
        if len(user_times) > 1:
            print(f"  StdDev: {stdev(user_times):8.2f}s")
    
    # Memory analysis
    max_rss = [r['max_rss'] for r in results]
    if max_rss:
        rss_mb = [m / 1024 for m in max_rss]
        print(f"\nMemory Usage (MB):")
        print(f"  Min:    {min(rss_mb):8.1f}MB")
        print(f"  Max:    {max(rss_mb):8.1f}MB")
        print(f"  Mean:   {mean(rss_mb):8.1f}MB")
    
    # Cache effectiveness
    all_hits = sum(r['cache_ops']['hits'] for r in results)
    all_misses = sum(r['cache_ops']['misses'] for r in results)
    total_cache_ops = all_hits + all_misses
    
    if total_cache_ops > 0:
        hit_rate = 100.0 * all_hits / total_cache_ops
        print(f"\nCache Operations: {total_cache_ops} ({all_hits} hits, {all_misses} misses, {hit_rate:.0f}% hit rate)")
        
        if hit_rate > 50 and all_misses > 0:
            first_run_wall = wall_times[0] if wall_times else 0
            cached_run_wall = min(wall_times) if len(wall_times) > 1 else wall_times[0] if wall_times else 0
            
            if first_run_wall > 0 and cached_run_wall > 0:
                cache_benefit = (first_run_wall - cached_run_wall) / first_run_wall * 100
                print(f"Cache Benefit:    {cache_benefit:.1f}% speedup on cached runs")
    
    print(f"\n=== Interpretation ===\n")
    
    if wall_times and max(wall_times) > 30:
        print("⚠ Render time is significant (>30s)")
        print("  Architecture breakdown likely: I/O ~14%, compute ~38%, rendering ~48%")
        print("  Main optimization opportunities:")
        print("    1. PDF rendering (48%) - largest bottleneck")
        print("    2. Pixel operations (38%) - SIMD already optimized")
        print("    3. I/O (14%) - metadata cached, column I/O efficient")
    
    if hit_rate > 90:
        print("\n✓ Metadata caching is very effective")
        print("  Further I/O optimization should focus on:")
        print("    - Column data layout/access patterns")
        print("    - Sparse map expansion efficiency")
        print("    - Binary table streaming (if supported by fitsrs)")

if __name__ == '__main__':
    main()