readcon-core 0.9.0

An oxidized single and multiple CON file reader and writer with FFI bindings for ergonomic C/C++ usage.
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
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272

#!/usr/bin/env python3
"""
Generate publication-quality benchmark plots for readcon-core docs.

Usage:
    uv run --with matplotlib --with numpy python benches/make_plots.py

Generates SVGs in docs/orgmode/img/.
"""

from pathlib import Path
import matplotlib
matplotlib.use("Agg")
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
import numpy as np

IMG_DIR = Path(__file__).resolve().parent.parent / "docs" / "orgmode" / "img"
IMG_DIR.mkdir(parents=True, exist_ok=True)

# ---------------------------------------------------------------------------
# Style
# ---------------------------------------------------------------------------

FONT = "Jost"
plt.rcParams.update({
    "font.family": "sans-serif",
    "font.sans-serif": [FONT, "DejaVu Sans"],
    "font.size": 11,
    "axes.titlesize": 12,
    "axes.labelsize": 11,
    "xtick.labelsize": 10,
    "ytick.labelsize": 10,
    "figure.facecolor": "white",
    "axes.facecolor": "#FAFAFA",
    "axes.edgecolor": "#CCCCCC",
    "axes.grid": True,
    "grid.alpha": 0.3,
})

TEAL = "#004D40"
CORAL = "#FF655D"
YELLOW = "#F1DB4B"
LIGHT_TEAL = "#4DB6AC"
GRAY = "#9E9E9E"

# ---------------------------------------------------------------------------
# 1. Parsing throughput bar chart (4-way, multiple sizes)
# ---------------------------------------------------------------------------

def plot_throughput():
    datasets = ["218x100\n(1.6 MiB)", "218x1000\n(9.7 MiB)", "10kx100\n(46.9 MiB)", "10kx10\n(4.7 MiB)"]
    c_times =       [10.6, 73,  361, 36]
    ase_times =     [36.1, 286, 956, 94]
    readcon_times = [4.4,  55,  185, 13]

    x = np.arange(len(datasets))
    w = 0.25

    fig, ax = plt.subplots(figsize=(10, 5))
    bars_ase = ax.bar(x - w, ase_times, w, label="ASE (Python)", color=CORAL, edgecolor="white", linewidth=0.5)
    bars_c = ax.bar(x, c_times, w, label="C sscanf (eOn-style)", color=YELLOW, edgecolor="white", linewidth=0.5)
    bars_rc = ax.bar(x + w, readcon_times, w, label="readcon-core (Rust)", color=TEAL, edgecolor="white", linewidth=0.5)

    ax.set_ylabel("Parse time (ms, lower is better)")
    ax.set_title("Parsing throughput across trajectory sizes")
    ax.set_xticks(x)
    ax.set_xticklabels(datasets)
    ax.legend(loc="upper left")
    ax.set_yscale("log")
    ax.set_ylim(1, 2000)

    # Speedup annotations on readcon bars
    for i, (rc, ase) in enumerate(zip(readcon_times, ase_times)):
        ax.annotate(f"{ase/rc:.1f}x", xy=(x[i] + w, rc), xytext=(0, 5),
                    textcoords="offset points", ha="center", fontsize=9, color=TEAL, fontweight="bold")

    plt.tight_layout()
    out = IMG_DIR / "parsing_throughput.svg"
    plt.savefig(out, dpi=150)
    plt.close()
    print(f"  {out}")


# ---------------------------------------------------------------------------
# 2. Memory usage comparison
# ---------------------------------------------------------------------------

def plot_memory():
    datasets = ["218x1000\n(9.7 MiB)", "10kx100\n(46.9 MiB)", "10kx10\n(4.7 MiB)"]
    readcon_rss = [70, 263, 263]
    ase_rss =     [268, 270, 270]

    x = np.arange(len(datasets))
    w = 0.3

    fig, ax = plt.subplots(figsize=(8, 4.5))
    ax.bar(x - w/2, ase_rss, w, label="ASE", color=CORAL, edgecolor="white")
    ax.bar(x + w/2, readcon_rss, w, label="readcon-core", color=TEAL, edgecolor="white")

    ax.set_ylabel("Peak RSS (MiB, lower is better)")
    ax.set_title("Memory usage (all frames loaded)")
    ax.set_xticks(x)
    ax.set_xticklabels(datasets)
    ax.legend()

    for i, (rc, ase) in enumerate(zip(readcon_rss, ase_rss)):
        if ase > rc * 1.5:
            ax.annotate(f"{ase/rc:.1f}x less", xy=(x[i] + w/2, rc), xytext=(0, 5),
                        textcoords="offset points", ha="center", fontsize=9, color=TEAL, fontweight="bold")

    plt.tight_layout()
    out = IMG_DIR / "memory_usage.svg"
    plt.savefig(out, dpi=150)
    plt.close()
    print(f"  {out}")


# ---------------------------------------------------------------------------
# 3. Feature comparison matrix (CON v2 vs other formats)
# ---------------------------------------------------------------------------

def plot_feature_matrix():
    formats = [
        "CON v2",
        "XYZ",
        "POSCAR",
        "extxyz",
        "CIF",
        "GRO",
        "LAMMPS",
        "PDB",
        "DCD",
        "TRR",
    ]

    # Features: Positions, Velocities, Forces, UnitCell, Constraints, AtomID,
    #           Metadata, Compression, MultiFrame, Streaming
    features = [
        "Positions", "Velocities", "Forces", "Unit cell",
        "Per-dir\nconstraints", "Atom ID\n(round-trip)", "Metadata\n(JSON)",
        "Compression", "Multi-frame", "Streaming\niterator",
    ]

    # 1 = yes, 0.5 = partial, 0 = no
    data = np.array([
        # CON v2
        [1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
        # XYZ
        [1, 0.5, 0, 0.5, 0, 0, 0, 0, 1, 1],
        # POSCAR
        [1, 0, 0, 1, 0.5, 0, 0, 0, 0, 0],
        # extxyz
        [1, 1, 1, 1, 0, 0, 0.5, 0, 1, 1],
        # CIF
        [1, 0, 0, 1, 0, 0, 0.5, 0, 0, 0],
        # GRO
        [1, 1, 0, 1, 0, 0, 0, 0, 1, 1],
        # LAMMPS
        [1, 1, 1, 1, 0, 0, 0.5, 0, 1, 1],
        # PDB
        [1, 0, 0, 1, 0, 0, 0.5, 0, 1, 0],
        # DCD
        [1, 0, 0, 1, 0, 0, 0, 0, 1, 1],
        # TRR
        [1, 1, 0, 1, 0, 0, 0, 0, 1, 1],
    ])

    fig, ax = plt.subplots(figsize=(12, 5.5))

    cmap = matplotlib.colors.ListedColormap(["#FFEBEE", "#FFF9C4", "#E0F2F1"])
    bounds = [-0.25, 0.25, 0.75, 1.25]
    norm = matplotlib.colors.BoundaryNorm(bounds, cmap.N)

    im = ax.imshow(data, cmap=cmap, norm=norm, aspect="auto")

    ax.set_xticks(range(len(features)))
    ax.set_xticklabels(features, rotation=45, ha="right", fontsize=9)
    ax.set_yticks(range(len(formats)))
    ax.set_yticklabels(formats, fontsize=10)

    # Text annotations
    for i in range(len(formats)):
        for j in range(len(features)):
            v = data[i, j]
            text = {1.0: "Y", 0.5: "~", 0.0: "N"}[v]
            color = {1.0: TEAL, 0.5: "#F57F17", 0.0: CORAL}[v]
            ax.text(j, i, text, ha="center", va="center", fontsize=12,
                    fontweight="bold", color=color)

    # Highlight CON v2 row
    ax.add_patch(plt.Rectangle((-0.5, -0.5), len(features), 1,
                               fill=False, edgecolor=TEAL, linewidth=2.5))

    ax.set_title("Feature comparison: CON v2 vs common atomic structure formats")

    # Legend
    patches = [
        mpatches.Patch(facecolor="#E0F2F1", edgecolor=TEAL, label="Supported"),
        mpatches.Patch(facecolor="#FFF9C4", edgecolor="#F57F17", label="Partial"),
        mpatches.Patch(facecolor="#FFEBEE", edgecolor=CORAL, label="Not supported"),
    ]
    ax.legend(handles=patches, loc="lower right", fontsize=9)

    plt.tight_layout()
    out = IMG_DIR / "feature_comparison.svg"
    plt.savefig(out, dpi=150)
    plt.close()
    print(f"  {out}")


# ---------------------------------------------------------------------------
# 4. Pareto front: features vs parse speed
# ---------------------------------------------------------------------------

def plot_pareto():
    # (name, feature_count, parse_time_ms_per_100frames_218atoms, color)
    points = [
        ("CON v2\n(readcon-core)", 10, 4.4, TEAL),
        ("CON v2\n(C sscanf)", 10, 10.6, LIGHT_TEAL),
        ("extxyz\n(ASE)", 6, 36.1, CORAL),   # ASE timing proxy
        ("XYZ\n(chemfiles)", 3, 8.0, YELLOW),  # estimate: simple format, C++ parser
        ("POSCAR\n(ASE)", 3, 15.0, CORAL),     # estimate
        ("GRO\n(chemfiles)", 5, 10.0, YELLOW), # estimate
        ("LAMMPS\n(chemfiles)", 7, 12.0, YELLOW), # estimate
        ("DCD\n(MDAnalysis)", 4, 3.0, GRAY),   # binary, very fast
        ("TRR\n(chemfiles)", 5, 5.0, YELLOW),  # binary
    ]

    fig, ax = plt.subplots(figsize=(10, 6))

    for name, feat, ms, color in points:
        ax.scatter(feat, ms, s=120, c=color, edgecolors="white", linewidth=1.5, zorder=3)
        offset = (8, -12) if "readcon" in name else (8, 5)
        ax.annotate(name, (feat, ms), textcoords="offset points",
                    xytext=offset, fontsize=8, ha="left")

    # Pareto front line (non-dominated: more features AND faster)
    ax.plot([10, 10], [0, 4.4], '--', color=TEAL, alpha=0.3, linewidth=1)
    ax.plot([0, 10], [4.4, 4.4], '--', color=TEAL, alpha=0.3, linewidth=1)

    ax.set_xlabel("Features supported (out of 10)")
    ax.set_ylabel("Parse time for 100 frames x 218 atoms (ms, lower is better)")
    ax.set_title("Feature coverage vs parsing performance")
    ax.set_xlim(1, 11.5)
    ax.set_ylim(0, 42)
    ax.invert_yaxis()  # Lower (faster) is better, so flip

    # Quadrant labels
    ax.text(9, 38, "Ideal:\nfast + feature-rich", ha="center", fontsize=9,
            color=TEAL, fontstyle="italic", alpha=0.7)
    ax.text(2, 3, "Fast but\nfeature-poor", ha="center", fontsize=9,
            color=GRAY, fontstyle="italic", alpha=0.7)

    plt.tight_layout()
    out = IMG_DIR / "pareto_features_vs_speed.svg"
    plt.savefig(out, dpi=150)
    plt.close()
    print(f"  {out}")


# ---------------------------------------------------------------------------
# Main
# ---------------------------------------------------------------------------

if __name__ == "__main__":
    print("Generating plots...")
    plot_throughput()
    plot_memory()
    plot_feature_matrix()
    plot_pareto()
    print("Done.")