File size: 9,076 Bytes
dffbc1f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
#!/usr/bin/env python
"""F0 figures and number macros from ``python -m pulse.tasks.f0`` output.

    python figures/f0/plot_f0.py --in runs/f0 --out figures/out --tables tables

Writes ``f0_grasp_phase_timing.{pdf,png}`` (event-aligned, per-event min-max
scaled pressure / EMG envelope / hand speed, mean +- 0.5 SD, median onsets),
``f0_delay_distributions.{pdf,png}`` (EMG onset, motion onset, paired
motion - EMG difference) and ``<tables>/numbers_f0.tex``
(``\\pulsenum{f0/emg/median}`` etc.).
"""
from __future__ import annotations

import argparse
import json
import sys
from pathlib import Path

import numpy as np

REPO = Path(__file__).resolve().parents[2]
sys.path.insert(0, str(REPO))

import matplotlib  # noqa: E402

matplotlib.use("Agg")
import matplotlib.pyplot as plt  # noqa: E402

# reference palette (categorical slots 1-3 validate all-pairs) and ink tokens
SERIES = {"emg": "#2a78d6", "motion": "#eb6834", "pressure": "#1baf7a"}
INK, INK2, MUTED, GRID, AXIS, SURFACE = "#0b0b0b", "#52514e", "#898781", "#e1e0d9", "#c3c2b7", "#fcfcfb"


def style():
    plt.rcParams.update({
        "pdf.fonttype": 42, "ps.fonttype": 42, "font.size": 9, "axes.titlesize": 9,
        "axes.labelsize": 9, "legend.fontsize": 8, "xtick.labelsize": 8, "ytick.labelsize": 8,
        "axes.edgecolor": AXIS, "axes.labelcolor": INK2, "xtick.color": MUTED, "ytick.color": MUTED,
        "text.color": INK, "axes.grid": True, "grid.color": GRID, "grid.linewidth": 0.6,
        "grid.linestyle": "-", "axes.spines.top": False, "axes.spines.right": False,
        "figure.facecolor": "white", "axes.facecolor": "white", "savefig.facecolor": "white",
        "legend.frameon": False,
    })


def minmax(a):
    a = a - a.min(axis=1, keepdims=True)
    return a / (a.max(axis=1, keepdims=True) + 1e-8)


def phase_figure(res, z, out: Path):
    tr, t = z["traces"], z["time_ms"]
    st = res["stats"]
    fig, ax = plt.subplots(figsize=(5.2, 2.9))
    ax.axvline(0, color=INK2, lw=0.8)
    ax.text(8, 1.07, "contact", color=INK2, fontsize=8, va="bottom")
    for ch, key, label in ((1, "emg", "EMG envelope"), (2, "motion", "Hand speed"),
                           (0, "pressure", "Contact pressure")):
        a = minmax(tr[:, ch, :].astype(np.float64))
        m, s = a.mean(0), a.std(0)
        ax.fill_between(t, m - 0.5 * s, m + 0.5 * s, color=SERIES[key], alpha=0.12, lw=0)
        ax.plot(t, m, color=SERIES[key], lw=1.6, label=label, solid_capstyle="round")
    meds = {k: st[f"{k}_delay_ms"]["median"] for k in ("emg", "motion")}
    later = max(meds, key=meds.get)
    for key, y in (("emg", 0.93), ("motion", 0.93)):
        med = meds[key]
        ax.plot([med, med], [0, y], color=SERIES[key], lw=1.0)
        ax.plot([med], [y], "o", ms=4, color=SERIES[key], mec="white", mew=1.0)
        right = key == later            # later onset labelled to the right, earlier to the left
        ax.text(med + (12 if right else -12), y, f"{'EMG' if key == 'emg' else 'motion'} onset\n"
                f"median {med:.0f} ms", ha="left" if right else "right", va="center",
                fontsize=7, color=INK)
    c = res["counts"]
    ax.set_title(f"Right-hand contacts: {c['used']} events, {c['recordings_with_events']} recordings, "
                 f"{c['volunteers_with_events']} volunteers", loc="left", color=INK)
    ax.set_xlabel("time relative to contact onset (ms)")
    ax.set_ylabel("scaled amplitude")
    ax.set_xlim(t[0], t[-1])
    ax.set_ylim(-0.05, 1.18)
    ax.legend(loc="upper left", bbox_to_anchor=(0.0, 0.93), handlelength=1.6)
    fig.tight_layout()
    for ext in ("pdf", "png"):
        fig.savefig(out / f"f0_grasp_phase_timing.{ext}", dpi=200)
    plt.close(fig)


def delay_figure(res, events_ms, out: Path):
    emg, mot = events_ms
    d = mot - emg
    st = res["stats"]
    pe = st["paired_motion_minus_emg"]["event_level"]
    fig, axes = plt.subplots(1, 3, figsize=(7.4, 2.5), sharey=False)
    bins = np.arange(-1500, 1, 50)
    for ax, x, key, title in ((axes[0], emg, "emg", "EMG onset"), (axes[1], mot, "motion", "Motion onset")):
        s = st[f"{key}_delay_ms"]
        ax.hist(x, bins=bins, color=SERIES[key], edgecolor="white", linewidth=0.6)
        ax.axvline(s["median"], color=INK, lw=1.0)
        ax.set_title(f"{title}\nmedian {s['median']:.0f} ms, IQR {s['iqr']:.0f} ms", loc="left")
        ax.set_xlabel("ms relative to contact")
    lim = max(50, float(np.percentile(np.abs(d), 99)))
    b = np.arange(-np.ceil(lim / 50) * 50, np.ceil(lim / 50) * 50 + 1, 50)
    axes[2].hist(d, bins=b, color=INK2, edgecolor="white", linewidth=0.6)
    axes[2].axvline(0, color=AXIS, lw=1.0)
    axes[2].axvline(pe["median_diff_ms"], color=INK, lw=1.0)
    axes[2].set_title(f"Motion - EMG (paired)\n"
                      f"median {pe['median_diff_ms']:.0f} ms, p = {pe['wilcoxon']['p']:.2g}", loc="left")
    axes[2].set_xlabel("ms (negative: motion first)")
    axes[0].set_ylabel("events")
    fig.tight_layout()
    for ext in ("pdf", "png"):
        fig.savefig(out / f"f0_delay_distributions.{ext}", dpi=200)
    plt.close(fig)


def numbers(res) -> dict:
    st, c = res["stats"], res["counts"]
    pm = st["paired_motion_minus_emg"]
    ev, vl = pm["event_level"], pm["volunteer_level"]
    n = {"f0/events": str(c["used"]), "f0/onsets": str(c["onsets"]),
         "f0/recordings": str(c["recordings_with_events"]), "f0/recordingsscanned": str(c["recordings"]),
         "f0/volunteers": str(c["volunteers_with_events"])}
    for key in ("emg", "motion"):
        s = st[f"{key}_delay_ms"]
        for k in ("mean", "sd", "median", "p25", "p75", "iqr"):
            n[f"f0/{key}/{k}"] = f"{s[k]:.0f}"
    n.update({"f0/diff/mean": f"{ev['mean_diff_ms']:.1f}", "f0/diff/median": f"{ev['median_diff_ms']:.0f}",
              "f0/diff/sd": f"{ev['sd_ms']:.0f}",
              "f0/diff/cilo": f"{ev['ci95_mean_ms'][0]:.0f}", "f0/diff/cihi": f"{ev['ci95_mean_ms'][1]:.0f}",
              "f0/diff/pt": _p(ev["paired_t"]["p"]), "f0/diff/pw": _p(ev["wilcoxon"]["p"]),
              "f0/diff/motionfirst": f"{100 * ev['motion_first_fraction']:.0f}",
              "f0/diff/r": ("--" if ev["pearson_r_emg_motion"] is None
                            else f"{ev['pearson_r_emg_motion']:.2f}"),
              "f0/vol/n": str(vl["n_volunteers"]),
              "f0/vol/median": f"{vl['median_of_volunteer_median_diffs_ms']:.0f}"})
    if "wilcoxon_volunteer_median_diffs" in vl:
        n["f0/vol/pw"] = _p(vl["wilcoxon_volunteer_median_diffs"]["p"])
        n["f0/vol/pt"] = _p(vl["t_volunteer_mean_diffs"]["p"])
        # CI of f0/vol/median (median of the per-volunteer median differences)
        lo, hi = vl["bootstrap_median_of_volunteer_medians_ci95_ms"]
        n["f0/vol/cilo"], n["f0/vol/cihi"] = f"{lo:.0f}", f"{hi:.0f}"
        # pooled event-level median difference (= f0/diff/median), volunteers as clusters
        lo, hi = vl["cluster_bootstrap_pooled_median_diff_ci95_ms"]
        n["f0/diff/medcilo"], n["f0/diff/medcihi"] = f"{lo:.0f}", f"{hi:.0f}"
    return n


def _p(p):
    from pulse.stats import format_p
    return format_p(p)


def write_numbers(nums: dict, path: Path, res: dict):
    from pulse.aggregate import NUMBERS_PREAMBLE
    prov = res["provenance"]
    lines = [NUMBERS_PREAMBLE, f"% ---- F0 case study (python -m pulse.tasks.f0): git {prov.get('git_sha')} "
             f"data {prov.get('data_tag')} config md5 {prov.get('config_md5')}"]
    for k, v in sorted(nums.items()):
        lines.append(r"\expandafter\def\csname pulse@" + k + r"\endcsname{" + v + "}")
    path.write_text("\n".join(lines) + "\n", encoding="utf-8", newline="\n")


def main(argv=None) -> int:
    ap = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
    ap.add_argument("--in", dest="inp", default="runs/f0")
    ap.add_argument("--out", default="figures/out")
    ap.add_argument("--tables", default="tables")
    args = ap.parse_args(argv)
    inp, out, tab = Path(args.inp), Path(args.out), Path(args.tables)
    out.mkdir(parents=True, exist_ok=True)
    tab.mkdir(parents=True, exist_ok=True)
    res = json.loads((inp / "f0_timing.json").read_text(encoding="utf-8"))
    if res.get("schema") != "pulse-f0/1":
        raise SystemExit(f"{inp / 'f0_timing.json'} is not an F0 result")
    z = np.load(inp / "f0_traces.npz")
    import csv
    with open(inp / "f0_events.csv", encoding="utf-8") as f:
        rows = list(csv.DictReader(f))
    emg = np.array([float(r["emg_delay_ms"]) for r in rows])
    mot = np.array([float(r["motion_delay_ms"]) for r in rows])
    if len(rows) != res["counts"]["used"] or z["traces"].shape[0] != len(rows):
        raise SystemExit("f0_events.csv / f0_traces.npz do not match f0_timing.json")
    style()
    phase_figure(res, z, out)
    delay_figure(res, (emg, mot), out)
    write_numbers(numbers(res), tab / "numbers_f0.tex", res)
    print(f"wrote {out / 'f0_grasp_phase_timing.pdf'}, {out / 'f0_delay_distributions.pdf'}, "
          f"{tab / 'numbers_f0.tex'}")
    return 0


if __name__ == "__main__":
    sys.exit(main())