visualdynamics.report.plot_conventions

plot_conventions

Plot conventions for a report's figures (report feature plot_conventions, with its option shared_decades).

Three habits a figure in a test report is expected to keep:

• An ASD or an SRS opens on log-log axes. A specification's breakpoints are joined by straight lines, and its tolerance band has a constant width, only on log-log axes; a narrowband comparison otherwise opens on a linear frequency axis.
• Curves are told apart by line style as well as color, and a figure of up to END_LABEL_CURVES curves names each one at its end, so the figure survives a monochrome copy and a reader who does not see color.
• With shared_decades, every figure of a kind (the ASD figures of one quantity and unit, the SRS figures likewise) opens on the same whole decades, so two figures can be compared by eye.

Each acts on the figures already built for the page, so the figure builders are unchanged and a report without the feature is drawn as it always was.

Functions:

Name	Description
figure_kind	'asd' for a plot of a power spectral density (a measured PSD, a
apply	Put the page's figures under the conventions, in place.
share_decades	Give every log-log figure of a kind the same opening window, in

Functions:

figure_kind

figure_kind(block: Mapping[str, Any], objects: Mapping[str, Any], object_groups: Sequence[Mapping[str, Any]] | None = None) -> str | None

'asd' for a plot of a power spectral density (a measured PSD, a specification, or the two compared), 'srs' for a plot of a shock response spectrum (with its specification or alone), and None for every other figure.

Source code in src/visualdynamics/report/plot_conventions.py

def figure_kind(block: Mapping[str, Any], objects: Mapping[str, Any],
                object_groups: Sequence[Mapping[str, Any]] | None = None
                ) -> str | None:
    """'asd' for a plot of a power spectral density (a measured PSD, a
    specification, or the two compared), 'srs' for a plot of a shock
    response spectrum (with its specification or alone), and None for
    every other figure."""
    from ..core.data import Psd, Srs
    from ..core.report import resolve_binding

    if block.get('kind') == 'prediction':
        # a response read against its prediction: an ASD when the
        # response is a PSD
        measured = objects.get(resolve_binding(
            block.get('measured'), objects, object_groups) or '')
        return 'asd' if isinstance(measured, Psd) else None
    if block.get('kind') != 'plot' or block.get('mode', 'curves') not in (
            'curves', 'pair'):
        return None
    source = objects.get(resolve_binding(block.get('source'), objects,
                                         object_groups) or '')
    if isinstance(source, Srs):
        return 'srs'
    if isinstance(source, Psd):
        return 'asd'
    return None

apply

apply(payload: dict[str, Any], figures: Sequence[tuple[Mapping[str, Any], str | None]], shared: bool = False) -> None

Put the page's figures under the conventions, in place.

payload is the page's data, which carries the line patterns and the end-label limit for the page to draw with. figures is each built block with its kind (figure_kind). An ASD or SRS figure on log ordinates goes to log frequency; a figure already there, or one on linear ordinates (a phase, a ratio in decibels), is left on its axes. With shared, the figures of each kind and unit open on the same whole decades (shared_decades).

Source code in src/visualdynamics/report/plot_conventions.py

def apply(payload: dict[str, Any],
          figures: Sequence[tuple[Mapping[str, Any], str | None]],
          shared: bool = False) -> None:
    """Put the page's figures under the conventions, in place.

    `payload` is the page's data, which carries the line patterns and
    the end-label limit for the page to draw with. `figures` is each
    built block with its kind (`figure_kind`). An ASD or SRS figure on
    log ordinates goes to log frequency; a figure already there, or one
    on linear ordinates (a phase, a ratio in decibels), is left on its
    axes. With `shared`, the figures of each kind and unit open on the
    same whole decades (`shared_decades`).
    """
    from . import _into_decades

    payload['line_patterns'] = [list(pattern) for pattern in LINE_PATTERNS]
    payload['end_labels'] = END_LABEL_CURVES
    for built, kind in figures:
        if kind not in KINDS:
            continue
        for cell in _cells(built):
            if cell.get('logy') and not cell.get('logx'):
                _into_decades(cell)
                if cell.get('marks'):
                    # a mode bookmark is a frequency too
                    cell['marks'] = [math.log10(f) for f in cell['marks']
                                     if f > 0]
    if shared:
        share_decades(figures)

share_decades

share_decades(figures: Sequence[tuple[Mapping[str, Any], str | None]]) -> None

Give every log-log figure of a kind the same opening window, in whole decades, in place.

Figures are grouped by kind and by their axis labels, which carry the quantity and the unit: an acceleration ASD in g²/Hz and a force ASD are different scales and are never forced onto one. Within a group the window is the union of each figure's opening window (_opening), widened outward to whole decades, at least one decade on each axis. A figure with nothing finite takes no part.

Source code in src/visualdynamics/report/plot_conventions.py

def share_decades(figures: Sequence[tuple[Mapping[str, Any], str | None]]
                  ) -> None:
    """Give every log-log figure of a kind the same opening window, in
    whole decades, in place.

    Figures are grouped by kind and by their axis labels, which carry
    the quantity and the unit: an acceleration ASD in g²/Hz and a force
    ASD are different scales and are never forced onto one. Within a
    group the window is the union of each figure's opening window
    (`_opening`), widened outward to whole decades, at least one decade
    on each axis. A figure with nothing finite takes no part.
    """
    groups: dict[tuple[str, str, str], list[dict[str, Any]]] = {}
    for built, kind in figures:
        if kind not in KINDS:
            continue
        for cell in _cells(built):
            if not (cell.get('logx') and cell.get('logy')):
                continue
            key = (kind, str(cell.get('xlabel', '')),
                   str(cell.get('ylabel', '')))
            groups.setdefault(key, []).append(cell)
    for cells in groups.values():
        windows = [window for window in map(_opening, cells)
                   if window is not None]
        if not windows:
            continue
        x0 = math.floor(min(w[0] for w in windows))
        x1 = math.ceil(max(w[1] for w in windows))
        y0 = math.floor(min(w[2] for w in windows))
        y1 = math.ceil(max(w[3] for w in windows))
        home = {'x0': float(x0), 'x1': float(max(x1, x0 + 1)),
                'y0': float(y0), 'y1': float(max(y1, y0 + 1))}
        for cell in cells:
            cell['home'] = dict(home)