endaq.batch

class endaq.batch.GetDataBuilder(*, preferred_chs=[], accel_highpass_cutoff, accel_start_time=None, accel_end_time=None, accel_start_margin=None, accel_end_margin=None, accel_integral_tukey_percent=0, accel_integral_zero='start')

The main interface for calculations in endaq.batch.

This object has two types of functions:

  • configuration functions - these determine what calculations will be performed on IDE recordings, and pass in any requisite parameters for said calculations. This includes the following functions:

  • execution functions - these functions take recording files as parameters, perform the configured calculations on the data therein, and return the calculated data as a OutputStruct object that wraps pandas objects.

    This includes the functions _get_data() & aggregate_data(), which operates on one & multiple file(s), respectively.

A typical use case will look something like this:

filenames = [...]

calc_output = (
    GetDataBuilder(accel_highpass_cutoff=1)
    .add_psd(freq_bin_width=1)
    .add_pvss(init_freq=1, bins_per_octave=12)
    .add_pvss_halfsine_envelope()
    .add_metrics()
    .add_peaks(margin_len=100)
    .add_vc_curves(init_freq=1, bins_per_octave=3)
    .aggregate_data(filenames)
)
file_data = calc_output.dataframes
Parameters:
  • preferred_chs – a sequence of channels; each channel listed is prioritized over others of the same type of physical measurement (e.g., acceleration, temperature, pressure, etc.)

  • accel_highpass_cutoff – the cutoff frequency used when pre-filtering acceleration data

  • accel_start_time – the relative timestamp before which to reject recording data; cannot be used in conjunction with accel_start_margin

  • accel_end_time – the relative timestamp after which to reject recording data; cannot be used in conjunction with accel_end_margin

  • accel_start_margin – the number of samples before which to reject recording data; cannot be used in conjunction with accel_start_time

  • accel_end_margin – the number of samples after which to reject recording data; cannot be used in conjunction with accel_end_time

  • accel_integral_tukey_percent – the alpha parameter of a Tukey window applied to the acceleration before integrating into velocity & displacement; see the tukey_percent parameter in endaq.calc.integrate.integrals() for details

  • accel_integral_zero – the output quantity driven to zero when integrating the acceleration into velocity & displacement; see the zero parameter in endaq.calc.integrate.integrals() for details

add_metrics(include=[], exclude=[])

Add broad channel metrics to the calculation queue.

The output units for each metric are listed below:

  • RMS Acceleration: G

  • RMS Velocity: mmsec

  • RMS Displacement: mm

  • Peak Absolute Acceleration: G

  • Peak Pseudo Velocity Shock Spectrum: mmsec

  • GPS Position: degrees

  • GPS Speed: kmhr

  • RMS Angular Velocity: degreessec

  • RMS Microphone: Pascals

  • Average Temperature: C

  • Average Pressure: Pascals

  • Average Relative Humidity: %

where G is the acceleration of gravity (1G9.80665msec2)

add_peaks(margin_len=1000)

Add windows about the acceleration’s peak value to the calculation queue.

calculation output units: G, where G is the acceleration of gravity (1G9.80665msec2)

Parameters:

margin_len (int) – the number of samples on each side of a peak to include in the windows

add_psd(freq_bin_width=None, freq_start_octave=None, bins_per_octave=None, window=None)

Add the acceleration PSD to the calculation queue.

calculation output units: G2Hz, where G is the acceleration of gravity (1G9.80665msec2)

Parameters:
  • freq_bin_width (float | None) – the desired spacing between adjacent PSD samples; a default is provided only if bins_per_octave is used, otherwise this parameter is required

  • freq_start_octave (float | None) – the first frequency to use in octave-spacing; this is only used if bins_per_octave is set

  • bins_per_octave (float | None) – the number of frequency bins per octave in a log-spaced PSD; if not set, the PSD will be linearly-spaced as specified by freq_bin_width

  • window (str | None) – the window type used in the PSD calculation; see the documentation for scipy.signal.welch for details

add_pvss(init_freq=1.0, bins_per_octave=3.0)

Add the acceleration PVSS (Pseudo Velocity Shock Spectrum) to the calculation queue.

calculation output units: mmsec

Parameters:
  • init_freq (float) – the first frequency sample in the spectrum

  • bins_per_octave (float) – the number of samples per frequency octave

add_pvss_halfsine_envelope(tstart=None, tstop=None, dt=None, tpulse=None)

Add the half-sine envelope for the acceleration’s PVSS (Pseudo Velocity Shock Spectrum) to the calculation queue.

calculation output units: mmsec

add_vc_curves(init_freq=1.0, bins_per_octave=3.0)

Add Vibration Criteria (VC) Curves to the calculation queue.

calculation output units: μmsec

Parameters:
  • init_freq (float) – the first frequency

  • bins_per_octave (float) – the number of samples per frequency octave

aggregate_data(filenames)

Compile configured data from the given files into a dataframe.

Parameters:

filenames – a sequence of paths of recording files to process

class endaq.batch.core.OutputStruct(data)

A data wrapper class with methods for common export operations.

Objects of this class are generated by GetDataBuilder.aggregate_data(). This class is not intended be instantiated manually.

dataframes
to_csv_folder(folder_path)

Write data to a folder as CSV’s.

Parameters:

folder_path – the output directory path for .CSV files

to_html_plots(folder_path=None, show=False, theme='endaq')

Generate plots in HTML.

Parameters:
  • folder_path – The output directory for saving .HTML plots. If None (default), plots are not saved.

  • show (bool) – Whether to open plots after generation. Defaults to False.

  • theme (Literal[None, 'endaq', 'endaq_light', 'endaq_arial', 'endaq_light_arial']) – The enDAQ plotly theme to use; see endaq.plot.utilities.set_theme() for details on the supported options. Defaults to “endaq”. If None, the default Plotly theme is used.