Home /Modules/signal-processing/SLSignalProcessingSmoke.sleela

SLSignalProcessingSmoke.sleela

lib/signal-processing/tests/SLSignalProcessingSmoke.sleela · 97 line(s) · sleela

// @author Max Rupplin — MEARVK LLC — 2026
// Deterministic smoke checks for the SLeeLa source mirror.
#sleela 2.5
material;
import math;

class SLSignalProcessingSmoke {
  void check(String name, boolean passed) {
    if (passed) { print("PASS " + name); }
    else { print("FAIL " + name); }
  }

  void near(String name, double actual, double expected, double tolerance) {
    double difference = actual - expected;
    if (difference < 0.0) { difference = 0.0 - difference; }
    check(name, difference <= tolerance);
  }

  void main() {
    SLSignalProcessing api = new SLSignalProcessing();
    SLSignalSeries values = new SLSignalSeries();
    SLSignalScalar scalar = new SLSignalScalar();
    values.clear();
    values.append(1.0);
    values.append(2.0);
    values.append(3.0);

    int status = api.mean(values, scalar);
    check("mean status", status == SLSignalProcessingStatus.OK);
    near("mean", scalar.get(), 2.0, 0.000001);

    status = api.rms(values, scalar);
    check("rms status", status == SLSignalProcessingStatus.OK);
    near("rms", scalar.get(), math.sqrt(14.0 / 3.0), 0.000001);

    SLSignalSeries average = new SLSignalSeries();
    average.clear();
    status = api.movingAverage(values, 2, average);
    check("moving average status", status == SLSignalProcessingStatus.OK);
    check("moving average length", average.size() == 3);
    near("moving average tail", average.sampleAt(2).value, 2.5, 0.000001);

    SLSignalSeries kernel = new SLSignalSeries();
    SLSignalSeries convolved = new SLSignalSeries();
    kernel.clear();
    convolved.clear();
    kernel.append(1.0);
    kernel.append(1.0);
    status = api.convolve(values, kernel, convolved);
    check("convolution status", status == SLSignalProcessingStatus.OK);
    check("convolution length", convolved.size() == 4);
    near("convolution middle", convolved.sampleAt(1).value, 3.0, 0.000001);
    near("convolution last", convolved.sampleAt(3).value, 3.0, 0.000001);

    SLSignalSeries oneWindow = new SLSignalSeries();
    oneWindow.clear();
    status = api.hannWindow(1, oneWindow);
    check("single-point Hann status", status == SLSignalProcessingStatus.OK);
    near("single-point Hann value", oneWindow.sampleAt(0).value, 1.0, 0.000001);

    SLSignalComplexSeries impulse = new SLSignalComplexSeries();
    impulse.clear();
    impulse.appendParts(1.0, 0.0);
    impulse.appendParts(0.0, 0.0);
    impulse.appendParts(0.0, 0.0);
    impulse.appendParts(0.0, 0.0);

    SLSignalComplexSeries dft = new SLSignalComplexSeries();
    dft.clear();
    status = api.dft(impulse, dft, false);
    check("DFT status", status == SLSignalProcessingStatus.OK);
    near("DFT impulse bin", dft.sampleAt(3).re, 1.0, 0.000001);
    near("DFT impulse imaginary", dft.sampleAt(3).im, 0.0, 0.000001);

    SLSignalComplexSeries transformed = impulse.cloneSeries();
    status = api.fft(transformed, false);
    check("FFT status", status == SLSignalProcessingStatus.OK);
    near("FFT impulse bin", transformed.sampleAt(3).re, 1.0, 0.000001);
    near("FFT impulse imaginary", transformed.sampleAt(3).im, 0.0, 0.000001);

    SLSignalSeries magnitudes = new SLSignalSeries();
    magnitudes.clear();
    status = api.magnitudeSpectrum(impulse, magnitudes);
    check("magnitude status", status == SLSignalProcessingStatus.OK);
    check("magnitude bin count", magnitudes.size() == 3);
    near("magnitude impulse bin", magnitudes.sampleAt(2).value, 1.0, 0.000001);

    SLSignalComplexSeries invalidLength = new SLSignalComplexSeries();
    invalidLength.clear();
    invalidLength.appendParts(1.0, 0.0);
    invalidLength.appendParts(0.0, 0.0);
    invalidLength.appendParts(0.0, 0.0);
    status = api.fft(invalidLength, false);
    check("non-radix-two FFT rejection", status == SLSignalProcessingStatus.ERR_SIZE);
  }
}