m-thesis-introduction/simulations/airshower_beacon_simulation/lib/lib.py

# vim: fdm=indent ts=4
"""
Library for this simulation
"""
import numpy as np
from numpy.polynomial import Polynomial

from earsim import Antenna

""" Beacon utils """
def sine_beacon(f, t, t0=0, amplitude=1, baseline=0, phase=0):
    """
    Return a sine appropriate as a beacon
    """
    return amplitude * np.sin(2*np.pi*f*(t+t0) + phase) + baseline

def phase_mod(phase, low=np.pi):
    """
    Modulo phase such that it falls within the
    interval $[-low, 2\pi - low)$.
    """
    return (phase + low) % (2*np.pi) - low

def distance(x1, x2):
    """
    Calculate the Euclidean distance between two locations x1 and x2
    """

    assert type(x1) in [Antenna]
    x1 = np.array([x1.x, x1.y, x1.z])

    assert type(x2) in [Antenna]
    x2 = np.array([x2.x, x2.y, x2.z])

    return np.sqrt( np.sum( (x1-x2)**2 ) )

def geometry_time(dist, x2=None, c_light=3e8):
    if x2 is not None:
        dist = distance(dist, x2)

    return dist/c_light

def beacon_from(tx, rx, f, t=0, t0=0, c_light=3e8, radiate_rsq=True, amplitude=1,**kwargs):
    dist = distance(tx,rx)
    t0 = t0 + dist/c_light

    if radiate_rsq:
        if np.isclose(dist, 0):
            dist = 1
        amplitude *= 1/(dist**2)

    return sine_beacon(f, t, t0=t0, amplitude=amplitude,**kwargs)

""" Fourier """
def ft_corr_vectors(freqs, time):
    """
    Get the cosine and sine terms for freqs at time.

    Takes the outer product of freqs and time.
    """
    freqtime = np.outer(freqs, time)

    c_k = np.cos(2*np.pi*freqtime)
    s_k = np.sin(2*np.pi*freqtime)

    return c_k, s_k

def direct_fourier_transform(freqs, time, samplesets_iterable):
    """
    Determine the fourier transform of each sampleset in samplesets_iterable at freqs.

    The samplesets are expected to have the same time vector.

    Returns either a generator to return the fourier transform for each sampleset
    if samplesets_iterable is a generator
    or a numpy array.
    """

    c_k, s_k = ft_corr_vectors(freqs, time)

    if not hasattr(samplesets_iterable, '__len__') and hasattr(samplesets_iterable, '__iter__'):
        # samplesets_iterable is an iterator
        # return a generator containing (real, imag) amplitudes
        return ( (np.dot(c_k, samples), np.dot(s_k, samples)) for samples in samplesets_iterable )

    # Numpy array
    return np.dot(c_k, samplesets_iterable), np.dot(s_k, samplesets_iterable)

def phase_field_from_tx(x, y, tx, f_beacon, c_light=3e8, t0=0, wrap_phase=True, return_meshgrid=True):
    """
    """

    assert type(tx) in [Antenna]

    xs, ys = np.meshgrid(x, y, sparse=True)

    x_distances = (tx.x - xs)**2
    y_distances = (tx.y - ys)**2

    dist = np.sqrt( x_distances + y_distances )

    phase = (dist/c_light + t0) * f_beacon*2*np.pi
    if wrap_phase:
        phase = phase_mod(phase)

    if return_meshgrid:
        return phase, (xs, ys)
    else:
        return phase, (np.repeat(xs, len(ys), axis=0), np.repeat(ys, len(xs[0]), axis=1))


def find_beacon_in_traces(
        traces,
        t_trace,
        f_beacon_estimate = 50e6,
        frequency_fit = False,
        N_test_freqs = 5e2,
        f_beacon_estimate_band = 0.01,
        amp_cut = 0.8
        ):
    """
    f_beacon_band is inclusive

    traces is [trace, trace, trace, ..  ]
    """

    amplitudes = np.zeros(len(traces))
    phases = np.zeros(len(traces))
    frequencies = np.zeros(len(traces))

    if frequency_fit: # fit frequency
        test_freqs = f_beacon_estimate + f_beacon_estimate_band * np.linspace(-1, 1, int(N_test_freqs)+1)
        ft_amp_gen = direct_fourier_transform(test_freqs, t_trace, (x for x in traces))

        n_samples = len(t_trace)

        for i, ft_amp in enumerate(ft_amp_gen):
            real, imag = ft_amp
            amps = 1/n_samples * ( real**2 + imag**2)**0.5

            # find frequency peak and surrounding bins
            # that are valid for the parabola fit
            max_amp_idx = np.argmax(amps)
            max_amp = amps[max_amp_idx]

            if True:
                frequencies[i] = test_freqs[max_amp_idx]
                continue

            valid_mask = amps >= amp_cut*max_amp

            if True: # make sure not to use other peaks
                lower_mask = valid_mask[0:max_amp_idx]
                upper_mask = valid_mask[max_amp_idx:]

                if any(lower_mask):
                    lower_end = np.argmin(lower_mask[::-1])
                else:
                    lower_end = max_amp_idx

                if any(upper_mask):
                    upper_end = np.argmin(upper_mask)
                else:
                    upper_end = 0


                valid_mask[0:(max_amp_idx - lower_end)] = False
                valid_mask[(max_amp_idx + upper_end):] = False

            if all(~valid_mask):
                frequencies[i] = np.nan
                continue

            # fit Parabola
            parafit = Polynomial.fit(test_freqs[valid_mask], amps[valid_mask], 2)
            func = parafit.convert()

            # find frequency where derivative is 0
            deriv = func.deriv(1)
            freq = deriv.roots()[0]

            frequencies[i] = freq

    else: # no frequency finding
        frequencies[:] = f_beacon_estimate


    n_samples = len(t_trace)
    # evaluate fourier transform at freq for each trace
    for i, freq in enumerate(frequencies):
        if freq is np.nan:
            phases[i] = np.nan
            amplitudes[i] = np.nan
            continue

        real, imag = direct_fourier_transform(freq, t_trace, traces[i])

        phases[i] = np.arctan2(real, imag)
        amplitudes[i] = 2/n_samples * (real**2 + imag**2)**0.5

    return frequencies, phases, amplitudes

def coherence_sum_maxima(ref_x, y, k_step=1, k_start=0, k_end=None, periodic=True):
    """
    Use the maximum of a coherent sum to determine
    the best number of samples to move
    """
    N_samples = int( len(ref_x) )
    k_end = N_samples if k_end is None or k_end > max_k else k_end

    ks = np.arange(k_start, k_end, step=k_step)

    maxima = np.empty(len(ks))

    if periodic is False:
        # prepend zeros
        N_zeros = N_samples
        preshift = 0 # only required for testing purposes

        ref_x = np.pad(ref_x, (N_zeros-0,0), 'constant')
        y = np.pad(y, (N_zeros-preshift,preshift), 'constant')

    for i,k in enumerate(ks, 0):
        augmented_y = np.roll(y, k)
        maxima[i] = max(ref_x + augmented_y)

    return ks, maxima