import numpy as np def complex_exp(freq, samp_rate, periods, phase_deg=0): t_max = 1.0 * samp_rate / freq * periods t = np.arange(t_max) fac = t / samp_rate * freq phase = 1j * phase_deg/360*2*np.pi ret = np.exp(phase + 1j * 2 * np.pi * fac - 1j * np.pi / 2, dtype=np.complex128) return ret ret = complex_exp(10,40,2) def gen_file(frequency_0, frequency_1, coefs = [], samp_rate = 4000000, path = "./np_twotone", count = 1): tone_0 = complex_exp(frequency_0, samp_rate, samp_rate/frequency_1 * count) tone_1 = complex_exp(frequency_1, samp_rate, samp_rate/frequency_0 * count) min_len = min(tone_0.shape[0], tone_1.shape[0]) two_tone = (tone_0[0:min_len] + tone_1[0:min_len]) res = two_tone for idx, coef in enumerate(coefs): res += two_tone * np.abs(two_tone)**(idx+1) * coef #+1 because first correction term is squared res = res / np.max(res) * 0.9 res = res.astype(np.complex64) res.tofile(path) a_load = np.fromfile(path, dtype=np.complex64) assert(np.isclose(a_load, res).all()), "Inconsistent stored file" return path def gen_file_d(frequency_0, frequency_1, x1 = 0, x2 = 0, x3 = 0, x4 = 0, samp_rate = 4000000, path = "./np_twotone", count = 1): tone_0 = complex_exp(frequency_0, samp_rate, samp_rate/frequency_1 * count) tone_1 = complex_exp(frequency_1, samp_rate, samp_rate/frequency_0 * count) min_len = min(tone_0.shape[0], tone_1.shape[0]) two_tone = (tone_0[0:min_len] + tone_1[0:min_len]) two_tone_1 = np.gradient(two_tone) two_tone_2 = np.gradient(two_tone_1) two_tone_3 = np.gradient(two_tone_2) two_tone_4 = np.gradient(two_tone_3) two_tone = two_tone \ + two_tone_1 * x1 \ + two_tone_2 * x2 \ + two_tone_3 * x3 \ + two_tone_4 * x4 two_tone = two_tone / np.max(two_tone) * 0.9 two_tone = two_tone.astype(np.complex64) two_tone.tofile(path) a_load = np.fromfile(path, dtype=np.complex64) assert(np.isclose(a_load, two_tone).all()), "Inconsistent stored file" return path def gen_file_i(frequency_0, frequency_1, x1 = 0, x2 = 0, x3 = 0, x4 = 0, samp_rate = 4000000, path = "./np_twotone", count = 1): if frequency_0 > frequency_1: f = frequency_0 frequency_1 = frequency_0 frequency_0 = f tone_0 = complex_exp(frequency_0, samp_rate, samp_rate/frequency_1 * count) tone_1 = complex_exp(frequency_1, samp_rate, samp_rate/frequency_0 * count) df = frequency_1 - frequency_0 tone_0_0 = complex_exp(frequency_0 - df, samp_rate, samp_rate/frequency_1 * count, pahse = x2) tone_0_1 = complex_exp(frequency_1 + df, samp_rate, samp_rate/frequency_1 * count, pahse = x2) min_len = min(tone_0.shape[0], tone_1.shape[0]) two_tone = (tone_0[0:min_len] + tone_1[0:min_len]) two_tone_1 = np.gradient(two_tone) two_tone_2 = np.gradient(two_tone_1) two_tone_3 = np.gradient(two_tone_2) two_tone_4 = np.gradient(two_tone_3) two_tone = two_tone \ + two_tone_1 * x1 \ + two_tone_2 * x2 \ + two_tone_3 * x3 \ + two_tone_4 * x4 two_tone = two_tone / np.max(two_tone) * 0.9 two_tone = two_tone.astype(np.complex64) two_tone.tofile(path) a_load = np.fromfile(path, dtype=np.complex64) assert(np.isclose(a_load, two_tone).all()), "Inconsistent stored file" return path def gen_sin(samples, oversampling, phi): t = np.arange(samples, dtype=np.float) sig = np.sin(((2*np.pi)/oversampling) * t - np.pi*phi/180.) return sig