def process_data(self, data): y = data.rangefinder[0].readings self.mean_readings = weighted_average(self.mean_readings, self.num_samples, y) yn = y - self.mean_readings T = outer(yn, yn) self.cov_readings = weighted_average(self.cov_readings, self.num_samples, T) self.num_samples += 1
def process_data(self, data): y = data.sensels self.mean_sensels = weighted_average(self.mean_sensels, self.num_samples, y) yn = y - self.mean_sensels T = outer(yn, yn) self.cov_sensels = weighted_average(self.cov_sensels, self.num_samples, T) self.num_samples += 1
def process_data(self, data): y = data.sensels y_dot = data.sensels_dot u = data.commands y_n = y - self.y_mean T = outer(u, outer(y_n, y_dot)) self.T = weighted_average(self.T, self.num_samples, T) self.y_mean = weighted_average(self.y_mean, self.num_samples, y) self.num_samples += 1
def process_data(self, data): y = data.optics[0].luminance # Update mean estimate self.mean_luminance = weighted_average(self.mean_luminance, self.num_samples, y) # Subtract the mean yn = y - self.mean_luminance # Compute the exterior product of normalized luminance T = outer(yn, yn) # Update covariance estimate self.cov_luminance = weighted_average(self.cov_luminance, self.num_samples, T) # Keep track of how many we integrated so far self.num_samples += 1
def process_data(self, data): y = data.sensels # y_dot = numpy.sign(data.sensels_dot) y_dot = data.sensels_dot u = data.commands self.y_mean = weighted_average(self.y_mean, self.num_samples, y.mean()) y_n = y - self.y_mean T = outer(u, outer(y_n, y_dot)) # if self.num_samples > 50: # delay execution self.T = weighted_average(self.T, self.num_samples, T) self.num_samples += 1
def process_data(self, data): y = data.sensels y_dot = data.sensels_dot u = data.commands T = outer(u, outer(y, y_dot)) self.T = weighted_average(self.T, self.num_samples, T) self.num_samples += 1