def next(self): self.currfile += 1 if self.currfile > self.n_files: raise StopIteration else: return (self.filelist[self.currfile - 1], read_timeseries( os.path.join(self.fullpath, self.filelist[self.currfile - 1])))
def detect_SMA(path, window, threshold): series = read_timeseries(path) values = np.array(map(lambda x: x[1], series)) s_ma = moving_avg(values, window) anomalies = [] times = list() values = list() for i in range(window + 1, len(series) - window): dist = abs(float(s_ma[i] - series[i][1])) #if dist >= values.ptp()*threshold: anomalies.append((i, series[i][0], dist)) times.append(series[i][0]) values.append(dist) filtered_anomalies = naive.get_anomalies_from_series(times, values, 3) #filtered_anomalies = naive.get_anomalies_from_series(map(lambda x:x[1:],anomalies),3) return filtered_anomalies
def detect_SMA(path, window, threshold): series = read_timeseries(path) values = np.array(map(lambda x:x[1],series)) s_ma = moving_avg(values,window) anomalies = [] times = list() values= list() for i in range(window+1,len(series)-window): dist = abs(float(s_ma[i]-series[i][1])) #if dist >= values.ptp()*threshold: anomalies.append((i, series[i][0], dist)) times.append(series[i][0]) values.append(dist) filtered_anomalies = naive.get_anomalies_from_series(times, values, 3) #filtered_anomalies = naive.get_anomalies_from_series(map(lambda x:x[1:],anomalies),3) return filtered_anomalies
def next(self): self.currfile+= 1 if self.currfile > self.n_files: raise StopIteration else: return (self.filelist[self.currfile-1], read_timeseries(os.path.join(self.fullpath, self.filelist[self.currfile-1])))