dataio.load_other_data(other_file_location, datatype = 'f')
data, times = dataio.load_traffic_data(files[0])
sd, st = strip_data_days(data, times, ["Sun"])
sun_data, sun_times = strip_data_days(data, times, ["Sun"])
sd, st = strip_data_std(sun_data, sun_times, 4)
sun_avg = np.average(sd, axis=0)
sun_res = sun_data - sun_avg
sun_1d = np.reshape(sun_data, -1)
sun_avg_1d = np.resize(sun_avg, sun_data.size)
ps = arima(sun_1d)
mase = analysis.mase(sun_1d, sun_avg_1d)
mape = analysis.mape(sun_1d, sun_avg_1d)
print "Mean Mase:", mase
print "Mean Mape:", mape
mase = analysis.mase(sun_1d, ps)
mape = analysis.mape(sun_1d, ps)
print "SARIMA Mase:", mase
print "SARIMA Mape:", mape
"""
xdata = range(24)
mpl.subplot(111)
for p in weekend_res[0:300]:
mpl.plot(xdata, p, 'k', alpha = 0.2)
resEventData.append(r[9:17])
#Convert to list
tmp = np.array(resEventData)
resEventData = tmp.tolist()
resEventData = pybb.data.stripZero(resEventData, threshold=6)
#Train HMM
m, mdata, mout = khmm.train(resEventData, 1, actLength, \
iterations = 20, outliers = False, \
clustering = "kmeans++",
verbose = False)
mase = analysis.mase(sunData1d, fData1d)
mape = analysis.mape(sunData1d, fData1d)
print "Mean Mase:", mase
print "Mean Mape:", mape
#Setup forecast
model = []
model.append(gaussian.Gaussian(resData.mean(), resData.std()))
#model.append(gaussian.Gaussian(0, resData.std()))
model.append(m[0])
bf = bayesianforecast.BayesianForecast(model)
bf.setStd(0, resData.std())
bf.setStd(1, resData.std())
bfData1d, probs, windowLens, models = bf.windowForecast(resData1d, \
dataio.load_other_data(other_file_location, datatype = 'f')
data, times = dataio.load_traffic_data(files[0])
sd, st = strip_data_days(data, times, ["Sun"])
sun_data, sun_times = strip_data_days(data, times, ["Sun"])
sd, st = strip_data_std(sun_data, sun_times, 4)
sun_avg = np.average(sd, axis = 0)
sun_res = sun_data - sun_avg
sun_1d = np.reshape(sun_data, -1)
sun_avg_1d = np.resize(sun_avg, sun_data.size)
ps = arima(sun_1d)
mase = analysis.mase(sun_1d, sun_avg_1d)
mape = analysis.mape(sun_1d, sun_avg_1d)
print "Mean Mase:", mase
print "Mean Mape:", mape
mase = analysis.mase(sun_1d, ps)
mape = analysis.mape(sun_1d, ps)
print "SARIMA Mase:", mase
print "SARIMA Mape:", mape
"""
xdata = range(24)
mpl.subplot(111)
st = time.time()
ftd = R.r('ftd = forecast.Arima(tmod)')
print "Forecast Time = " + str(time.time() - st)
fTData1d = np.array(ftd[8])
fTData = np.array(fTData1d)
fTData = np.reshape(fTData, (len(ftd[8]) / periodLength, periodLength))
tres1d = np.array(ftd[9])
tres = np.array(tres1d)
tres = np.reshape(tres, (len(ftd[9]) / periodLength, periodLength))
print "Arima Training data Std:", mod[1][0]**0.5
mase = analysis.mase(tdata1d, fTData1d)
mape = analysis.mape(tdata1d, fTData1d)
print "Testing Mase:", mase
print "Testing Mape:", mape
#Make residual events
resEventData = []
for a in tacts:
resEventData.append(tres[a[0]][a[1]:a[1] + actLength])
resEventData = np.array(resEventData)
resEventData = resEventData.tolist()
#Train HMM
m, mdata, mout = khmm.train(resEventData, 1, actLength, \
iterations = 20, outliers = False, \
#Convert to list
tmp = np.array(resEventData)
resEventData = tmp.tolist()
resEventData = pybb.data.stripZero(resEventData, threshold = 6)
#Train HMM
m, mdata, mout = khmm.train(resEventData, 1, actLength, \
iterations = 20, outliers = False, \
clustering = "kmeans++",
verbose = False)
mase = analysis.mase(sunData1d, fData1d)
mape = analysis.mape(sunData1d, fData1d)
print "Mean Mase:", mase
print "Mean Mape:", mape
#Setup forecast
model = []
model.append(gaussian.Gaussian(resData.mean(), resData.std()))
#model.append(gaussian.Gaussian(0, resData.std()))
model.append(m[0])
bf = bayesianforecast.BayesianForecast(model)
bf.setStd(0, resData.std())
bf.setStd(1, resData.std())
st = time.time()
ftd = R.r('ftd = forecast.Arima(tmod)')
print "Forecast Time = " + str(time.time() - st)
fTData1d = np.array(ftd[8])
fTData = np.array(fTData1d)
fTData = np.reshape(fTData, (len(ftd[8])/periodLength, periodLength))
tres1d = np.array(ftd[9])
tres = np.array(tres1d)
tres = np.reshape(tres, (len(ftd[9])/periodLength, periodLength))
print "Arima Training data Std:", mod[1][0]**0.5
mase = analysis.mase(tdata1d, fTData1d)
mape = analysis.mape(tdata1d, fTData1d)
print "Testing Mase:", mase
print "Testing Mape:", mape
#Make residual events
resEventData = []
for a in tacts:
resEventData.append(tres[a[0]][a[1]:a[1] + actLength])
resEventData = np.array(resEventData)
resEventData = resEventData.tolist()
#Train HMM
m, mdata, mout = khmm.train(resEventData, 1, actLength, \
iterations = 20, outliers = False, \