- Several bugfixes in benchmarks methods and optimizations

This commit is contained in:
Petrônio Cândido de Lima e Silva 2017-05-08 17:50:35 -03:00
parent 38a1410b16
commit 8df4f9c749
3 changed files with 21 additions and 10 deletions

View File

@ -24,11 +24,11 @@ class ARIMA(fts.FTS):
self.benchmark_only = True
self.min_order = 1
def train(self, data, sets, order=(2,1,1), parameters=None):
def train(self, data, sets, order, parameters=None):
self.p = order[0]
self.d = order[1]
self.q = order[2]
self.order = max([self.p, self.d, self.q])
self.order = self.p + self.q
self.shortname = "ARIMA(" + str(self.p) + "," + str(self.d) + "," + str(self.q) + ")"
old_fit = self.model_fit
@ -51,13 +51,19 @@ class ARIMA(fts.FTS):
ret = []
if self.d == 0:
ar = np.array([self.ar(ndata[k - self.p: k]) for k in np.arange(self.p, l)])
else:
ar = np.array([ndata[k] + self.ar(ndata[k - self.p: k]) for k in np.arange(self.p, l)])
residuals = np.array([ar[k - self.p] - ndata[k] for k in np.arange(self.p, l)])
if self.q > 0:
residuals = np.array([ndata[k] - ar[k - self.p] for k in np.arange(self.p, l)])
ma = np.array([self.ma(residuals[k - self.q: k]) for k in np.arange(self.q, len(ar) + 1)])
ma = np.array([self.ma(residuals[k - self.q: k]) for k in np.arange(self.q, len(residuals))])
ret = ar + ma
ret = ar[self.q:] + ma
else:
ret = ar
ret = self.doInverseTransformations(ret, params=[data[self.order - 1:]])

View File

@ -101,7 +101,7 @@ def point_sliding_window(data, windowsize, train=0.8, models=None, partitioners=
quantreg.QuantileRegression, quantreg.QuantileRegression]
if benchmark_models_parameters is None:
benchmark_models_parameters = [None, (1, 0, 1), (1, 1, 1), (2, 1, 1), (2, 1, 2), 1, 2]
benchmark_models_parameters = [None, (1, 0, 0), (1, 0, 1), (2, 0, 1), (2, 0, 2), 1, 2]
cluster = dispy.JobCluster(run_point, nodes=nodes) #, depends=dependencies)

View File

@ -64,9 +64,14 @@ from pyFTS.benchmarks import arima, quantreg
#Util.plot_dataframe_point("experiments/taiex_point_sintetic.csv","experiments/taiex_point_analitic.csv",11)
#tmp = arima.ARIMA("")
#tmp.train(taiex[:1600],(2,1,1))
#teste = tmp.forecast(taiex[1600:2000])
#tmp.train(taiex[:1600], None, order=(1,0,1))
#teste = tmp.forecast(taiex[1600:1610])
#tmp = quantreg.QuantileRegression("")
#tmp.train(taiex[:1600], None, order=1)
#teste = tmp.forecast(taiex[1600:1610])
#print(taiex[1600:1610])
#print(teste)
#bchmk.teste(taiex,['192.168.0.109', '192.168.0.101'])
@ -75,7 +80,7 @@ bchmk.point_sliding_window(taiex,2000,train=0.8, #models=[yu.WeightedFTS], # #
partitioners=[Grid.GridPartitioner], #Entropy.EntropyPartitioner], # FCM.FCMPartitioner, ],
partitions= np.arange(10,200,step=10), #transformation=diff,
dump=True, save=True, file="experiments/taiex_point_analytic.csv",
nodes=['192.168.0.102', '192.168.0.109']) #, depends=[hofts, ifts])
nodes=['192.168.0.102', '192.168.0.109', '192.168.0.106']) #, depends=[hofts, ifts])
#bchmk.testa(taiex,[10,20],partitioners=[Grid.GridPartitioner], nodes=['192.168.0.109', '192.168.0.101'])