Weighted High Order FTS

This commit is contained in:
Petrônio Cândido 2018-11-12 21:46:14 -02:00
parent 91a9fa04ae
commit 75e69a1ae1
3 changed files with 86 additions and 4 deletions

View File

@ -19,6 +19,7 @@ from mpl_toolkits.mplot3d import Axes3D
from pyFTS.probabilistic import ProbabilityDistribution
from pyFTS.common import Transformations
from pyFTS.models import song, chen, yu, ismailefendi, sadaei, hofts, pwfts, ifts, cheng, hwang
from pyFTS.models.multivariate import mvfts, wmvfts, cmvfts
from pyFTS.models.ensemble import ensemble
from pyFTS.benchmarks import Measures, naive, arima, ResidualAnalysis, quantreg, knn
from pyFTS.benchmarks import Util as bUtil
@ -57,10 +58,16 @@ def get_benchmark_point_methods():
def get_point_methods():
"""Return all FTS methods for point forecasting"""
return [song.ConventionalFTS, chen.ConventionalFTS, yu.WeightedFTS, ismailefendi.ImprovedWeightedFTS,
cheng.TrendWeightedFTS, sadaei.ExponentialyWeightedFTS, hofts.HighOrderFTS, hwang.HighOrderFTS,
cheng.TrendWeightedFTS, sadaei.ExponentialyWeightedFTS,
hofts.HighOrderFTS, hofts.WeightedHighOrderFTS, hwang.HighOrderFTS,
pwfts.ProbabilisticWeightedFTS]
def get_point_multivariate_methods():
"""Return all multivariate FTS methods por point forecasting"""
return [mvfts.MVFTS, wmvfts.WeightedMVFTS, cmvfts.ClusteredMVFTS]
def get_benchmark_interval_methods():
"""Return all non FTS methods for point_to_interval forecasting"""
return [ arima.ARIMA, quantreg.QuantileRegression]

View File

@ -36,6 +36,48 @@ class HighOrderFLRG(flrg.FLRG):
return len(self.RHS)
class WeightedHighOrderFLRG(flrg.FLRG):
"""Weighted High Order Fuzzy Logical Relationship Group"""
def __init__(self, order, **kwargs):
super(WeightedHighOrderFLRG, self).__init__(order, **kwargs)
self.LHS = []
self.RHS = {}
self.count = 0.0
self.strlhs = ""
self.w = None
def append_rhs(self, fset, **kwargs):
if fset not in self.RHS:
self.RHS[fset] = 1.0
else:
self.RHS[fset] += 1.0
self.count += 1.0
def append_lhs(self, c):
self.LHS.append(c)
def weights(self):
if self.w is None:
self.w = np.array([self.RHS[c] / self.count for c in self.RHS.keys()])
return self.w
def get_midpoint(self, sets):
mp = np.array([sets[c].centroid for c in self.RHS.keys()])
return mp.dot(self.weights())
def __str__(self):
_str = ""
for k in self.RHS.keys():
_str += ", " if len(_str) > 0 else ""
_str += k + " (" + str(round(self.RHS[k] / self.count, 3)) + ")"
return self.get_key() + " -> " + _str
def __len__(self):
return len(self.RHS)
class HighOrderFTS(fts.FTS):
"""Conventional High Order Fuzzy Time Series"""
def __init__(self, **kwargs):
@ -145,7 +187,6 @@ class HighOrderFTS(fts.FTS):
else:
self.generate_flrg_fuzzyfied(data)
def forecast(self, ndata, **kwargs):
explain = kwargs.get('explain', False)
@ -181,7 +222,6 @@ class HighOrderFTS(fts.FTS):
if explain:
print("\t {} -> {} (Naïve)\t Midpoint: {}\n".format(str(flrg.LHS), flrg.LHS[-1],
mp))
else:
flrg = self.flrgs[flrg.get_key()]
mp = flrg.get_midpoint(self.partitioner.sets)
@ -197,3 +237,38 @@ class HighOrderFTS(fts.FTS):
print("Deffuzyfied value: {} \n".format(final))
return ret
class WeightedHighOrderFTS(HighOrderFTS):
"""Weighted High Order Fuzzy Time Series"""
def __init__(self, **kwargs):
super(WeightedHighOrderFTS, self).__init__(**kwargs)
self.name = "Weighted High Order FTS"
self.shortname = "WHOFTS"
def generate_lhs_flrg_fuzzyfied(self, sample, explain=False):
lags = {}
flrgs = []
for ct, o in enumerate(self.lags):
lags[ct] = sample[o-1]
if explain:
print("\t (Lag {}) {} -> {} \n".format(o, sample[o-1], lhs))
root = tree.FLRGTreeNode(None)
tree.build_tree_without_order(root, lags, 0)
# Trace the possible paths
for p in root.paths():
flrg = WeightedHighOrderFLRG(self.order)
path = list(reversed(list(filter(None.__ne__, p))))
for lhs in path:
flrg.append_lhs(lhs)
flrgs.append(flrg)
return flrgs

View File

@ -25,7 +25,7 @@ p = Grid.GridPartitioner(data=dataset, npart=20)
print(p)
model = hofts.HighOrderFTS(partitioner=p, order=2)
model = hofts.WeightedHighOrderFTS(partitioner=p, order=2)
model.fit(dataset) #[22, 22, 23, 23, 24])