sam/pyFTS
1
0
Fork 0
Code Issues 3 Pull Requests Actions Packages Projects Releases Wiki Activity

Bugfixes and improvements on benchmarking methods

Browse Source
This commit is contained in:
Petrônio Cândido 2019-05-30 11:36:55 -03:00
parent 99c36fc5e9
commit 5f667c8680
4 changed files with 18 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
pyFTS/benchmarks/quantreg.py
View File

@ -97,7 +97,7 @@ class QuantileRegression(fts.FTS):
def forecast_ahead_interval(self, ndata, steps, **kwargs): def forecast_ahead_interval(self, ndata, steps, **kwargs):
smoothing = kwargs.get("smoothing", 0.1) smoothing = kwargs.get("smoothing", 0.01)
l = len(ndata) l = len(ndata)
@ -113,7 +113,7 @@ class QuantileRegression(fts.FTS):
tmpk = k-self.order tmpk = k-self.order
ret.append([intl[0]*(1 + (tmpk*smoothing)), intl[1]*(1 + (tmpk*smoothing))]) ret.append([intl[0]*(1 - (tmpk*smoothing)), intl[1]*(1 + (tmpk*smoothing))])
return ret[-steps:] return ret[-steps:]
@ -139,7 +139,7 @@ class QuantileRegression(fts.FTS):
return ret return ret
def forecast_ahead_distribution(self, ndata, steps, **kwargs): def forecast_ahead_distribution(self, ndata, steps, **kwargs):
smoothing = kwargs.get("smoothing", 0.1) smoothing = kwargs.get("smoothing", 0.01)
l = len(ndata) l = len(ndata)
@ -158,7 +158,7 @@ class QuantileRegression(fts.FTS):
for qt in self.dist_qt: for qt in self.dist_qt:
intl1 = self.point_to_interval(nmeans[k - self.order: k], qt[0], qt[1]) intl1 = self.point_to_interval(nmeans[k - self.order: k], qt[0], qt[1])
tmpk = k - self.order tmpk = k - self.order
intl2 = [intl1[0] * (1 + (tmpk * smoothing)), intl1[1] * (1 + (tmpk * smoothing))] intl2 = [intl1[0] * (1 - (tmpk * smoothing)), intl1[1] * (1 + (tmpk * smoothing))]
intervals.append(intl2) intervals.append(intl2)
dist.append_interval(intervals) dist.append_interval(intervals)

6
pyFTS/common/Util.py
View File

@ -177,8 +177,6 @@ def plot_distribution2(probabilitydist, data, **kwargs):
import matplotlib.colorbar as cbar import matplotlib.colorbar as cbar
import matplotlib.cm as cm import matplotlib.cm as cm
order = kwargs.get('order', 1)
ax = kwargs.get('ax',None) ax = kwargs.get('ax',None)
if ax is None: if ax is None:
fig, ax = plt.subplots(nrows=1, ncols=1, figsize=[15, 5]) fig, ax = plt.subplots(nrows=1, ncols=1, figsize=[15, 5])
@ -188,7 +186,7 @@ def plot_distribution2(probabilitydist, data, **kwargs):
cmap = kwargs.get('cmap','Blues') cmap = kwargs.get('cmap','Blues')
cmap = plt.get_cmap(cmap) cmap = plt.get_cmap(cmap)
start_at = kwargs.get('start_at',0) + order - 1 start_at = kwargs.get('start_at',0)
x = [k + start_at for k in range(l + 1)] x = [k + start_at for k in range(l + 1)]
@ -275,7 +273,7 @@ def plot_interval2(intervals, data, **kwargs):
if ax is None: if ax is None:
fig, ax = plt.subplots(nrows=1, ncols=1, figsize=[15, 5]) fig, ax = plt.subplots(nrows=1, ncols=1, figsize=[15, 5])
for k in np.arange(0, start_at-1): for k in np.arange(0, start_at):
intervals.insert(0, [None,None]) intervals.insert(0, [None,None])
intervals.insert(start_at, [data[start_at], data[start_at]]) intervals.insert(start_at, [data[start_at], data[start_at]])

1
pyFTS/models/ensemble/ensemble.py
View File

@ -303,6 +303,7 @@ class SimpleEnsembleFTS(EnsembleFTS):
"""Possible variations of number of partitions on internal models""" """Possible variations of number of partitions on internal models"""
self.orders = kwargs.get('orders', [1,2,3]) self.orders = kwargs.get('orders', [1,2,3])
"""Possible variations of order on internal models""" """Possible variations of order on internal models"""
self.uod_clip = False
def train(self, data, **kwargs): def train(self, data, **kwargs):
for k in self.partitions: for k in self.partitions:

17
pyFTS/tests/general.py
View File

@ -12,6 +12,7 @@ from pyFTS.common import Util as cUtil, FuzzySet
from pyFTS.partitioners import Grid, Entropy, Util as pUtil, Simple from pyFTS.partitioners import Grid, Entropy, Util as pUtil, Simple
from pyFTS.benchmarks import benchmarks as bchmk, Measures from pyFTS.benchmarks import benchmarks as bchmk, Measures
from pyFTS.models import chen, yu, cheng, ismailefendi, hofts, pwfts, tsaur, song, sadaei from pyFTS.models import chen, yu, cheng, ismailefendi, hofts, pwfts, tsaur, song, sadaei
from pyFTS.models.ensemble import ensemble
from pyFTS.common import Transformations, Membership from pyFTS.common import Transformations, Membership
from pyFTS.benchmarks import arima, quantreg, BSTS from pyFTS.benchmarks import arima, quantreg, BSTS
from pyFTS.fcm import fts, common, GA from pyFTS.fcm import fts, common, GA
@ -20,17 +21,19 @@ from pyFTS.data import Enrollments, TAIEX
data = TAIEX.get_data() data = TAIEX.get_data()
train = data[:500] train = data[:800]
test = data[500:1000] test = data[800:1000]
model = quantreg.QuantileRegression(order=1, dist=True) model = ensemble.SimpleEnsembleFTS(fts_method=hofts.HighOrderFTS)
#model = quantreg.QuantileRegression(order=2, dist=True)
#model = arima.ARIMA(order = (2,0,0))
model.fit(train) model.fit(train)
horizon=5 horizon=5
#points = model.predict(test[:10], type='point', steps_ahead=horizon) #points = model.predict(test[:10], type='point', steps_ahead=horizon)
intervals = model.predict(test[:10], type='interval', alpha=.05, smoothing=0.01, steps_ahead=horizon) intervals = model.predict(test[:10], type='interval', alpha=.05, steps_ahead=horizon)
print(test[:10]) print(test[:10])
print(intervals) print(intervals)
distributions = model.predict(test[:10], type='distribution', steps_ahead=horizon, num_bins=100) distributions = model.predict(test[:10], type='distribution', steps_ahead=horizon, num_bins=100)
@ -39,8 +42,10 @@ distributions = model.predict(test[:10], type='distribution', steps_ahead=horizo
fig, ax = plt.subplots(nrows=1, ncols=1,figsize=[15,5]) fig, ax = plt.subplots(nrows=1, ncols=1,figsize=[15,5])
ax.plot(test[:10], label='Original',color='black') ax.plot(test[:10], label='Original',color='black')
cUtil.plot_interval2(intervals, test[:10], start_at=3, ax=ax) cUtil.plot_interval2(intervals, test[:10], start_at=model.order, ax=ax)
cUtil.plot_distribution2(distributions, test[:10], start_at=2, ax=ax, cmap="Blues") cUtil.plot_distribution2(distributions, test[:10], start_at=model.order, ax=ax, cmap="Blues")
print("")
''' '''
model = fts.FCM_FTS(partitioner=fs, order=1) model = fts.FCM_FTS(partitioner=fs, order=1)