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

ResidualAnalysis bugfix

Browse Source
This commit is contained in:
Petrônio Cândido 2019-09-30 19:28:04 -03:00
parent aa4598f79f
commit 039dca9d5d
2 changed files with 17 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
pyFTS/benchmarks/ResidualAnalysis.py
View File

@ -69,7 +69,7 @@ def plot_residuals_by_model(targets, models, tam=[8, 8], save=False, file=None):
else:
ax = axes
forecasts = mfts.predict(targets)
res = residuals(targets, forecasts, mfts.order+1)
res = residuals(targets, forecasts, mfts.order)
mu = np.mean(res)
sig = np.std(res)

20
pyFTS/tests/general.py
View File

@ -17,7 +17,11 @@ from pyFTS.common import Transformations, Membership, Util
from pyFTS.benchmarks import arima, quantreg, BSTS, gaussianproc, knn
from pyFTS.fcm import fts, common, GA
from pyFTS.data import TAIEX, NASDAQ, SP500
from pyFTS.common import Transformations
tdiff = Transformations.Differential(1)
boxcox = Transformations.BoxCox(0)
from pyFTS.data import TAIEX, NASDAQ, SP500
from pyFTS.common import Util
@ -28,10 +32,18 @@ test = TAIEX.get_data()[1800:2000]
from pyFTS.models import pwfts
from pyFTS.partitioners import Grid
fs = Grid.GridPartitioner(data=train, npart=45)
fs = Grid.GridPartitioner(data=train, npart=15, transformation=tdiff)
model = pwfts.ProbabilisticWeightedFTS(partitioner=fs, order=1)
#model = pwfts.ProbabilisticWeightedFTS(partitioner=fs, order=1)
model = chen.ConventionalFTS(partitioner=fs)
model.append_transformation(tdiff)
model.fit(train)
from pyFTS.benchmarks import ResidualAnalysis as ra
ra.plot_residuals_by_model(test, [model])
horizon = 10
'''
forecasts = model.predict(test[9:20], type='point')
@ -41,7 +53,7 @@ distributions = model.predict(test[9:20], type='distribution')
forecasts = model.predict(test[9:20], type='point', steps_ahead=horizon)
intervals = model.predict(test[9:20], type='interval', steps_ahead=horizon)
'''
distributions = model.predict(test[9:20], type='distribution', steps_ahead=horizon)
#distributions = model.predict(test[9:20], type='distribution', steps_ahead=horizon)
'''