pyFTS/common/Transformations.py

92 lines
2.1 KiB
Python
Raw Normal View History

import numpy as np
import math
from pyFTS import *
class Transformation(object):
def __init__(self, parameters):
self.isInversible = True
self.parameters = parameters
self.minimalLength = 1
def apply(self,data,param):
pass
def inverse(self,data, param):
pass
def __str__(self):
return self.__class__.__name__ + '(' + str(self.parameters) + ')'
class Differential(Transformation):
def __init__(self, parameters):
super(Differential, self).__init__(parameters)
self.lag = parameters
self.minimalLength = 2
def apply(self, data, param=None):
if param is not None:
self.lag = param
if not isinstance(data, (list, np.ndarray, np.generic)):
data = [data]
if isinstance(data, (np.ndarray, np.generic)):
data = data.tolist()
n = len(data)
diff = [data[t - self.lag] - data[t] for t in np.arange(self.lag, n)]
for t in np.arange(0, self.lag): diff.insert(0, 0)
return diff
def inverse(self,data, param):
if isinstance(data, (np.ndarray, np.generic)):
data = data.tolist()
if not isinstance(data, list):
data = [data]
n = len(data)
inc = [data[t] + param[t] for t in np.arange(0, n)]
if n == 1:
return inc[0]
else:
return inc
def boxcox(original, plambda):
n = len(original)
if plambda != 0:
modified = [(original[t] ** plambda - 1) / plambda for t in np.arange(0, n)]
else:
modified = [math.log(original[t]) for t in np.arange(0, n)]
return np.array(modified)
def Z(original):
mu = np.mean(original)
sigma = np.std(original)
z = [(k - mu)/sigma for k in original]
return z
# retrieved from Sadaei and Lee (2014) - Multilayer Stock ForecastingModel Using Fuzzy Time Series
def roi(original):
n = len(original)
roi = []
for t in np.arange(0, n-1):
roi.append( (original[t+1] - original[t])/original[t] )
return roi
def smoothing(original, lags):
pass
def aggregate(original, operation):
pass