PEP 256 documentation

pyFTS/benchmarks/Measures.py

@@ -53,7 +53,6 @@ def rmse_interval(targets, forecasts):
     return np.sqrt(np.nanmean((fmean - targets) ** 2))
 
 
-
 def mape(targets, forecasts):
     """
     Mean Average Percentual Error
@@ -68,7 +67,6 @@ def mape(targets, forecasts):
     return np.mean(np.abs(targets - forecasts) / targets) * 100
 
 
-
 def smape(targets, forecasts, type=2):
     """
     Symmetric Mean Average Percentual Error
@@ -89,13 +87,11 @@ def smape(targets, forecasts, type=2):
         return sum(np.abs(forecasts - targets)) / sum(forecasts + targets)
 
 
-
 def mape_interval(targets, forecasts):
     fmean = [np.mean(i) for i in forecasts]
     return np.mean(abs(fmean - targets) / fmean) * 100
 
 
-
 def UStatistic(targets, forecasts):
     """
     Theil's U Statistic
@@ -117,7 +113,6 @@ def UStatistic(targets, forecasts):
     return np.sqrt(sum(y) / sum(naive))
 
 
-
 def TheilsInequality(targets, forecasts):
     """
     Theil’s Inequality Coefficient
@@ -194,7 +189,7 @@ def pinball(tau, target, forecast):
     :param tau: quantile value in the range (0,1)
     :param target: 
     :param forecast: 
-    :return: distance of forecast to the tau-quantile of the target
+    :return: float, distance of forecast to the tau-quantile of the target
     """
     if target >= forecast:
         return (target - forecast) * tau
@@ -208,7 +203,7 @@ def pinball_mean(tau, targets, forecasts):
     :param tau: quantile value in the range (0,1)
     :param targets: list of target values
     :param forecasts: list of prediction intervals
-    :return: 
+    :return: float, the pinball loss mean for tau quantile
     """
     try:
         if tau <= 0.5:
@@ -220,7 +215,6 @@ def pinball_mean(tau, targets, forecasts):
         print(ex)
 
 
-
 def pmf_to_cdf(density):
     ret = []
     for row in density.index:
@@ -244,7 +238,12 @@ def heavyside_cdf(bins, targets):
 
 
 def crps(targets, densities):
-    """Continuous Ranked Probability Score"""
+    '''
+    Continuous Ranked Probability Score
+    :param targets: a list with the target values
+    :param densities: a list with pyFTS.probabil objectsistic.ProbabilityDistribution
+    :return: float
+    '''
     _crps = float(0.0)
     if isinstance(densities, pd.DataFrame):
         l = len(densities.columns)
@@ -269,7 +268,13 @@ def crps(targets, densities):
 
 
 def get_point_statistics(data, model, **kwargs):
-    """Condensate all measures for point forecasters"""
+    '''
+    Condensate all measures for point forecasters
+    :param data: test data
+    :param model: FTS model with point forecasting capability
+    :param kwargs:
+    :return: a list with the RMSE, SMAPE and U Statistic
+    '''
 
     steps_ahead = kwargs.get('steps_ahead',1)
 
@@ -307,7 +312,14 @@ def get_point_statistics(data, model, **kwargs):
 
 
 def get_interval_statistics(data, model, **kwargs):
-    """Condensate all measures for point_to_interval forecasters"""
+    '''
+    Condensate all measures for point interval forecasters
+    :param data: test data
+    :param model: FTS model with interval forecasting capability
+    :param kwargs:
+    :return: a list with the sharpness, resolution, coverage, .05 pinball mean,
+    .25 pinball mean, .75 pinball mean and .95 pinball mean.
+    '''
 
     steps_ahead = kwargs.get('steps_ahead', 1)
 
@@ -341,6 +353,13 @@ def get_interval_statistics(data, model, **kwargs):
 
 
 def get_distribution_statistics(data, model, **kwargs):
+    '''
+    Get CRPS statistic and time for a forecasting model
+    :param data: test data
+    :param model: FTS model with probabilistic forecasting capability
+    :param kwargs:
+    :return: a list with the CRPS and execution time
+    '''
     steps_ahead = kwargs.get('steps_ahead', 1)
 
     ret = list()

pyFTS/benchmarks/ResidualAnalysis.py

@@ -31,7 +31,7 @@ def chi_squared(q, h):
 def compare_residuals(data, models):
     """
     Compare residual's statistics of several models
-    :param data: 
+    :param data: test data
     :param models: 
     :return: 
     """

pyFTS/benchmarks/Util.py

@@ -1,5 +1,5 @@
 """
-Benchmark utility functions
+Facilities for pyFTS Benchmark module
 """
 
 import matplotlib as plt

pyFTS/benchmarks/benchmarks.py

@@ -51,22 +51,31 @@ def __pop(key, default, kwargs):
 def sliding_window_benchmarks(data, windowsize, train=0.8, **kwargs):
     """
     Sliding window benchmarks for FTS point forecasters
-    :param data:
+    :param data: test data
     :param windowsize: size of sliding window
     :param train: percentual of sliding window data used to train the models
-    :param models: FTS point forecasters
-    :param partitioners: Universe of Discourse partitioner
-    :param partitions: the max number of partitions on the Universe of Discourse
-    :param max_order: the max order of the models (for high order models)
-    :param transformation: data transformation
-    :param indexer: seasonal indexer
-    :param dump:
-    :param benchmark_methods: Non FTS models to benchmark
-    :param benchmark_methods_parameters: Non FTS models parameters
-    :param save: save results
-    :param file: file path to save the results
-    :param sintetic: if true only the average and standard deviation of the results
-    :return: DataFrame with the results
+    :param kwargs: dict, optional arguments
+
+    :keyword
+        models: FTS point forecasters
+        partitioners: Universe of Discourse partitioner
+        partitions: the max number of partitions on the Universe of Discourse
+        max_order: the max order of the models (for high order models)
+        type: the forecasting type, one of these values: point(default), interval or distribution.
+        steps_ahead: The forecasting horizon, i. e., the number of steps ahead to forecast
+        start: in the multi step forecasting, the index of the data where to start forecasting
+        transformation: data transformation
+        indexer: seasonal indexer
+        progress: If true a progress bar will be displayed during the benchmarks
+        distributed: boolean, indicate if the forecasting procedure will be distributed in a dispy cluster
+        nodes: a list with the dispy cluster nodes addresses
+        benchmark_methods: Non FTS models to benchmark
+        benchmark_methods_parameters: Non FTS models parameters
+        save: save results
+        file: file path to save the results
+        sintetic: if true only the average and standard deviation of the results
+
+    :return: DataFrame with the benchmark results
     """
     distributed = __pop('distributed', False, kwargs)
     save = __pop('save', False, kwargs)
@@ -226,6 +235,12 @@ def sliding_window_benchmarks(data, windowsize, train=0.8, **kwargs):
             if job.status == dispy.DispyJob.Finished and job is not None:
                 tmp = job()
                 jobs2.append(tmp)
+                print(tmp)
+            else:
+                print("status",job.status)
+                print("result",job.result)
+                print("stdout",job.stdout)
+                print("stderr",job.exception)
 
         jobs = deepcopy(jobs2)
 
@@ -234,6 +249,8 @@ def sliding_window_benchmarks(data, windowsize, train=0.8, **kwargs):
     file = kwargs.get('file', None)
     sintetic = kwargs.get('sintetic', False)
 
+    print(jobs)
+
     return synthesis_method(jobs, experiments, save, file, sintetic)
 
 

pyFTS/common/FLR.py

@@ -1,12 +1,17 @@
-import numpy as np
-from pyFTS.common import FuzzySet
-
 """
 This module implements functions for Fuzzy Logical Relationship generation
 """
 
+import numpy as np
+from pyFTS.common import FuzzySet
+
+
 class FLR(object):
-    """Fuzzy Logical Relationship"""
+    """
+    Fuzzy Logical Relationship
+
+    Represents a temporal transition of the fuzzy set LHS on time t for the fuzzy set RHS on time t+1.
+    """
     def __init__(self, LHS, RHS):
         """
         Creates a Fuzzy Logical Relationship

pyFTS/common/Transformations.py

@@ -1,3 +1,7 @@
+"""
+Common data transformation used on pre and post processing of the FTS
+"""
+
 import numpy as np
 import math
 from pyFTS import *
@@ -5,7 +9,7 @@ from pyFTS import *
 
 class Transformation(object):
     """
-    Data transformation used to pre and post processing of the FTS
+    Data transformation used on pre and post processing of the FTS
     """
 
     def __init__(self, **kwargs):
@@ -13,9 +17,23 @@ class Transformation(object):
         self.minimal_length = 1
 
     def apply(self, data, param, **kwargs):
+        """
+        Apply the transformation on input data
+        :param data: input data
+        :param param:
+        :param kwargs:
+        :return: numpy array with transformed data
+        """
         pass
 
     def inverse(self,data, param, **kwargs):
+        """
+
+        :param data: transformed data
+        :param param:
+        :param kwargs:
+        :return: numpy array with inverse transformed data
+        """
         pass
 
     def __str__(self):
@@ -73,6 +91,11 @@ class Differential(Transformation):
 
 
 class Scale(Transformation):
+    """
+    Scale data inside a interval [min, max]
+
+    
+    """
     def __init__(self, min=0, max=1):
         super(Scale, self).__init__()
         self.data_max = None
@@ -130,6 +153,9 @@ class AdaptiveExpectation(Transformation):
 
 
 class BoxCox(Transformation):
+    """
+    Box-Cox power transformation
+    """
     def __init__(self, plambda):
         super(BoxCox, self).__init__()
         self.plambda = plambda

pyFTS/common/Util.py

@@ -1,3 +1,7 @@
+"""
+Common facilities for pyFTS
+"""
+
 import time
 import matplotlib.pyplot as plt
 import dill

pyFTS/common/flrg.py

@@ -2,6 +2,12 @@ import numpy as np
 
 
 class FLRG(object):
+    """
+    Fuzzy Logical Relationship Group
+
+    Group a set of FLR's with the same LHS. Represents the temporal patterns for time t+1 (the RHS fuzzy sets)
+    when the LHS pattern is identified on time t.
+    """
 
     def __init__(self, order, **kwargs):
         self.LHS = None

pyFTS/common/fts.py

@@ -120,7 +120,6 @@ class FTS(object):
 
         return ret
 
-
     def forecast(self, data, **kwargs):
         """
         Point forecast one step ahead 

pyFTS/common/tree.py

@@ -1,9 +1,17 @@
+"""
+Tree data structure
+"""
+
 from pyFTS import *
 from functools import reduce
 import numpy as np
 
 
 class FLRGTreeNode:
+    """
+    Tree node for
+    """
+
     def __init__(self, value):
         self.isRoot = False
         self.children = []

pyFTS/data/AirPassengers.py

@@ -4,11 +4,19 @@ import numpy as np
 
 
 def get_data():
+    """
+    Get a simple univariate time series data.
+    :return: numpy array
+    """
     dat = get_dataframe()
     dat = np.array(dat["Passengers"])
     return dat
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('AirPassengers.csv',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/AirPassengers.csv',
                                sep=",")

pyFTS/data/Enrollments.py

@@ -4,6 +4,10 @@ import numpy as np
 
 
 def get_data():
+    """
+    Get a simple univariate time series data.
+    :return: numpy array
+    """
     dat = get_dataframe()
     dat = np.array(dat["Enrollments"])
     return dat

pyFTS/data/INMET.py

@@ -18,6 +18,10 @@ import pandas as pd
 
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('INMET.csv.bz2',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/INMET.csv.bz2',
                                sep=";", compression='bz2')

pyFTS/data/NASDAQ.py

@@ -4,12 +4,21 @@ import numpy as np
 
 
 def get_data(field):
+    """
+    Get a simple univariate time series data.
+    :param field: the dataset field name to extract
+    :return: numpy array
+    """
     dat = get_dataframe()
     dat = np.array(dat[field])
     return dat
 
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('NASDAQ.csv.bz2',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/NASDAQ.csv.bz2',
                                sep=";", compression='bz2')

pyFTS/data/SONDA.py

@@ -4,12 +4,21 @@ import numpy as np
 
 
 def get_data(field):
+    """
+    Get a simple univariate time series data.
+    :param field: the dataset field name to extract
+    :return: numpy array
+    """
     dat = get_dataframe()
     dat = np.array(dat[field])
     return dat
 
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('SONDA_BSB.csv.bz2',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/SONDA_BSB.csv.bz2',
                                sep=";", compression='bz2')

pyFTS/data/SP500.py

@@ -4,6 +4,10 @@ import numpy as np
 
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('SP500.csv.bz2',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/SP500.csv.bz2',
                                sep=",", compression='bz2')

pyFTS/data/TAIEX.py

@@ -4,12 +4,20 @@ import numpy as np
 
 
 def get_data():
+    """
+    :param field: the dataset field name to extract
+    :return: numpy array
+    """
     dat = get_dataframe()
     dat = np.array(dat["avg"])
     return dat
 
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('TAIEX.csv.bz2',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/TAIEX.csv.bz2',
                                sep=",", compression='bz2')

pyFTS/data/__init__.py

@@ -0,0 +1,3 @@
+"""
+Module for pyFTS standard datasets facilities
+"""

pyFTS/data/artificial.py

@@ -1,3 +1,7 @@
+"""
+Facilities to generate synthetic stochastic processes
+"""
+
 import numpy as np
 
 

pyFTS/data/common.py

@@ -1,3 +1,4 @@
+
 import pandas as pd
 import numpy as np
 import os
@@ -7,7 +8,16 @@ from urllib import request
 
 
 def get_dataframe(filename, url, sep=";", compression='infer'):
-    #filename = pkg_resources.resource_filename('pyFTS', path)
+    """
+    This method check if filename already exists, read the file and return its data.
+    If the file don't already exists, it will be downloaded and decompressed.
+    :param filename: dataset local filename
+    :param url: dataset internet URL
+    :param sep: CSV field separator
+    :param compression: type of compression
+    :return:  Pandas dataset
+    """
+
     tmp_file = Path(filename)
 
     if tmp_file.is_file():

pyFTS/data/sunspots.py

@@ -3,11 +3,19 @@ import pandas as pd
 import numpy as np
 
 def get_data():
+    """
+    Get a simple univariate time series data.
+    :return: numpy array
+    """
     dat = get_dataframe()
     dat = np.array(dat["SUNACTIVITY"])
     return dat
 
 def get_dataframe():
+    """
+    Get the complete multivariate time series data.
+    :return: Pandas DataFrame
+    """
     dat = common.get_dataframe('sunspots.csv',
                                'https://github.com/petroniocandido/pyFTS/raw/8f20f3634aa6a8f58083bdcd1bbf93795e6ed767/pyFTS/data/sunspots.csv',
                                sep=",")

pyFTS/models/__init__.py

@@ -0,0 +1,3 @@
+"""
+Fuzzy Time Series methods
+"""

pyFTS/models/ensemble/__init__.py

@@ -0,0 +1,3 @@
+"""
+Meta FTS that aggregates other FTS methods
+"""

pyFTS/models/ifts.py

@@ -1,6 +1,13 @@
 #!/usr/bin/python
 # -*- coding: utf8 -*-
 
+"""
+High Order Interval Fuzzy Time Series
+
+SILVA, Petrônio CL; SADAEI, Hossein Javedani; GUIMARÃES, Frederico Gadelha. Interval Forecasting with Fuzzy Time Series.
+In: Computational Intelligence (SSCI), 2016 IEEE Symposium Series on. IEEE, 2016. p. 1-8.
+"""
+
 import numpy as np
 from pyFTS.common import FuzzySet, FLR, fts, tree
 from pyFTS.models import hofts
@@ -9,9 +16,6 @@ from pyFTS.models import hofts
 class IntervalFTS(hofts.HighOrderFTS):
     """
     High Order Interval Fuzzy Time Series
-
-    SILVA, Petrônio CL; SADAEI, Hossein Javedani; GUIMARÃES, Frederico Gadelha. Interval Forecasting with Fuzzy Time Series.
-    In: Computational Intelligence (SSCI), 2016 IEEE Symposium Series on. IEEE, 2016. p. 1-8.
     """
     def __init__(self, name, **kwargs):
         super(IntervalFTS, self).__init__(name="IFTS " + name, **kwargs)

pyFTS/models/multivariate/__init__.py

@@ -0,0 +1,3 @@
+"""
+Multivariate Fuzzy Time Series methods
+"""

pyFTS/models/nonstationary/__init__.py

@@ -0,0 +1,3 @@
+"""
+Fuzzy time series with nonstationary fuzzy sets, for heteroskedastic data
+"""

pyFTS/notebooks/__init__.py

@@ -0,0 +1,3 @@
+"""
+Jupyter notebooks with pyFTS usage examples
+"""

pyFTS/partitioners/Entropy.py

@@ -1,3 +1,8 @@
+"""
+C. H. Cheng, R. J. Chang, and C. A. Yeh, “Entropy-based and trapezoidal fuzzification-based fuzzy time series approach for forecasting IT project cost,”
+Technol. Forecast. Social Change, vol. 73, no. 5, pp. 524–542, Jun. 2006.
+"""
+
 import numpy as np
 import math
 import random as rnd
@@ -5,9 +10,6 @@ import functools, operator
 from pyFTS.common import FuzzySet, Membership
 from pyFTS.partitioners import partitioner
 
-# C. H. Cheng, R. J. Chang, and C. A. Yeh, “Entropy-based and trapezoidal fuzzification-based fuzzy time series approach for forecasting IT project cost,”
-# Technol. Forecast. Social Change, vol. 73, no. 5, pp. 524–542, Jun. 2006.
-
 
 def splitBelow(data,threshold):
     return [k for k in data if k <= threshold]

pyFTS/partitioners/FCM.py

@@ -1,3 +1,7 @@
+"""
+S. T. Li, Y. C. Cheng, and S. Y. Lin, “A FCM-based deterministic forecasting model for fuzzy time series,”
+Comput. Math. Appl., vol. 56, no. 12, pp. 3052–3063, Dec. 2008. DOI: 10.1016/j.camwa.2008.07.033.
+"""
 import numpy as np
 import math
 import random as rnd
@@ -6,11 +10,6 @@ from pyFTS.common import FuzzySet, Membership
 from pyFTS.partitioners import partitioner
 
 
-# import CMeans
-
-# S. T. Li, Y. C. Cheng, and S. Y. Lin, “A FCM-based deterministic forecasting model for fuzzy time series,”
-# Comput. Math. Appl., vol. 56, no. 12, pp. 3052–3063, Dec. 2008. DOI: 10.1016/j.camwa.2008.07.033.
-
 def fuzzy_distance(x, y):
     if isinstance(x, list):
         tmp = functools.reduce(operator.add, [(x[k] - y[k]) ** 2 for k in range(0, len(x))])

pyFTS/partitioners/Grid.py

@@ -1,3 +1,5 @@
+"""Even Length Grid Partitioner"""
+
 import numpy as np
 import math
 import random as rnd

pyFTS/partitioners/Huarng.py

@@ -1,15 +1,16 @@
+"""
+K. H. Huarng, “Effective lengths of intervals to improve forecasting in fuzzy time series,”
+Fuzzy Sets Syst., vol. 123, no. 3, pp. 387–394, Nov. 2001.
+"""
+
 import numpy as np
 import math
 import random as rnd
 import functools, operator
 from pyFTS.common import FuzzySet, Membership, Transformations
 
-
-# K. H. Huarng, “Effective lengths of intervals to improve forecasting in fuzzy time series,”
-# Fuzzy Sets Syst., vol. 123, no. 3, pp. 387–394, Nov. 2001.
 from pyFTS.partitioners import partitioner
 
-
 class HuarngPartitioner(partitioner.Partitioner):
     """Huarng Empirical Partitioner"""
     def __init__(self, **kwargs):

pyFTS/partitioners/Util.py

@@ -1,3 +1,7 @@
+"""
+Facility methods for pyFTS partitioners module
+"""
+
 import numpy as np
 import pandas as pd
 import matplotlib as plt

pyFTS/probabilistic/__init__.py

@@ -0,0 +1,3 @@
+"""
+Probability Distribution objects
+"""

pyFTS/probabilistic/kde.py

@@ -1,10 +1,12 @@
-from pyFTS.common import Transformations
-import numpy as np
+# -*- coding: utf8 -*-
 
 """
 Kernel Density Estimation 
 """
 
+from pyFTS.common import Transformations
+import numpy as np
+
 
 class KernelSmoothing(object):
     """Kernel Density Estimation"""

pyFTS/tests/general.py

@@ -20,10 +20,10 @@ from pyFTS.benchmarks import benchmarks as bchmk
 from pyFTS.models import pwfts
 
 #'''
-bchmk.sliding_window_benchmarks(dataset, 1000, train=0.8, inc=0.2, methods=[pwfts.ProbabilisticWeightedFTS],
-                                benchmark_models=False, orders=[1], partitions=[10], #np.arange(10,100,2),
+bchmk.sliding_window_benchmarks(dataset[:2000], 1000, train=0.8, inc=0.2, methods=[pwfts.ProbabilisticWeightedFTS],
+                                benchmark_models=False, orders=[1,2,3], partitions=[30,50,70], #np.arange(10,100,2),
                                 progress=False, type='distribution', steps_ahead=[1,4,7,10],
-                                #distributed=False, nodes=['192.168.0.106', '192.168.0.105', '192.168.0.110'],
+                                distributed=True, nodes=['192.168.0.102','192.168.0.106','192.168.0.110'],
                                 save=True, file="pwfts_taiex_distribution.csv")
 #'''