notebook.community

Edit and run 





In [1]:



    


import pandas as pd
import numpy as np
import pyaf.ForecastEngine as autof

from IPython.display import Image
from IPython.display import display, HTML 

%matplotlib inline  

import logging
import logging.config


# remove the notebook root logger.
logger = logging.getLogger()
logger.handlers = []

logging.basicConfig(level=logging.INFO)



    











In [2]:



    


def analyze_dataset(name , horizon):
    signal = name.replace(".csv" , "") 
    url="https://raw.githubusercontent.com/antoinecarme/TimeSeriesData/master/R_TSData/expsmooth/" + name 
    df = pd.read_csv(url)
    df = df [[df.columns[0] , df.columns[-1]]].dropna()
    df.columns = ['Date' , signal]
    lEngine = autof.cForecastEngine()
    N = df.shape[0]
    lEngine.train(df , 'Date' , signal , horizon)
    lEngine.getModelInfo()
    print("PERFORMANCE MAPE_FORECAST" , signal, lEngine.mSignalDecomposition.mBestModel.mForecastPerf.mMAPE)
    # lEngine.standrdPlots()
    prefix = '<img src="data:image/png;base64,'
    lDict = lEngine.getPlotsAsDict()
    uri = prefix + lDict['Prediction_Intervals'] + '">'
    display(HTML(uri))
    return lEngine



    











In [3]:



    


csvs="ausgdp.csv dji.csv gasprice.csv partx.csv utility.csv bonds.csv enplanements.csv hospital.csv ukcars.csv vehicles.csv canadagas.csv fmsales.csv jewelry.csv unemp.cci.csv visitors.csv carparts.csv freight.csv mcopper.csv usgdp.csv xrates.csv djiclose.csv frexport.csv msales.csv usnetelec.csv"



    











In [4]:



    


lEngine_djiclose = analyze_dataset('djiclose.csv' , 4)



    


















    






INFO:pyaf.std:START_TRAINING 'djiclose'
INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'djiclose' 7.205088376998901
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1928.83333333 TimeMax=1991.75 TimeDelta=0.0833333333333 Estimation = (0 , 756) Validation = (756 , 946) Test = (946 , 950) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='CumSum_djiclose' Min=-145.87 Max=539.57  Mean=209.252263158 StdDev=170.913496014
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'CumSum_'
INFO:pyaf.std:BEST_DECOMPOSITION  'CumSum_djiclose_ConstantTrend_residue_zeroCycle_residue_NoAR' [ConstantTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL 'CumSum_djiclose_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL 'CumSum_djiclose_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'CumSum_djiclose_ConstantTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=1.0089 MAPE_Forecast=1.0 MAPE_Test=1.0
INFO:pyaf.std:MODEL_L2 L2_Fit=7.26607495003 L2_Forecast=4.08732508076 L2_Test=2.87330384749
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END
/usr/lib/python3/dist-packages/matplotlib/__init__.py:1403: UserWarning:  This call to matplotlib.use() has no effect
because the backend has already been chosen;
matplotlib.use() must be called *before* pylab, matplotlib.pyplot,
or matplotlib.backends is imported for the first time.

  warnings.warn(_use_error_msg)










    






PERFORMANCE MAPE_FORECAST djiclose 1.0










    























In [ ]:



    






    











In [5]:



    


for csv in csvs.split():
    print("analyzing " , csv);
    analyze_dataset(csv , 4)



    


















    






INFO:pyaf.std:START_TRAINING 'ausgdp'










    






analyzing  ausgdp.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'ausgdp' 2.2820041179656982
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1971.5 TimeMax=1991.75 TimeDelta=0.25 Estimation = (0 , 82) Validation = (82 , 103) Test = (103 , 107) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_ausgdp' Min=4612 Max=7618  Mean=5870.18691589 StdDev=802.972365196
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_ausgdp_ConstantTrend_residue_zeroCycle_residue_AR(26)' [ConstantTrend + NoCycle + AR(26)]
INFO:pyaf.std:TREND_DETAIL '_ausgdp_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_ausgdp_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_ausgdp_ConstantTrend_residue_zeroCycle_residue_AR(26)' [AR(26)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0021 MAPE_Forecast=0.0018 MAPE_Test=0.0023
INFO:pyaf.std:MODEL_L2 L2_Fit=15.9088393793 L2_Forecast=13.4130551088 L2_Test=19.2273397978
INFO:pyaf.std:MODEL_COMPLEXITY 26
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag2 -3.09318331897
INFO:pyaf.std:AR_MODEL_COEFF 2 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag1 2.60854394661
INFO:pyaf.std:AR_MODEL_COEFF 3 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag3 2.41926926061
INFO:pyaf.std:AR_MODEL_COEFF 4 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag4 -1.43190340152
INFO:pyaf.std:AR_MODEL_COEFF 5 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag17 1.24905779387
INFO:pyaf.std:AR_MODEL_COEFF 6 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag16 -1.18804293427
INFO:pyaf.std:AR_MODEL_COEFF 7 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag9 -1.09337222611
INFO:pyaf.std:AR_MODEL_COEFF 8 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag10 1.0784689542
INFO:pyaf.std:AR_MODEL_COEFF 9 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag18 -0.931476443481
INFO:pyaf.std:AR_MODEL_COEFF 10 _ausgdp_ConstantTrend_residue_zeroCycle_residue_Lag15 0.878232923779
INFO:pyaf.std:AR_MODEL_DETAIL_END
/usr/lib/python3/dist-packages/matplotlib/__init__.py:1403: UserWarning:  This call to matplotlib.use() has no effect
because the backend has already been chosen;
matplotlib.use() must be called *before* pylab, matplotlib.pyplot,
or matplotlib.backends is imported for the first time.

  warnings.warn(_use_error_msg)










    






PERFORMANCE MAPE_FORECAST ausgdp 0.0018










    
















    






INFO:pyaf.std:START_TRAINING 'dji'










    






analyzing  dji.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'dji' 2.6621153354644775
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1990.0 TimeMax=2003.41666667 TimeDelta=0.0833333333334 Estimation = (0 , 162) Validation = (162 , 203) Test = (203 , 207) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_dji' Min=2442.33 Max=12621.7  Mean=7394.09135266 StdDev=3206.55127118
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_dji_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_dji_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_dji_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_dji_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0347 MAPE_Forecast=0.0197 MAPE_Test=0.0169
INFO:pyaf.std:MODEL_L2 L2_Fit=348.594912422 L2_Forecast=251.999985501 L2_Test=232.050204697
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST dji 0.0197










    
















    






INFO:pyaf.std:START_TRAINING 'gasprice'










    






analyzing  gasprice.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'gasprice' 2.559203624725342
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1991.0 TimeMax=2003.33333333 TimeDelta=0.0833333333333 Estimation = (0 , 149) Validation = (149 , 187) Test = (187 , 191) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_gasprice' Min=11.35 Max=74.41  Mean=27.9483769634 StdDev=14.3490218571
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_gasprice_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_gasprice_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_gasprice_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_gasprice_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0576 MAPE_Forecast=0.0582 MAPE_Test=0.0625
INFO:pyaf.std:MODEL_L2 L2_Fit=1.64291523512 L2_Forecast=3.51846108705 L2_Test=5.27727912849
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST gasprice 0.0582










    
















    






INFO:pyaf.std:START_TRAINING 'partx'










    






analyzing  partx.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'partx' 1.8329050540924072
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1.0 TimeMax=4.0 TimeDelta=0.0833333333333 Estimation = (0 , 37) Validation = (37 , 47) Test = (47 , 51) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='CumSum_partx' Min=0 Max=32  Mean=11.8823529412 StdDev=11.4748609438
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'CumSum_'
INFO:pyaf.std:BEST_DECOMPOSITION  'CumSum_partx_PolyTrend_residue_zeroCycle_residue_NoAR' [PolyTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL 'CumSum_partx_PolyTrend' [PolyTrend]
INFO:pyaf.std:CYCLE_DETAIL 'CumSum_partx_PolyTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'CumSum_partx_PolyTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.6516 MAPE_Forecast=0.3558 MAPE_Test=0.2396
INFO:pyaf.std:MODEL_L2 L2_Fit=1.02193312074 L2_Forecast=1.28723890717 L2_Test=1.43023894853
INFO:pyaf.std:MODEL_COMPLEXITY 48
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST partx 0.3558










    
















    






INFO:pyaf.std:START_TRAINING 'utility'










    






analyzing  utility.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'utility' 13.675541877746582
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1.0 TimeMax=101.625 TimeDelta=0.0416666666667 Estimation = (0 , 2416) Validation = (2416 , 3020) Test = (3020 , 3024) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_utility' Min=3798 Max=14940  Mean=8352.43650794 StdDev=2433.02565296
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_utility_ConstantTrend_residue_zeroCycle_residue_AR(256)' [ConstantTrend + NoCycle + AR(256)]
INFO:pyaf.std:TREND_DETAIL '_utility_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_utility_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_utility_ConstantTrend_residue_zeroCycle_residue_AR(256)' [AR(256)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0164 MAPE_Forecast=0.0239 MAPE_Test=0.02
INFO:pyaf.std:MODEL_L2 L2_Fit=192.201940465 L2_Forecast=246.34533778 L2_Test=172.629957382
INFO:pyaf.std:MODEL_COMPLEXITY 256
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _utility_ConstantTrend_residue_zeroCycle_residue_Lag1 1.34570750102
INFO:pyaf.std:AR_MODEL_COEFF 2 _utility_ConstantTrend_residue_zeroCycle_residue_Lag2 -0.384218719323
INFO:pyaf.std:AR_MODEL_COEFF 3 _utility_ConstantTrend_residue_zeroCycle_residue_Lag169 -0.363507874625
INFO:pyaf.std:AR_MODEL_COEFF 4 _utility_ConstantTrend_residue_zeroCycle_residue_Lag25 -0.317904106335
INFO:pyaf.std:AR_MODEL_COEFF 5 _utility_ConstantTrend_residue_zeroCycle_residue_Lag24 0.233874469523
INFO:pyaf.std:AR_MODEL_COEFF 6 _utility_ConstantTrend_residue_zeroCycle_residue_Lag145 -0.214704835051
INFO:pyaf.std:AR_MODEL_COEFF 7 _utility_ConstantTrend_residue_zeroCycle_residue_Lag167 0.200407393799
INFO:pyaf.std:AR_MODEL_COEFF 8 _utility_ConstantTrend_residue_zeroCycle_residue_Lag168 0.190452144146
INFO:pyaf.std:AR_MODEL_COEFF 9 _utility_ConstantTrend_residue_zeroCycle_residue_Lag144 0.136468679476
INFO:pyaf.std:AR_MODEL_COEFF 10 _utility_ConstantTrend_residue_zeroCycle_residue_Lag27 0.0978124794361
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST utility 0.0239










    
















    






INFO:pyaf.std:START_TRAINING 'bonds'










    






analyzing  bonds.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'bonds' 2.2585999965667725
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1994.0 TimeMax=2001.91666667 TimeDelta=0.0833333333334 Estimation = (0 , 96) Validation = (96 , 121) Test = (121 , 125) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_bonds' Min=3.32 Max=8.12  Mean=5.68304 StdDev=1.09253336718
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_bonds_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_bonds_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_bonds_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_bonds_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0336 MAPE_Forecast=0.044 MAPE_Test=0.0704
INFO:pyaf.std:MODEL_L2 L2_Fit=0.242074368738 L2_Forecast=0.246981780705 L2_Test=0.343474889912
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST bonds 0.044










    
















    






INFO:pyaf.std:START_TRAINING 'enplanements'










    






analyzing  enplanements.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'enplanements' 3.0647730827331543
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1979.0 TimeMax=1997.41666667 TimeDelta=0.0833333333333 Estimation = (0 , 222) Validation = (222 , 278) Test = (278 , 282) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_enplanements' Min=20.14 Max=56.14  Mean=35.6665602837 StdDev=9.30945370252
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_enplanements_ConstantTrend_residue_zeroCycle_residue_AR(70)' [ConstantTrend + NoCycle + AR(70)]
INFO:pyaf.std:TREND_DETAIL '_enplanements_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_enplanements_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_enplanements_ConstantTrend_residue_zeroCycle_residue_AR(70)' [AR(70)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0246 MAPE_Forecast=0.0253 MAPE_Test=0.0427
INFO:pyaf.std:MODEL_L2 L2_Fit=1.00599964097 L2_Forecast=2.55053193921 L2_Test=2.17256844652
INFO:pyaf.std:MODEL_COMPLEXITY 70
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag1 0.704189389185
INFO:pyaf.std:AR_MODEL_COEFF 2 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag13 -0.413772710901
INFO:pyaf.std:AR_MODEL_COEFF 3 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag12 0.409416460175
INFO:pyaf.std:AR_MODEL_COEFF 4 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag60 0.291761480512
INFO:pyaf.std:AR_MODEL_COEFF 5 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag27 0.210801684483
INFO:pyaf.std:AR_MODEL_COEFF 6 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag6 -0.207796695427
INFO:pyaf.std:AR_MODEL_COEFF 7 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag32 -0.190965414109
INFO:pyaf.std:AR_MODEL_COEFF 8 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag17 -0.187766204461
INFO:pyaf.std:AR_MODEL_COEFF 9 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag24 0.170934374671
INFO:pyaf.std:AR_MODEL_COEFF 10 _enplanements_ConstantTrend_residue_zeroCycle_residue_Lag8 0.159484964706
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST enplanements 0.0253










    
















    






analyzing  hospital.csv










    






INFO:pyaf.std:START_TRAINING 'hospital'
INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'hospital' 1.9305205345153809
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=2000.0 TimeMax=2005.25 TimeDelta=0.0833333333333 Estimation = (0 , 64) Validation = (64 , 80) Test = (80 , 84) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_hospital' Min=35 Max=108  Mean=60.5119047619 StdDev=18.351394165
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_hospital_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_hospital_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_hospital_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_hospital_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.1962 MAPE_Forecast=0.1595 MAPE_Test=0.2583
INFO:pyaf.std:MODEL_L2 L2_Fit=17.0687946265 L2_Forecast=9.41740410092 L2_Test=11.874342087
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST hospital 0.1595










    
















    






INFO:pyaf.std:START_TRAINING 'ukcars'










    






analyzing  ukcars.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'ukcars' 2.315995454788208
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1977.0 TimeMax=1998.5 TimeDelta=0.25 Estimation = (0 , 87) Validation = (87 , 109) Test = (109 , 113) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='CumSum_ukcars' Min=330.371 Max=37682.986  Mean=17058.7638673 StdDev=10825.8338602
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'CumSum_'
INFO:pyaf.std:BEST_DECOMPOSITION  'CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_AR(28)' [ConstantTrend + NoCycle + AR(28)]
INFO:pyaf.std:TREND_DETAIL 'CumSum_ukcars_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL 'CumSum_ukcars_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_AR(28)' [AR(28)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.1065 MAPE_Forecast=0.0649 MAPE_Test=0.0358
INFO:pyaf.std:MODEL_L2 L2_Fit=57.8803549864 L2_Forecast=31.4455404884 L2_Test=15.1005810101
INFO:pyaf.std:MODEL_COMPLEXITY 60
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag1 1.28217837913
INFO:pyaf.std:AR_MODEL_COEFF 2 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag9 -0.507750809972
INFO:pyaf.std:AR_MODEL_COEFF 3 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag10 0.384523591418
INFO:pyaf.std:AR_MODEL_COEFF 4 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag5 -0.37889546755
INFO:pyaf.std:AR_MODEL_COEFF 5 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag7 0.326784262879
INFO:pyaf.std:AR_MODEL_COEFF 6 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag16 0.274252764905
INFO:pyaf.std:AR_MODEL_COEFF 7 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag17 -0.223484103552
INFO:pyaf.std:AR_MODEL_COEFF 8 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag20 0.201791156117
INFO:pyaf.std:AR_MODEL_COEFF 9 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag18 -0.201051931214
INFO:pyaf.std:AR_MODEL_COEFF 10 CumSum_ukcars_ConstantTrend_residue_zeroCycle_residue_Lag4 0.182354226343
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST ukcars 0.0649










    
















    






INFO:pyaf.std:START_TRAINING 'vehicles'










    






analyzing  vehicles.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'vehicles' 9.159932613372803
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1.0 TimeMax=57.125 TimeDelta=0.0416666666667 Estimation = (0 , 1348) Validation = (1348 , 1685) Test = (1685 , 1689) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_vehicles' Min=154 Max=5549  Mean=2060.41799882 StdDev=1339.14923305
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_vehicles_Lag1Trend_residue_bestCycle_byL2_residue_NoAR' [Lag1Trend + Cycle_Length_24 + NoAR]
INFO:pyaf.std:TREND_DETAIL '_vehicles_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_vehicles_Lag1Trend_residue_bestCycle_byL2' [Cycle_Length_24]
INFO:pyaf.std:AUTOREG_DETAIL '_vehicles_Lag1Trend_residue_bestCycle_byL2_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.1647 MAPE_Forecast=0.1941 MAPE_Test=0.1303
INFO:pyaf.std:MODEL_L2 L2_Fit=357.952867328 L2_Forecast=368.606027515 L2_Test=418.625224138
INFO:pyaf.std:MODEL_COMPLEXITY 40
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST vehicles 0.1941










    
















    






INFO:pyaf.std:START_TRAINING 'canadagas'










    






analyzing  canadagas.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'canadagas' 4.538783073425293
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1960.0 TimeMax=1995.75 TimeDelta=0.0833333333333 Estimation = (0 , 430) Validation = (430 , 538) Test = (538 , 542) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_canadagas' Min=0.966 Max=19.5284  Mean=9.77650498155 StdDev=5.13574993339
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_canadagas_ConstantTrend_residue_zeroCycle_residue_AR(135)' [ConstantTrend + NoCycle + AR(135)]
INFO:pyaf.std:TREND_DETAIL '_canadagas_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_canadagas_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_canadagas_ConstantTrend_residue_zeroCycle_residue_AR(135)' [AR(135)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0277 MAPE_Forecast=0.017 MAPE_Test=0.0149
INFO:pyaf.std:MODEL_L2 L2_Fit=0.226490192208 L2_Forecast=0.362180239017 L2_Test=0.316425525264
INFO:pyaf.std:MODEL_COMPLEXITY 135
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag1 0.691600463706
INFO:pyaf.std:AR_MODEL_COEFF 2 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag48 0.310071372461
INFO:pyaf.std:AR_MODEL_COEFF 3 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag28 -0.209548151031
INFO:pyaf.std:AR_MODEL_COEFF 4 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag76 0.207650626842
INFO:pyaf.std:AR_MODEL_COEFF 5 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag24 0.201843429093
INFO:pyaf.std:AR_MODEL_COEFF 6 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag15 -0.189865834659
INFO:pyaf.std:AR_MODEL_COEFF 7 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag13 -0.1773696465
INFO:pyaf.std:AR_MODEL_COEFF 8 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag84 0.170779453098
INFO:pyaf.std:AR_MODEL_COEFF 9 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag36 0.169284186628
INFO:pyaf.std:AR_MODEL_COEFF 10 _canadagas_ConstantTrend_residue_zeroCycle_residue_Lag74 -0.16305997978
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST canadagas 0.017










    
















    






INFO:pyaf.std:START_TRAINING 'fmsales'










    






analyzing  fmsales.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'fmsales' 1.9707601070404053
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1 TimeMax=46 TimeDelta=1.0 Estimation = (0 , 46) Validation = (46 , 58) Test = (58 , 62) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_fmsales' Min=22.2413797644 Max=51.9140807231  Mean=32.4748610823 StdDev=5.44569033576
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_fmsales_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_fmsales_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_fmsales_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_fmsales_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0656 MAPE_Forecast=0.0998 MAPE_Test=0.0621
INFO:pyaf.std:MODEL_L2 L2_Fit=2.88818047032 L2_Forecast=5.90836732884 L2_Test=2.95562686812
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST fmsales 0.0998










    
















    






analyzing  jewelry.csv










    






INFO:pyaf.std:START_TRAINING 'jewelry'
INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'jewelry' 2.231196641921997
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1998.07692308 TimeMax=1999.90384615 TimeDelta=0.0192307692308 Estimation = (0 , 96) Validation = (96 , 120) Test = (120 , 124) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_jewelry' Min=46 Max=426  Mean=124.725806452 StdDev=64.4336791705
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_jewelry_ConstantTrend_residue_zeroCycle_residue_NoAR' [ConstantTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_jewelry_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_jewelry_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_jewelry_ConstantTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.3201 MAPE_Forecast=0.2299 MAPE_Test=0.1537
INFO:pyaf.std:MODEL_L2 L2_Fit=48.9025823971 L2_Forecast=111.809540614 L2_Test=26.8732739271
INFO:pyaf.std:MODEL_COMPLEXITY 0
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST jewelry 0.2299










    
















    






INFO:pyaf.std:START_TRAINING 'unemp.cci'










    






analyzing  unemp.cci.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'unemp.cci' 2.176116466522217
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1997.41666667 TimeMax=2003.66666667 TimeDelta=0.0833333333333 Estimation = (0 , 76) Validation = (76 , 96) Test = (96 , 100) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_unemp.cci' Min=0 Max=1  Mean=0.49 StdDev=0.499899989998
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_unemp.cci_ConstantTrend_residue_zeroCycle_residue_AR(25)' [ConstantTrend + NoCycle + AR(25)]
INFO:pyaf.std:TREND_DETAIL '_unemp.cci_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_unemp.cci_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_unemp.cci_ConstantTrend_residue_zeroCycle_residue_AR(25)' [AR(25)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0627 MAPE_Forecast=0.0 MAPE_Test=0.0
INFO:pyaf.std:MODEL_L2 L2_Fit=0.122911295077 L2_Forecast=3.9876729474e-08 L2_Test=3.9876729474e-08
INFO:pyaf.std:MODEL_COMPLEXITY 25
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag1 0.599029115614
INFO:pyaf.std:AR_MODEL_COEFF 2 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag2 0.228808761914
INFO:pyaf.std:AR_MODEL_COEFF 3 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag3 0.0873971701272
INFO:pyaf.std:AR_MODEL_COEFF 4 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag4 0.033382748468
INFO:pyaf.std:AR_MODEL_COEFF 5 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag5 0.0127510752769
INFO:pyaf.std:AR_MODEL_COEFF 6 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag6 0.00487047736268
INFO:pyaf.std:AR_MODEL_COEFF 7 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag7 0.00186035681115
INFO:pyaf.std:AR_MODEL_COEFF 8 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag8 0.000710593070753
INFO:pyaf.std:AR_MODEL_COEFF 9 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag9 0.000271422401114
INFO:pyaf.std:AR_MODEL_COEFF 10 _unemp.cci_ConstantTrend_residue_zeroCycle_residue_Lag10 0.000103674132587
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST unemp.cci 0.0










    
















    






INFO:pyaf.std:START_TRAINING 'visitors'










    






analyzing  visitors.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'visitors' 2.906276226043701
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1985.33333333 TimeMax=2000.91666667 TimeDelta=0.0833333333334 Estimation = (0 , 188) Validation = (188 , 236) Test = (236 , 240) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_visitors' Min=75.4 Max=593.1  Mean=288.185833333 StdDev=115.117105445
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_visitors_ConstantTrend_residue_bestCycle_byL2_residue_AR(60)' [ConstantTrend + Cycle_Length_12 + AR(60)]
INFO:pyaf.std:TREND_DETAIL '_visitors_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_visitors_ConstantTrend_residue_bestCycle_byL2' [Cycle_Length_12]
INFO:pyaf.std:AUTOREG_DETAIL '_visitors_ConstantTrend_residue_bestCycle_byL2_residue_AR(60)' [AR(60)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0469 MAPE_Forecast=0.0586 MAPE_Test=0.059
INFO:pyaf.std:MODEL_L2 L2_Fit=13.2459051682 L2_Forecast=28.4450781044 L2_Test=29.4693654141
INFO:pyaf.std:MODEL_COMPLEXITY 68
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag1 0.524116249666
INFO:pyaf.std:AR_MODEL_COEFF 2 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag24 0.301893759658
INFO:pyaf.std:AR_MODEL_COEFF 3 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag34 -0.289752823013
INFO:pyaf.std:AR_MODEL_COEFF 4 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag12 0.243638366195
INFO:pyaf.std:AR_MODEL_COEFF 5 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag25 -0.22860431585
INFO:pyaf.std:AR_MODEL_COEFF 6 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag2 0.20967386029
INFO:pyaf.std:AR_MODEL_COEFF 7 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag35 0.202223439619
INFO:pyaf.std:AR_MODEL_COEFF 8 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag31 -0.169711202098
INFO:pyaf.std:AR_MODEL_COEFF 9 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag28 -0.155497505707
INFO:pyaf.std:AR_MODEL_COEFF 10 _visitors_ConstantTrend_residue_bestCycle_byL2_residue_Lag36 0.151736531677
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST visitors 0.0586










    
















    






analyzing  carparts.csv










    






INFO:pyaf.std:START_TRAINING 'carparts'
INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'carparts' 1.828833818435669
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1998.0 TimeMax=2001.0 TimeDelta=0.0833333333333 Estimation = (0 , 37) Validation = (37 , 47) Test = (47 , 51) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='Diff_carparts' Min=-5.0 Max=5.0  Mean=0.0196078431373 StdDev=1.85229265547
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'Diff_'
INFO:pyaf.std:BEST_DECOMPOSITION  'Diff_carparts_LinearTrend_residue_zeroCycle_residue_NoAR' [LinearTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL 'Diff_carparts_LinearTrend' [LinearTrend]
INFO:pyaf.std:CYCLE_DETAIL 'Diff_carparts_LinearTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'Diff_carparts_LinearTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.593 MAPE_Forecast=0.3962 MAPE_Test=0.818
INFO:pyaf.std:MODEL_L2 L2_Fit=1.9054663567 L2_Forecast=1.09333765889 L2_Test=0.970913494537
INFO:pyaf.std:MODEL_COMPLEXITY 48
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST carparts 0.3962










    
















    






INFO:pyaf.std:START_TRAINING 'freight'










    






analyzing  freight.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'freight' 1.7231676578521729
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1947 TimeMax=1980 TimeDelta=1.0 Estimation = (0 , 34) Validation = (34 , 43) Test = (43 , 47) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_freight' Min=298.1 Max=6243.1  Mean=2177.34680851 StdDev=1474.62692939
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_freight_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_freight_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_freight_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_freight_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.5466 MAPE_Forecast=0.5883 MAPE_Test=0.2464
INFO:pyaf.std:MODEL_L2 L2_Fit=1692.4475095 L2_Forecast=503.730324678 L2_Test=532.543080417
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST freight 0.5883










    
















    






INFO:pyaf.std:START_TRAINING 'mcopper'










    






analyzing  mcopper.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'mcopper' 4.500873565673828
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1960.0 TimeMax=1997.25 TimeDelta=0.0833333333333 Estimation = (0 , 448) Validation = (448 , 560) Test = (560 , 564) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_mcopper' Min=216.6 Max=4306.0  Mean=997.810460993 StdDev=606.066125451
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_mcopper_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_mcopper_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_mcopper_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_mcopper_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0479 MAPE_Forecast=0.0404 MAPE_Test=0.0432
INFO:pyaf.std:MODEL_L2 L2_Fit=63.9234628845 L2_Forecast=122.134676836 L2_Test=202.289025901
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST mcopper 0.0404










    
















    






INFO:pyaf.std:START_TRAINING 'usgdp'










    






analyzing  usgdp.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'usgdp' 2.7824032306671143
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1947.0 TimeMax=1993.25 TimeDelta=0.25 Estimation = (0 , 186) Validation = (186 , 233) Test = (233 , 237) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_usgdp' Min=1568.0 Max=11403.6  Mean=5168.46919831 StdDev=2763.16681343
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_usgdp_Lag1Trend_residue_zeroCycle_residue_NoAR' [Lag1Trend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL '_usgdp_Lag1Trend' [Lag1Trend]
INFO:pyaf.std:CYCLE_DETAIL '_usgdp_Lag1Trend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_usgdp_Lag1Trend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0111 MAPE_Forecast=0.0084 MAPE_Test=0.009
INFO:pyaf.std:MODEL_L2 L2_Fit=51.8574255112 L2_Forecast=87.1605488544 L2_Test=108.521553159
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST usgdp 0.0084










    
















    






INFO:pyaf.std:START_TRAINING 'xrates'










    






analyzing  xrates.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'xrates' 2.0227911472320557
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=2000.0 TimeMax=2004.75 TimeDelta=0.0833333333333 Estimation = (0 , 58) Validation = (58 , 73) Test = (73 , 77) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='Diff_xrates' Min=-0.0315 Max=0.0276  Mean=0.000151948051948 StdDev=0.0105911456573
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'Diff_'
INFO:pyaf.std:BEST_DECOMPOSITION  'Diff_xrates_ConstantTrend_residue_zeroCycle_residue_NoAR' [ConstantTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL 'Diff_xrates_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL 'Diff_xrates_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'Diff_xrates_ConstantTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0679 MAPE_Forecast=0.0231 MAPE_Test=0.0165
INFO:pyaf.std:MODEL_L2 L2_Fit=0.0301692026863 L2_Forecast=0.0122010937893 L2_Test=0.00793568042333
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST xrates 0.0231










    
















    






INFO:pyaf.std:START_TRAINING 'djiclose'










    






analyzing  djiclose.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'djiclose' 7.152137994766235
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1928.83333333 TimeMax=1991.75 TimeDelta=0.0833333333333 Estimation = (0 , 756) Validation = (756 , 946) Test = (946 , 950) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='CumSum_djiclose' Min=-145.87 Max=539.57  Mean=209.252263158 StdDev=170.913496014
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'CumSum_'
INFO:pyaf.std:BEST_DECOMPOSITION  'CumSum_djiclose_ConstantTrend_residue_zeroCycle_residue_NoAR' [ConstantTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL 'CumSum_djiclose_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL 'CumSum_djiclose_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'CumSum_djiclose_ConstantTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=1.0089 MAPE_Forecast=1.0 MAPE_Test=1.0
INFO:pyaf.std:MODEL_L2 L2_Fit=7.26607495003 L2_Forecast=4.08732508076 L2_Test=2.87330384749
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST djiclose 1.0










    
















    






INFO:pyaf.std:START_TRAINING 'frexport'










    






analyzing  frexport.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'frexport' 1.6417994499206543
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1.0 TimeMax=6.75 TimeDelta=0.25 Estimation = (0 , 24) Validation = (0 , 24) Test = (0 , 24) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_frexport' Min=341 Max=854  Mean=548.708333333 StdDev=137.44649843
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_frexport_ConstantTrend_residue_zeroCycle_residue_AR(6)' [ConstantTrend + NoCycle + AR(6)]
INFO:pyaf.std:TREND_DETAIL '_frexport_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_frexport_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_frexport_ConstantTrend_residue_zeroCycle_residue_AR(6)' [AR(6)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0459 MAPE_Forecast=0.0459 MAPE_Test=0.0459
INFO:pyaf.std:MODEL_L2 L2_Fit=28.8831856505 L2_Forecast=28.8831856505 L2_Test=28.8831856505
INFO:pyaf.std:MODEL_COMPLEXITY 6
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _frexport_ConstantTrend_residue_zeroCycle_residue_Lag4 1.1538764942
INFO:pyaf.std:AR_MODEL_COEFF 2 _frexport_ConstantTrend_residue_zeroCycle_residue_Lag1 0.525740647775
INFO:pyaf.std:AR_MODEL_COEFF 3 _frexport_ConstantTrend_residue_zeroCycle_residue_Lag5 -0.51350224395
INFO:pyaf.std:AR_MODEL_COEFF 4 _frexport_ConstantTrend_residue_zeroCycle_residue_Lag6 -0.17146324926
INFO:pyaf.std:AR_MODEL_COEFF 5 _frexport_ConstantTrend_residue_zeroCycle_residue_Lag3 -0.0358585026967
INFO:pyaf.std:AR_MODEL_COEFF 6 _frexport_ConstantTrend_residue_zeroCycle_residue_Lag2 0.0354687995945
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST frexport 0.0459










    
















    






INFO:pyaf.std:START_TRAINING 'msales'










    






analyzing  msales.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'msales' 1.7027981281280518
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1.0 TimeMax=3.0 TimeDelta=0.0833333333333 Estimation = (0 , 25) Validation = (25 , 32) Test = (32 , 36) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='Diff_msales' Min=-1.0 Max=1.0  Mean=0.0 StdDev=0.57735026919
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE 'Diff_'
INFO:pyaf.std:BEST_DECOMPOSITION  'Diff_msales_ConstantTrend_residue_zeroCycle_residue_NoAR' [ConstantTrend + NoCycle + NoAR]
INFO:pyaf.std:TREND_DETAIL 'Diff_msales_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL 'Diff_msales_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL 'Diff_msales_ConstantTrend_residue_zeroCycle_residue_NoAR' [NoAR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0 MAPE_Forecast=0.0 MAPE_Test=0.0
INFO:pyaf.std:MODEL_L2 L2_Fit=0.8 L2_Forecast=0.654653670708 L2_Test=0.0
INFO:pyaf.std:MODEL_COMPLEXITY 32
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST msales 0.0










    
















    






INFO:pyaf.std:START_TRAINING 'usnetelec'










    






analyzing  usnetelec.csv










    






INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'usnetelec' 1.8384170532226562
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=1949 TimeMax=1988 TimeDelta=1.0 Estimation = (0 , 40) Validation = (40 , 51) Test = (51 , 55) Horizon=4
INFO:pyaf.std:SIGNAL_DETAIL SignalVariable='_usnetelec' Min=296.1 Max=3858.5  Mean=1972.06 StdDev=1119.35555706
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_usnetelec_ConstantTrend_residue_zeroCycle_residue_AR(13)' [ConstantTrend + NoCycle + AR(13)]
INFO:pyaf.std:TREND_DETAIL '_usnetelec_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_usnetelec_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_usnetelec_ConstantTrend_residue_zeroCycle_residue_AR(13)' [AR(13)]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.0218 MAPE_Forecast=0.0254 MAPE_Test=0.035
INFO:pyaf.std:MODEL_L2 L2_Fit=29.2038352035 L2_Forecast=93.3272324545 L2_Test=152.044717185
INFO:pyaf.std:MODEL_COMPLEXITY 13
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag1 1.57032624776
INFO:pyaf.std:AR_MODEL_COEFF 2 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag2 -0.883149965431
INFO:pyaf.std:AR_MODEL_COEFF 3 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag3 0.616586682143
INFO:pyaf.std:AR_MODEL_COEFF 4 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag7 0.435408114064
INFO:pyaf.std:AR_MODEL_COEFF 5 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag10 0.304877387419
INFO:pyaf.std:AR_MODEL_COEFF 6 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag9 -0.294386706028
INFO:pyaf.std:AR_MODEL_COEFF 7 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag8 -0.264322853028
INFO:pyaf.std:AR_MODEL_COEFF 8 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag6 -0.22439896756
INFO:pyaf.std:AR_MODEL_COEFF 9 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag4 -0.19655667889
INFO:pyaf.std:AR_MODEL_COEFF 10 _usnetelec_ConstantTrend_residue_zeroCycle_residue_Lag5 -0.0558237042079
INFO:pyaf.std:AR_MODEL_DETAIL_END










    






PERFORMANCE MAPE_FORECAST usnetelec 0.0254










    























In [ ]:



    






    











In [6]:



    


lDict = lEngine_djiclose.getPlotsAsDict()



    


















    






/usr/lib/python3/dist-packages/matplotlib/__init__.py:1403: UserWarning:  This call to matplotlib.use() has no effect
because the backend has already been chosen;
matplotlib.use() must be called *before* pylab, matplotlib.pyplot,
or matplotlib.backends is imported for the first time.

  warnings.warn(_use_error_msg)

















In [7]:



    


lDict.keys()



    


















    
Out[7]:








dict_keys(['Forecast', 'Trend', 'Cycle', 'Prediction_Intervals', 'AR'])

















In [8]:



    


from IPython.display import Image
from IPython.display import display, HTML 
prefix = '<img src="data:image/png;base64,'
uri = prefix + lDict['Prediction_Intervals'] + '">'
HTML(uri)



    


















    
Out[8]:























In [ ]:

Content source: antoinecarme/pyaf

Similar notebooks:

notebook.community | gallery | about