In [1]:

    

import numpy as np
import pandas as pd

%matplotlib inline

# generate a daily signal covering one year 2016 in a pandas dataframe
N = 360
df_train = pd.DataFrame({"Date" : pd.date_range(start="2016-01-25", periods=N, freq='D'),
                         "Signal" : (np.arange(N)//40 + np.arange(N) % 21 + np.random.randn(N))})


    

In [2]:

    

df_train.head(12)


    

    
Out[2]:







  

    

      

      
Date
      
Signal
    
  
  

    

      
0
      
2016-01-25
      
-0.870635
    
    

      
1
      
2016-01-26
      
0.951762
    
    

      
2
      
2016-01-27
      
2.285937
    
    

      
3
      
2016-01-28
      
3.582260
    
    

      
4
      
2016-01-29
      
3.085772
    
    

      
5
      
2016-01-30
      
5.518716
    
    

      
6
      
2016-01-31
      
4.474009
    
    

      
7
      
2016-02-01
      
7.780780
    
    

      
8
      
2016-02-02
      
7.429573
    
    

      
9
      
2016-02-03
      
9.298334
    
    

      
10
      
2016-02-04
      
11.248414
    
    

      
11
      
2016-02-05
      
13.022808
    
  

In [3]:

    

df_train.tail(12)


    

    
Out[3]:







  

    

      

      
Date
      
Signal
    
  
  

    

      
348
      
2017-01-07
      
19.106093
    
    

      
349
      
2017-01-08
      
20.896310
    
    

      
350
      
2017-01-09
      
22.297185
    
    

      
351
      
2017-01-10
      
22.450376
    
    

      
352
      
2017-01-11
      
23.943530
    
    

      
353
      
2017-01-12
      
24.033267
    
    

      
354
      
2017-01-13
      
26.744537
    
    

      
355
      
2017-01-14
      
27.757442
    
    

      
356
      
2017-01-15
      
28.483932
    
    

      
357
      
2017-01-16
      
7.236476
    
    

      
358
      
2017-01-17
      
8.608582
    
    

      
359
      
2017-01-18
      
11.386710
    
  

In [4]:

    

df_train.plot('Date' , ['Signal'])


    

    
Out[4]:





<matplotlib.axes._subplots.AxesSubplot at 0x7fee289a7a30>






    

In [5]:

    

import pyaf.ForecastEngine as autof
# create a forecast engine. This is the main object handling all the operations
lEngine = autof.cForecastEngine()

# get the best time series model for predicting one week
lEngine.train(iInputDS = df_train, iTime = 'Date', iSignal = 'Signal', iHorizon = 7);
lEngine.getModelInfo() # => relative error 7% (MAPE)


    

    




INFO:pyaf.std:START_TRAINING 'Signal'
INFO:pyaf.std:END_TRAINING_TIME_IN_SECONDS 'Signal' 3.4868886470794678
INFO:pyaf.std:TIME_DETAIL TimeVariable='Date' TimeMin=2016-01-25T00:00:00.000000 TimeMax=2016-11-01T00:00:00.000000 TimeDelta=<DateOffset: days=1> Horizon=7
INFO:pyaf.std:SIGNAL_DETAIL_ORIG SignalVariable='Signal' Length=360  Min=-1.8166907883271852 Max=28.483931937748174  Mean=13.930487319531547 StdDev=6.838330205212032
INFO:pyaf.std:SIGNAL_DETAIL_TRANSFORMED TransformedSignalVariable='_Signal' Min=-1.8166907883271852 Max=28.483931937748174  Mean=13.930487319531547 StdDev=6.838330205212032
INFO:pyaf.std:BEST_TRANSOFORMATION_TYPE '_'
INFO:pyaf.std:BEST_DECOMPOSITION  '_Signal_ConstantTrend_residue_zeroCycle_residue_AR(64)' [ConstantTrend + NoCycle + AR]
INFO:pyaf.std:TREND_DETAIL '_Signal_ConstantTrend' [ConstantTrend]
INFO:pyaf.std:CYCLE_DETAIL '_Signal_ConstantTrend_residue_zeroCycle' [NoCycle]
INFO:pyaf.std:AUTOREG_DETAIL '_Signal_ConstantTrend_residue_zeroCycle_residue_AR(64)' [AR]
INFO:pyaf.std:MODEL_MAPE MAPE_Fit=0.4355 MAPE_Forecast=0.0738 MAPE_Test=0.0963
INFO:pyaf.std:MODEL_SMAPE SMAPE_Fit=0.1613 SMAPE_Forecast=0.0776 SMAPE_Test=0.1042
INFO:pyaf.std:MODEL_MASE MASE_Fit=0.5233 MASE_Forecast=0.5615 MASE_Test=0.2979
INFO:pyaf.std:MODEL_L1 L1_Fit=1.1991478592518148 L1_Forecast=1.153783468476882 L1_Test=1.4822005489079952
INFO:pyaf.std:MODEL_L2 L2_Fit=1.7379017776841135 L2_Forecast=1.494692508068081 L2_Test=1.8374021956503817
INFO:pyaf.std:MODEL_COMPLEXITY 64
INFO:pyaf.std:SIGNAL_TRANSFORMATION_DETAIL_START
INFO:pyaf.std:SIGNAL_TRANSFORMATION_MODEL_VALUES NoTransf None
INFO:pyaf.std:SIGNAL_TRANSFORMATION_DETAIL_END
INFO:pyaf.std:TREND_DETAIL_START
INFO:pyaf.std:CONSTANT_TREND ConstantTrend 12.840239575721542
INFO:pyaf.std:TREND_DETAIL_END
INFO:pyaf.std:CYCLE_MODEL_DETAIL_START
INFO:pyaf.std:ZERO_CYCLE_MODEL_VALUES _Signal_ConstantTrend_residue_zeroCycle 0.0 {}
INFO:pyaf.std:CYCLE_MODEL_DETAIL_END
INFO:pyaf.std:AR_MODEL_DETAIL_START
INFO:pyaf.std:AR_MODEL_COEFF 1 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag21 0.6495451430963992
INFO:pyaf.std:AR_MODEL_COEFF 2 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag1 0.5506630019690499
INFO:pyaf.std:AR_MODEL_COEFF 3 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag22 -0.32832538311217213
INFO:pyaf.std:AR_MODEL_COEFF 4 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag42 0.22677087830487136
INFO:pyaf.std:AR_MODEL_COEFF 5 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag2 0.19308134047715292
INFO:pyaf.std:AR_MODEL_COEFF 6 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag41 -0.17144721426695997
INFO:pyaf.std:AR_MODEL_COEFF 7 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag64 -0.14751182588829692
INFO:pyaf.std:AR_MODEL_COEFF 8 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag7 0.1378075845260201
INFO:pyaf.std:AR_MODEL_COEFF 9 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag23 -0.1371382948009829
INFO:pyaf.std:AR_MODEL_COEFF 10 _Signal_ConstantTrend_residue_zeroCycle_residue_Lag20 0.13168079640169356
INFO:pyaf.std:AR_MODEL_DETAIL_END

In [ ]:

    




    

In [6]:

    

lEngine.standardPlots()


    

    




INFO:pyaf.std:START_PLOTTING
INFO:pyaf.std:END_PLOTTING_TIME_IN_SECONDS 1.3545963764190674






    













    













    













    













    













    

In [7]:

    

# predict one week
df_forecast = lEngine.forecast(iInputDS = df_train, iHorizon = 7)
# list the columns of the forecast dataset
print(df_forecast.columns) #

# print the real forecasts
# Future dates : ['2017-01-19T00:00:00.000000000' '2017-01-20T00:00:00.000000000' '2017-01-21T00:00:00.000000000' '2017-01-22T00:00:00.000000000' '2017-01-23T00:00:00.000000000' '2017-01-24T00:00:00.000000000' '2017-01-25T00:00:00.000000000']
print(df_forecast['Date'].tail(7).values)

# signal forecast : [ 9.74934646  10.04419761  12.15136455  12.20369717  14.09607727 15.68086323  16.22296559]
print(df_forecast['Signal_Forecast'].tail(7).values)


    

    




INFO:pyaf.std:START_FORECASTING 'Signal'
INFO:pyaf.std:END_FORECAST_TIME_IN_SECONDS  'Signal' 0.4388258457183838






    




Index(['Date', 'Signal', 'row_number', 'Date_Normalized', '_Signal',
       '_Signal_ConstantTrend', '_Signal_ConstantTrend_residue',
       '_Signal_ConstantTrend_residue_zeroCycle',
       '_Signal_ConstantTrend_residue_zeroCycle_residue',
       '_Signal_ConstantTrend_residue_zeroCycle_residue_AR(64)',
       '_Signal_ConstantTrend_residue_zeroCycle_residue_AR(64)_residue',
       '_Signal_Trend', '_Signal_Trend_residue', '_Signal_Cycle',
       '_Signal_Cycle_residue', '_Signal_AR', '_Signal_AR_residue',
       '_Signal_TransformedForecast', 'Signal_Forecast',
       '_Signal_TransformedResidue', 'Signal_Residue',
       'Signal_Forecast_Lower_Bound', 'Signal_Forecast_Upper_Bound'],
      dtype='object')
['2017-01-19T00:00:00.000000000' '2017-01-20T00:00:00.000000000'
 '2017-01-21T00:00:00.000000000' '2017-01-22T00:00:00.000000000'
 '2017-01-23T00:00:00.000000000' '2017-01-24T00:00:00.000000000'
 '2017-01-25T00:00:00.000000000']
[11.39254404 12.17141899 11.12398491 13.23095585 12.71242861 12.98681181
 15.24008799]

In [ ]: