In [62]:

    

import pandas as pd
import numpy as np
import matplotlib.pylab as plt
from statsmodels.tsa.ar_model import AR
from sklearn.metrics import r2_score
%matplotlib inline
 
plt.rcParams['figure.figsize']=(20,10)
plt.style.use('ggplot')


    

In [3]:

    

data = pd.read_csv('DEOK_hourly.csv')
data['Datetime']=pd.to_datetime(data['Datetime'])
data.set_index('Datetime', inplace=True)


    

In [4]:

    

data.head()


    

    
Out[4]:







  

    

      

      
DEOK_MW
    
    

      
Datetime
      

    
  
  

    

      
2012-12-31 01:00:00
      
2945.0
    
    

      
2012-12-31 02:00:00
      
2868.0
    
    

      
2012-12-31 03:00:00
      
2812.0
    
    

      
2012-12-31 04:00:00
      
2812.0
    
    

      
2012-12-31 05:00:00
      
2860.0
    
  

In [5]:

    

data.plot()


    

    
Out[5]:





<matplotlib.axes._subplots.AxesSubplot at 0x7efbc832db50>






    

In [9]:

    

data['DEOK_MW'].hist(bins=20)


    

    
Out[9]:





<matplotlib.axes._subplots.AxesSubplot at 0x7efbc41bfcd0>






    

In [39]:

    

one, two, three = np.split(
        data['DEOK_MW'].sample(
        frac=1), [int(.25*len(data['DEOK_MW'])),
        int(.75*len(data['DEOK_MW']))])


    

In [34]:

    

mean1, mean2, mean3 = one.mean(), two.mean(), three.mean()
var1, var2, var3 = one.var(), two.var(), three.var()


    

In [40]:

    

print mean1, mean2, mean3
print var1, var2, var3


    

    




3093.27497575 3107.45445099 3112.20124697
353154.655416 363558.421407 358899.692558

In [41]:

    

from statsmodels.tsa.stattools import adfuller


    

In [44]:

    

adf_test = adfuller(data['DEOK_MW'])


    

In [45]:

    

adf_test


    

    
Out[45]:





(-14.913267801069782,
 1.4477674072055658e-27,
 57,
 57681,
 {'1%': -3.4304633751328555,
  '10%': -2.5667966716717614,
  '5%': -2.8615901096273602},
 669611.23911962728)

In [48]:

    

print "ADF = " + str(adf_test[0])
print "p-value = " +str(adf_test[1])


    

    




ADF = -14.9132678011
p-value = 1.44776740721e-27

In [50]:

    

pd.tools.plotting.lag_plot(data['DEOK_MW'])


    

    




/vagrant/pythondata/env/lib/python2.7/site-packages/ipykernel_launcher.py:1: FutureWarning: 'pandas.tools.plotting.lag_plot' is deprecated, import 'pandas.plotting.lag_plot' instead.
  """Entry point for launching an IPython kernel.






    
Out[50]:





<matplotlib.axes._subplots.AxesSubplot at 0x7efbcc9196d0>






    

In [149]:

    

#create train/test datasets
X = data['DEOK_MW'].dropna()

train_data = X[1:len(X)-12]
test_data = X[len(X)-12:]


    

In [150]:

    

#train the autoregression model
model = AR(train_data)
model_fitted = model.fit()


    

    




/vagrant/pythondata/env/local/lib/python2.7/site-packages/statsmodels/tsa/base/tsa_model.py:225: ValueWarning: A date index has been provided, but it has no associated frequency information and so will be ignored when e.g. forecasting.
  ' ignored when e.g. forecasting.', ValueWarning)

In [151]:

    

print('The lag value chose is: %s' % model_fitted.k_ar)


    

    




The lag value chose is: 59

In [152]:

    

# make predictions 
predictions = model_fitted.predict(
    start=len(train_data), 
    end=len(train_data) + len(test_data)-1, 
    dynamic=False)


# create a comparison dataframe
compare_df = pd.concat(
    [data['DEOK_MW'].reset_index().tail(12),
    predictions], axis=1).rename(
    columns={'DEOK_MW': 'actual', 0:'predicted'})
compare_df=compare_df[['actual', 'predicted']].dropna()


    

    




/vagrant/pythondata/env/local/lib/python2.7/site-packages/statsmodels/tsa/base/tsa_model.py:531: ValueWarning: No supported index is available. Prediction results will be given with an integer index beginning at `start`.
  ValueWarning)

In [153]:

    

predictions


    

    
Out[153]:





57726    3869.472137
57727    3805.973745
57728    3796.575978
57729    3797.108515
57730    3890.598661
57731    4097.286867
57732    4260.151854
57733    4310.203028
57734    4307.501185
57735    4252.261197
57736    4146.984356
57737    4053.067806
dtype: float64

In [ ]:

    




    

In [154]:

    

compare_df


    

    
Out[154]:







  

    

      

      
actual
      
predicted
    
  
  

    

      
57727
      
3865.0
      
3805.973745
    
    

      
57728
      
3824.0
      
3796.575978
    
    

      
57729
      
3766.0
      
3797.108515
    
    

      
57730
      
3776.0
      
3890.598661
    
    

      
57731
      
3885.0
      
4097.286867
    
    

      
57732
      
4200.0
      
4260.151854
    
    

      
57733
      
4393.0
      
4310.203028
    
    

      
57734
      
4426.0
      
4307.501185
    
    

      
57735
      
4419.0
      
4252.261197
    
    

      
57736
      
4355.0
      
4146.984356
    
    

      
57737
      
4224.0
      
4053.067806
    
  

In [155]:

    

compare_df.plot()


    

    
Out[155]:





<matplotlib.axes._subplots.AxesSubplot at 0x7efbbd7af0d0>






    

In [158]:

    

r2 = r2_score(compare_df.actual, compare_df.predicted)


    

In [160]:

    

r2


    

    
Out[160]:





0.75819470059241667

In [ ]: