Energy Efficiency (2)

Live coding demo on May 19, 2018



In [31]:

    
%matplotlib notebook 

import json

import numpy as np 
import pandas as pd 
import seaborn as sns 
import yellowbrick as yb
import matplotlib.pyplot as plt 

from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import PolynomialFeatures
from sklearn.base import BaseEstimator, TransformerMixin

from sklearn.neural_network import MLPRegressor 
from sklearn.linear_model import RidgeCV, LassoCV, ElasticNetCV 
from sklearn.ensemble import GradientBoostingRegressor, RandomForestRegressor

from sklearn.model_selection import cross_val_score

Data Loading



In [9]:

    
data = pd.read_csv("../data/energy/energy.csv")
data.head()



In [15]:

    
with open("../data/energy/meta.json", "r") as f:
    meta = json.load(f)
    
columns = meta['feature_names'] + list(meta['target_names'].values())
data.columns = columns
data.head()









    Out[15]:







  
    
      
      relative compactness
      surface area
      wall area
      roof area
      overall height
      orientation
      glazing area
      glazing area distribution
      heating load
      cooling load
    
  
  
    
      0
      0.98
      514.5
      294.0
      110.25
      7.0
      2
      0.0
      0
      15.55
      21.33
    
    
      1
      0.98
      514.5
      294.0
      110.25
      7.0
      3
      0.0
      0
      15.55
      21.33
    
    
      2
      0.98
      514.5
      294.0
      110.25
      7.0
      4
      0.0
      0
      15.55
      21.33
    
    
      3
      0.98
      514.5
      294.0
      110.25
      7.0
      5
      0.0
      0
      15.55
      21.33
    
    
      4
      0.90
      563.5
      318.5
      122.50
      7.0
      2
      0.0
      0
      20.84
      28.28



In [41]:

    
data.shape









    Out[41]:





(768, 10)



In [16]:

    
heating = data[meta['target_names']['Y1']]
cooling = data[meta['target_names']['Y2']]
features = data[meta['feature_names']]



In [18]:

    
features.head()









    Out[18]:







  
    
      
      relative compactness
      surface area
      wall area
      roof area
      overall height
      orientation
      glazing area
      glazing area distribution
    
  
  
    
      0
      0.98
      514.5
      294.0
      110.25
      7.0
      2
      0.0
      0
    
    
      1
      0.98
      514.5
      294.0
      110.25
      7.0
      3
      0.0
      0
    
    
      2
      0.98
      514.5
      294.0
      110.25
      7.0
      4
      0.0
      0
    
    
      3
      0.98
      514.5
      294.0
      110.25
      7.0
      5
      0.0
      0
    
    
      4
      0.90
      563.5
      318.5
      122.50
      7.0
      2
      0.0
      0



In [29]:

    
_, axes = plt.subplots(1, 2, figsize=(9,6), sharey=True)

# Heating first 
h = sns.violinplot(data=heating, ax=axes[0])
h.set_ylabel("energy load")
h.set_xlabel("heating")

c = sns.violinplot(data=cooling, ax=axes[1])
c.set_xlabel("cooling")









    














    











    Out[29]:





Text(0.5,0,'cooling')



In [30]:

    
sns.boxplot(data=features)









    














    











    Out[30]:





<matplotlib.axes._subplots.AxesSubplot at 0x10fe00f28>

Linear and Polynomial Regressions



In [59]:

    
def build_poly(model, degree=1):
    if degree == 1:
        return Pipeline([
            ("std", StandardScaler()), 
            ("reg", model), 
        ])
    
    return Pipeline([
        ("std", StandardScaler()),
        ("poly", PolynomialFeatures(degree)), 
        ("reg", model), 
    ])


alphas = np.logspace(-6, 0, 200)



In [38]:

    
plt.plot(alphas)









    














    











    Out[38]:





[<matplotlib.lines.Line2D at 0x110afe5c0>]



In [43]:

    
# Ridge 
ridge_reg = build_poly(RidgeCV(alphas=alphas))
cross_val_score(ridge_reg, X=features, y=heating, cv=3)









    Out[43]:





array([0.81812578, 0.9189592 , 0.90371794])



In [56]:

    
ridge_reg.fit(features, heating)
print(ridge_reg._final_estimator.alpha_)
print(ridge_reg._final_estimator.coef_)









    



0.11623224686798518
[-6.70178939 -3.66740443  0.73499571 -3.93119914  7.33813269 -0.02607769
  2.65330353  0.31587548]



In [53]:

    
# Lasso 
lasso_reg = build_poly(LassoCV(alphas=alphas))
cross_val_score(lasso_reg, X=features, y=heating, cv=3)









    



/Users/benjamin/.pyenv/versions/3.6.2/Python.framework/Versions/3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.
  ConvergenceWarning)






    Out[53]:





array([0.81823595, 0.90881363, 0.90284712])



In [57]:

    
lasso_reg.fit(features, heating)
print(lasso_reg._final_estimator.alpha_)
print(lasso_reg._final_estimator.coef_)









    



1e-06
[-6.84351433 -7.22744064  2.4270302  -0.46314998  7.29881752 -0.02608064
  2.65371268  0.31583515]






    



/Users/benjamin/.pyenv/versions/3.6.2/Python.framework/Versions/3.6/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.
  ConvergenceWarning)



In [61]:

    
scores = []

for d in range(1, 10):
    reg = build_poly(RidgeCV(alphas=alphas), degree=d)
    scores.append(
        cross_val_score(reg, X=features, y=heating, cv=3).mean()
    )

scores = pd.Series(scores, index=range(1,10), name="$r^2$ of RidgeCV by degree")
scores.plot()









    














    











    Out[61]:





<matplotlib.axes._subplots.AxesSubplot at 0x110b5ab38>



In [ ]:



In [64]:

    
gb = build_poly(GradientBoostingRegressor())
cross_val_score(gb, X=features, y=heating, cv=3)









    Out[64]:





array([0.82543409, 0.94746752, 0.93229382])



In [67]:

    
gb.fit(features, heating)









    Out[67]:





Pipeline(memory=None,
     steps=[('std', StandardScaler(copy=True, with_mean=True, with_std=True)), ('reg', GradientBoostingRegressor(alpha=0.9, criterion='friedman_mse', init=None,
             learning_rate=0.1, loss='ls', max_depth=3, max_features=None,
             max_leaf_nodes=None, min_impurity_decrease=0.0,
        ...s=100, presort='auto', random_state=None,
             subsample=1.0, verbose=0, warm_start=False))])



In [72]:

    
gb._final_estimator.feature_importances_









    Out[72]:





array([0.23281328, 0.22199104, 0.16951871, 0.04807181, 0.06242108,
       0.04405539, 0.14126829, 0.0798604 ])



In [75]:

    
from sklearn.model_selection import train_test_split as tts
from yellowbrick.regressor import PredictionError

X_train, X_test, y_train, y_test = tts(features, heating, train_size=0.6)

oz = PredictionError(build_poly(GradientBoostingRegressor(n_estimators=10)))
oz.fit(X_train, y_train)
oz.score(X_test, y_test)
oz.poof()









    



/Users/benjamin/.pyenv/versions/3.6.2/Python.framework/Versions/3.6/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2026: FutureWarning: From version 0.21, test_size will always complement train_size unless both are specified.
  FutureWarning)

	X1	X2	X3	X4	X5	X6	Y1	Y2
0	0.98	514.5	294.0	110.25	7.0	2	15.55	21.33
1	0.98	514.5	294.0	110.25	7.0	3	15.55	21.33
2	0.98	514.5	294.0	110.25	7.0	4	15.55	21.33
3	0.98	514.5	294.0	110.25	7.0	5	15.55	21.33
4	0.90	563.5	318.5	122.50	7.0	2	20.84	28.28