In [16]:

    

from scipy.stats import norm
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
%matplotlib inline


    

In [6]:

    

mae = 'test_neg_mean_absolute_error'
var = 'test_explained_variance'
acc = 'test_accuracy'
alg = 'algorithm'


    

In [8]:

    

#results/19jul2018/
low_rxtr_scr = pd.read_csv('../trainset_1_fissact_reactor_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})
low_burn_scr = pd.read_csv('../trainset_1_fissact_burnup_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})
low_enri_scr = pd.read_csv('../trainset_1_fissact_enrichment_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})
low_cool_scr = pd.read_csv('../trainset_1_fissact_cooling_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})

low_burn_scr['Score'] = low_burn_scr[var]
low_cool_scr['Score'] = low_cool_scr[var]
low_enri_scr['Score'] = low_enri_scr[var]
low_rxtr_scr['Score'] = low_rxtr_scr[acc]

low_burn_scr['MAE'] = low_burn_scr[mae]
low_cool_scr['MAE'] = low_cool_scr[mae]
low_enri_scr['MAE'] = low_enri_scr[mae]

#rxtr_scr = pd.read_csv('../results/19jul2018/trainset_2_fissact_reactor_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})
#burn_scr = pd.read_csv('../results/19jul2018/trainset_2_fissact_burnup_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})
#enri_scr = pd.read_csv('../results/19jul2018/trainset_2_fissact_enrichment_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})
#cool_scr = pd.read_csv('../trainset_2_fissact_cooling_scores.csv').rename(columns = {'Unnamed: 0':'CV_fold'})


    

    
Out[8]:







  

    

      

      
CV_fold
      
fit_time
      
score_time
      
test_explained_variance
      
test_neg_mean_absolute_error
      
algorithm
      
Score
      
MAE
    
  
  

    

      
27
      
7
      
0.755002
      
0.073503
      
0.999707
      
-35.348563
      
svr
      
0.999707
      
-35.348563
    
    

      
17
      
7
      
0.003636
      
0.001101
      
1.000000
      
-0.465195
      
rr
      
1.000000
      
-0.465195
    
    

      
21
      
1
      
0.800573
      
0.073401
      
0.999584
      
-37.529184
      
svr
      
0.999584
      
-37.529184
    
    

      
2
      
2
      
0.003336
      
0.011704
      
0.999967
      
-7.611162
      
knn
      
0.999967
      
-7.611162
    
    

      
3
      
3
      
0.003287
      
0.011838
      
0.999974
      
-7.214527
      
knn
      
0.999974
      
-7.214527
    
    

      
4
      
4
      
0.003224
      
0.011794
      
0.999926
      
-14.549559
      
knn
      
0.999926
      
-14.549559
    
    

      
22
      
2
      
0.797616
      
0.072507
      
0.999660
      
-26.040613
      
svr
      
0.999660
      
-26.040613
    
    

      
29
      
9
      
0.746682
      
0.072487
      
0.999568
      
-37.506451
      
svr
      
0.999568
      
-37.506451
    
    

      
28
      
8
      
0.758397
      
0.073320
      
0.999722
      
-37.694791
      
svr
      
0.999722
      
-37.694791
    
    

      
1
      
1
      
0.003313
      
0.011503
      
0.999956
      
-7.939545
      
knn
      
0.999956
      
-7.939545
    
  

In [15]:

    

# knn
low_knn_b = low_burn_scr['Score'].loc[low_burn_scr['algorithm']=='knn']
low_knn_c = low_cool_scr['Score'].loc[low_cool_scr['algorithm']=='knn']
low_knn_e = low_enri_scr['Score'].loc[low_enri_scr['algorithm']=='knn']
low_knn_r = low_rxtr_scr['Score'].loc[low_rxtr_scr['algorithm']=='knn']
# rr
low_rr_b = low_burn_scr['Score'].loc[low_burn_scr['algorithm']=='rr']
low_rr_c = low_cool_scr['Score'].loc[low_cool_scr['algorithm']=='rr']
low_rr_e = low_enri_scr['Score'].loc[low_enri_scr['algorithm']=='rr']
low_rr_r = low_rxtr_scr['Score'].loc[low_rxtr_scr['algorithm']=='rr']
# svr
low_svr_b = low_burn_scr['Score'].loc[low_burn_scr['algorithm']=='svr']
low_svr_c = low_cool_scr['Score'].loc[low_cool_scr['algorithm']=='svr']
low_svr_e = low_enri_scr['Score'].loc[low_enri_scr['algorithm']=='svr']
low_svr_r = low_rxtr_scr['Score'].loc[low_rxtr_scr['algorithm']=='svr']


    

In [32]:

    

parameters = norm.fit(low_knn_e)
x = np.linspace(0, 1, 100)
normal_pdf = norm.pdf(x)
fitted_pdf = norm.pdf(x,loc = parameters[0],scale = parameters[1])


    

In [33]:

    

plt.plot(x,fitted_pdf,"red",label="Fitted normal dist",linestyle="dashed", linewidth=2)
plt.plot(x,normal_pdf,"blue",label="Normal dist", linewidth=2)
#plt.hist(low_svr_b,normed=1,color="cyan",alpha=.3) #alpha, from 0 (transparent) to 1 (opaque)
plt.title("Normal distribution fitting")
plt.legend(loc=2)
#plt.ylim((0, 1))
plt.show()


    

In [ ]:

    

burn = burn_scr.loc[:, [alg, var]]
cool = cool_scr.loc[:, [alg, var]]
enri = enri_scr.loc[:, [alg, var]]
rxtr = rxtr_scr.loc[:, [alg, acc]]

burn['Score'] = burn[var]
cool['Score'] = cool[var]
enri['Score'] = enri[var]
rxtr['Score'] = rxtr[acc]


    

In [ ]:

    

low_burn = low_burn_scr.loc[:, [alg, mae]]
low_cool = low_cool_scr.loc[:, [alg, mae]]
low_enri = low_enri_scr.loc[:, [alg, mae]]

burn = burn_scr.loc[:, [alg, mae]]
cool = cool_scr.loc[:, [alg, mae]]
enri = enri_scr.loc[:, [alg, mae]]