In [1]:

    

%matplotlib inline
import PS1QLS
import matplotlib.pyplot as plt
import numpy as np
from sklearn.cross_validation import train_test_split
dataLoc = '../../'
pQSO = np.loadtxt(dataLoc+'pQSO/pSDSScolmag.txt')[:,2:]
lQSO = np.loadtxt(dataLoc+'lQSO/SDSScolmag.txt')[:,2:]
sinQSO = np.loadtxt(dataLoc+'sinQSO/sSDSScolmag.txt')[:,2:]
unlQSO = np.loadtxt(dataLoc+'unlQSO/nlSDSScolmag.txt')[:,2:]
unlQSO[:,3:5] = -unlQSO[:,3:5] #bug in WISE magnitudes for this file

duds = np.concatenate((pQSO,unlQSO,sinQSO),axis=0)
data = np.concatenate((lQSO,duds),axis=0) #all sims
truth = np.concatenate((np.ones(lQSO.shape[0]),np.zeros(duds.shape[0])),axis=0)


    

    




WARNING:root:Deprecation Warning: 'triangle' has been renamed to 'corner'. This shim should continue to work but you should use 'import corner' in new code. https://github.com/dfm/corner.py

In [2]:

    

train,test,truth_train,truth_test = train_test_split(data,truth,test_size=0.33)


    

In [3]:

    

trainds = PS1QLS.Dataset(train,truth_train)
trainds.noisify()
trainds.plot()


    

In [6]:

    

reload(PS1QLS)
RF = PS1QLS.Classifier(algorithm='RandomForest')
RF.train(trainds.data,trainds.truth,covmat=trainds.covmat)

testds = PS1QLS.Dataset(test,truth_test)
testds.noisify()

RF.test(testds.data,testds.covmat)
fpr_RF,tpr_RF = RF.make_roc(testds.truth)


    

    




Fitting 3 folds for each of 36 candidates, totalling 108 fits






    




[Parallel(n_jobs=-1)]: Done  53 tasks      | elapsed:   13.7s
[Parallel(n_jobs=-1)]: Done 108 out of 108 | elapsed:   40.7s finished






    

In [10]:

    

RF = PS1QLS.Classifier(algorithm='XD')

#RF.train(trainds.data,trainds.truth,covmat=trainds.covmat)
RF.load('xd.pkl')

testds = PS1QLS.Dataset(test,truth_test)
testds.noisify()

RF.test(testds.data,testds.covmat)
fpr_XD,tpr_XD = RF.make_roc(testds.truth)


    

In [13]:

    

plt.title('ROC Curve')
plt.plot(fpr_RF,tpr_RF,'r')
plt.plot(fpr_XD,tpr_XD,'b--')
plt.xlabel('FPR')
plt.ylabel('TPR')


    

    
Out[13]:





<matplotlib.text.Text at 0x10a57cc50>






    

In [ ]: