def GetProba(drive,driverID,tripInd):
driverDir = '/home/user1/Desktop/SharedFolder/Kaggle/DriversCleaned/'+str(driverID)
df = pd.read_csv(driverDir+'_' + str(tripInd)+'.csv')
trip = Trip(driverID,tripInd,df)
trip.getSpeed()
trip.getAcc()
#trip.getRadius()
#trip.getCacc()
trip.getFeatures()
X=trip.features[['v','acc']]
probas = np.zeros((X.shape[0],drive.shape[0]))
for i in range(drive.shape[0]):
probas[:,i]=multivariate_normal.pdf(X, mean=array(drive.ix[i,:2]), cov=[array(drive.ix[i,2:4]),array(drive.ix[i,4:])])
probas=np.max(probas,axis=1)
return probas.mean()
def GetProba(clf,driverID,tripID):
#print driverID,tripID
driverDir = '/home/user1/Desktop/SharedFolder/Kaggle/DriversCleaned/'+str(driverID)
df = pd.read_csv(driverDir+'_' + str(tripInd)+'.csv')
trip = Trip(driverID,tripInd,df)
trip.getSpeed()
trip.getAcc()
#trip.getRadius()
#trip.getCacc()
trip.getFeatures()
X=trip.features[['v','acc']]
probas = np.zeros(X.shape[0])
for i in range(X.shape[0]):
probas[i]=clf.score(X.loc[i])
print probas.mean()
return probas.mean()
def GetFeatures(driverID,j):
driverDir = '/home/user1/Desktop/SharedFolder/Kaggle/DriversCleaned/'+str(driverID)
tripFiles = range(1,201)
X = pd.DataFrame(columns=selectedCols)
for index,tripID in enumerate(tripFiles):
#print index,tripID
trip = Trip(driverID,tripID,pd.read_csv(driverDir+'_' + str(tripID) + '.csv'))
trip.getSpeed()
trip.getAcc()
#trip.getRadius()
#trip.getCacc()
trip.getFeatures()
'''z=array(list(set(np.asarray([range(x-5,x+5) for x in (trip.features.v<vlim[0]).nonzero()[0]]).flatten())))
z=z[z<trip.features.shape[0]]
z=z[z>=0]
#z=array(list(set(range(trip.features.shape[0]))-set(z)))
Xz=trip.features.loc[z]
Xz=Xz.reset_index(drop=True)
Xz=Xz.loc[Xz.v!=0]
Xz=Xz.reset_index(drop=True)
X = X.append(Xz)'''
X = X.append(trip.features)
X=X.reset_index(drop=True)
X=X[(X.v<vlim[1]) & (X.v>vlim[0])]
X=X[(X.acc<clim[1]) & (X.acc>clim[0])]
X=X.reset_index(drop=True)
clf=GetGmm(np.asanyarray(X[['v','acc']]))
cos = SaveGmm(clf)
cos.to_csv('/home/user1/Desktop/SharedFolder/Kaggle/FeaturesCleaned/GMM/All/' + str(driverID) + '.csv', index=False)
#del cos
return 0
import matplotlib.pyplot as plt
import matplotlib as mpl
from sklearn.mixture import GMM
from sklearn.neighbors import KernelDensity as kde
from scipy.stats import multivariate_normal
from scipy.spatial.distance import mahalanobis
# <codecell>
driverID = 1
driverDir = '/home/user1/Desktop/SharedFolder/Kaggle/DriversCleaned/'+str(driverID)
tripInd = 1
df = pd.read_csv(driverDir+'_' + str(tripInd)+'.csv')
trip = Trip(driverID,tripInd,df)
trip.getSpeed()
trip.getAcc()
#trip.getRadius()
#trip.getCacc()
trip.getFeatures()
X=trip.features[['v','acc']]
# <codecell>
def GetProba(drive,driverID,tripInd):
driverDir = '/home/user1/Desktop/SharedFolder/Kaggle/DriversCleaned/'+str(driverID)
df = pd.read_csv(driverDir+'_' + str(tripInd)+'.csv')
trip = Trip(driverID,tripInd,df)
trip.getSpeed()
trip.getAcc()
#trip.getRadius()
#trip.getCacc()
