def optimize_logit_for(self, pair):
# extract the right images from the set
[a,b] = [int(pair[0]), int(pair[1])]
data = np.array(pd.read_csv('./data/kaggle/train.csv', header=0)).astype('float64')
data_t = data[0:, :]
data_a = data_t[data_t[:,0]==a]
data_b = data_t[data_t[:,0]==b]
data_ab = np.vstack([data_a, data_b])
xt = data_ab[:, 1:]
yt = data_ab[:, 0].astype('int8')
yt = np.atleast_2d(yt).T
print 'xt shape', xt.shape
# Perform logistic regression
xt2 = np.column_stack([np.ones((xt.shape[0],1)), xt])
yt2 = yt.flatten()
count = 0
for i in range(yt2.size):
if yt2[i]==a:
yt2[i] = 1
count +=1
else:
yt2[i] = 0
ini_thetas = 0.005*np.random.rand(xt2.shape[1],1)
L = 1e+5
opt_thetas = logit.optimizeThetas(ini_thetas, xt2, yt2, L, visual=False)
return opt_thetas
def test_optimize_thetas_reg():
l = Logysterical(tr, None)
ini_thetas = np.zeros((l.tr["X"].shape[1], 1))
pp = l.optimizeThetas(ini_thetas)
print pp
diff = pp - np.array(
[
1.273005,
0.624876,
1.177376,
-2.020142,
-0.912616,
-1.429907,
0.125668,
-0.368551,
-0.360033,
-0.171068,
-1.460894,
-0.052499,
-0.618889,
-0.273745,
-1.192301,
-0.240993,
-0.207934,
-0.047224,
-0.278327,
-0.296602,
-0.453957,
-1.045511,
0.026463,
-0.294330,
0.014381,
-0.328703,
-0.143796,
-0.924883,
]
)
for e in diff:
assert e < 0.01
def check_accuracy(self):
logit_thetas = {}
soft_thetas = np.array(pd.read_csv('./data/kaggle/optimized_thetas.csv', header=None))
soft_thetas = soft_thetas.reshape(self.LABS, -1)
h = soft.h(soft_thetas, self.xe)
m = h.shape[0]
misses = 0.00
count = 0.0
for i in range(m):
true_label = self.ye[i,0]
[ml_1, ml_2] = h[i,:].argsort()[-2:][::-1] # 1st and 2nd model choices
p1,p2 = h[i,:][ml_1], h[i,:][ml_2]
right_order = True
if ml_1 > ml_2:
right_order = False
s = `ml_2`+`ml_1`
else:
s = `ml_1`+`ml_2`
if p1<0.99 and p2>0.01:
if s not in logit_thetas:
count +=1
logit_thetas[s] = self.optimize_logit_for(s)
l_t = logit_thetas[s]
logix = np.hstack([1, self.xe[i,:]])
p = logit.h(l_t, logix)
if (p>0.5):
prediction = (ml_1 if right_order else ml_2)
else:
prediction = (ml_2 if right_order else ml_1)
else:
prediction = ml_1
#print prediction, true_label
if prediction!=true_label:
misses +=1.0
print 'misses', misses
print 'logit thetas searched', count
acc = 1 - misses/m
print 'accuracy:', acc
pass
def test_model_submit(self):
logit_thetas = {}
soft_thetas = np.array(pd.read_csv('./data/kaggle/submit_optimized_thetas.csv', header=None))
soft_thetas = soft_thetas.reshape(self.LABS, -1)
m, n = self.x_test.shape
h = soft.h(soft_thetas, self.x_test)
predictions = np.zeros((m,2))
for i in range(m):
[ml_1, ml_2] = h[i,:].argsort()[-2:][::-1] # 1st and 2nd model choices
p1,p2 = h[i,:][ml_1], h[i,:][ml_2]
right_order = True
if ml_1 > ml_2:
right_order = False
s = `ml_2`+`ml_1`
else:
s = `ml_1`+`ml_2`
if p1<0.99 and p2>0.01:
if s not in logit_thetas:
logit_thetas[s] = self.optimize_logit_for(s)
l_t = logit_thetas[s]
logix = np.hstack([1, self.x_test[i,:]])
p = logit.h(l_t, logix)
if (p>0.5):
predictions[i,:] = ([i+1, ml_1] if right_order else [i+1, ml_2])
else:
predictions[i,:] = ([i+1, ml_2] if right_order else [i+1, ml_1])
else:
predictions[i,:]=[i+1, ml_1]
print 'To submitt add header: ImageId,Label'
print predictions[0:10,:]
np.savetxt('./data/kaggle/predictions_2steps.csv', predictions, fmt='%i,%i')
pass
def test_J_with_lam_and_zero_theta():
l = Logysterical(tr)
theta = np.zeros((l.tr["X"].shape[1], 1))
assert l.j(theta) - 0.69314718056 < epsilon
