def _fit(self, X, y):
self._check(X, y)
assert (dim(y) == 1)
beta = zeros(shape(X)[1]) # row vector
X_T = matrix_transpose(X)
if self.fit_intercept:
beta[0] = sum(minus(reshape(y, -1), dot(X,
beta[1:]))) / (shape(X)[0])
for _ in range(self.max_iter):
print(_)
start = 1 if self.fit_intercept else 0
for j in range(start, len(beta)):
tmp_beta = [x for x in beta]
tmp_beta[j] = 0.0
r_j = minus(reshape(y, -1), dot(X, beta))
# r_j = minus(reshape(y,-1) , dot(X, tmp_beta))
arg1 = dot(X_T[j], r_j)
arg2 = self.alpha * shape(X)[0]
if sum(square(X_T[j])) != 0:
beta[j] = self._soft_thresholding_operator(
arg1, arg2) / sum(square(X_T[j]))
else:
beta[j] = 0
if self.fit_intercept:
beta[0] = sum(minus(reshape(y, -1), dot(
X, beta[1:]))) / (shape(X)[0])
return beta
def _score_calc(y, y_):
y_ = [int(round(i)) for i in y_]
numerator = sqrt(mean(square(minus(y, y_))))
denominator = sqrt(mean(square(y))) + sqrt(mean(square(y_)))
if denominator == 0:
return 0
else:
return 1 - (numerator / float(denominator))
def fit(self, X, y):
self._check(X, y)
if dim(y) == 1:
raw_X = X
if self.fit_intercept:
X = hstack([ones(shape(X)[0], 1), X])
beta = zeros(shape(X)[1]) # row vector
X_T = matrix_transpose(X)
if self.fit_intercept:
beta[0] = sum(minus(reshape(y, -1), dot(
raw_X, beta[1:]))) / (shape(X)[0])
for _ in range(self.max_iter):
start = 1 if self.fit_intercept else 0
for j in range(start, len(beta)):
tmp_beta = [x for x in beta]
tmp_beta[j] = 0.0
r_j = minus(reshape(y, -1), dot(X, beta))
# r_j = minus(reshape(y,-1) , dot(X, tmp_beta))
arg1 = dot(X_T[j], r_j)
arg2 = self.alpha * shape(X)[0]
if sum(square(X_T[j])) != 0:
beta[j] = self._soft_thresholding_operator(
arg1, arg2) / sum(square(X_T[j]))
else:
beta[j] = 0
if self.fit_intercept:
beta[0] = sum(
minus(reshape(y, -1), dot(
raw_X, beta[1:]))) / (shape(X)[0])
# # add whatch
# self.beta = beta
# self._whatch(raw_X,y)
if self.fit_intercept:
self.intercept_ = beta[0]
self.coef_ = beta[1:]
else:
self.coef_ = beta
self.beta = beta
return self
elif dim(y) == 2:
if self.fit_intercept:
X = hstack([ones(shape(X)[0], 1), X])
y_t = matrix_transpose(y)
betas = []
for i in range(shape(y)[1]):
betas.append(self._fit(X, y_t[i]))
batas = matrix_transpose(betas)
self.betas = batas
def predict(self, X):
result = []
# dim_X = dim(X)
if dim(X) == 1:
X = [X]
for x in X:
loss = sum(square(minus(self.X, x)), axis=1)
index = argsort(loss)[:self.k]
if self.verbose:
print(index)
ys = []
for i in index:
ys.append(self.y[i])
k_loss_raw = sorted(loss)[:self.k]
k_loss = [1 / l if l != 0 else 0 for l in k_loss_raw]
k_loss_sum = sum(k_loss)
weights = [
l / float(k_loss_sum) if k_loss_sum != 0 else 1 for l in k_loss
]
weight_m = diag(weights)
ys = matrix_matmul(weight_m, ys)
result.append(sum(ys, axis=0))
if len(self.shape_Y) == 1:
result = matrix_transpose(result)[0]
return result
def standard_scaling(X, y=None, axis=1):
if axis == 0:
return matrix_transpose(standard_scaling(matrix_transpose(X), axis=1))
R = []
for j in range(shape(X)[1]):
col = fancy(X, None, j)
mean_ = mean(col)
std = sqrt(mean(square(minus(col, mean_))))
if y != None:
std_y = sqrt(mean(square(minus(y, mean(y)))))
if std == 0:
R.append(col)
else:
R.append([(x - mean_) * std_y / std for x in col])
return matrix_transpose(R)
def stdev(X):
# X = matrix_copy(X)
X_T = matrix_transpose(X)
m = mean(X, axis=1)
R = []
for j in range(shape(X)[1]):
R.append(sqrt(mean(square(minus(X_T[j], m[j])))))
return R
def stdev(X, axis=0):
assert (dim(X) == 2)
assert (axis == 0)
X_T = matrix_transpose(X)
m = mean(X, axis=0)
R = []
for j in range(shape(X)[1]):
R.append(sqrt(mean(square(minus(X_T[j], m[j])))))
return R
def normalize(X,
y=None,
norm='l2',
axis=1,
return_norm=False,
return_norm_inv=False):
assert (axis == 0 or axis == 1)
assert (norm == 'l2' or norm == 'l1')
X_T = matrix_transpose(X)
y_norm = None
if y != None:
if norm == 'l2':
y_norm = sqrt(sum(square(y)))
elif norm == 'l1':
y_norm = sqrt(sum(abs(y)))
if y and y_norm == 0:
return X
norms = []
if axis == 0:
A = matrix_copy(X)
for i in range(shape(X)[0]):
n = 0
if norm == 'l2':
n = sqrt(sum(square(
X_T[i]))) if not y else sqrt(sum(square(X_T[i]))) / y_norm
elif norm == 'l1':
n = sqrt(sum(abs(
X_T[i]))) if not y else sqrt(sum(square(X_T[i]))) / y_norm
if n != 0:
A[i] = (multiply(X[i], 1 / float(n)))
norms.append(n)
elif axis == 1:
A = matrix_transpose(X)
for j in range(shape(X)[1]):
n = 0
if norm == 'l2':
n = sum(square(
X_T[j])) if not y else sqrt(sum(square(X_T[j]))) / y_norm
elif norm == 'l1':
n = sum(abs(
X_T[j])) if not y else sqrt(sum(square(X_T[j]))) / y_norm
if n != 0:
A[j] = (multiply(X_T[j], 1 / float(n)))
norms.append(n)
A = matrix_transpose(A)
norms_inv = [0 if x == 0 else 1 / float(x) for x in norms]
if return_norm and return_norm_inv:
return A, norms, norms_inv
elif return_norm:
return A, norms
elif return_norm_inv:
return A, norms_inv
else:
return A
def _corr(A, i, j):
assert (dim(A) == 2)
m, n = shape(A)
A_T = matrix_transpose(A)
X, Y = A_T[i], A_T[j] # X,Y = col(A,i),col(A,j)
mean_X, mean_Y = mean(X), mean(Y)
X_ = [k - mean_X for k in X]
Y_ = [k - mean_Y for k in Y]
numerator = mean(multiply(X_, Y_))
# print(sqrt(mean(square(X_))))
denominator = sqrt(mean(square(X_))) * sqrt(mean(square(Y_)))
if denominator == 0:
return 0
else:
r = (numerator) / (denominator)
return r
def predict(self, X):
result = []
# dim_X = dim(X)
if dim(X) == 1:
X = [X]
for x in X:
loss = sum(square(minus(self.X, x)), axis=1)
# loss = sum(abs(minus(self.X,x)),axis=1)
from preprocessing import standard_scaling
new_X = standard_scaling(self.X, axis=0)
x = sqrt(square(minus(x, mean(x))))
loss = minus(loss, multiply(dot(new_X, x), self.alpha))
index = argsort(loss)[:self.k]
if self.verbose:
print(index, '/len', len(loss))
ys = []
for i in index:
ys.append(self.y[i])
result.append(mean(ys, axis=0))
return result
def predict(self, X):
result = []
# dim_X = dim(X)
if dim(X) == 1:
X = [X]
for x in X:
loss = sum(square(minus(self.X, x)), axis=1)
# loss = sum(abs(minus(self.X,x)),axis=1)
index = argsort(loss)[:self.k]
if self.verbose:
print(index, '/len', len(loss))
ys = []
for i in index:
ys.append(self.y[i])
result.append(mean(ys, axis=0))
return result
def fit(self,X,y):
assert(dim(X)==2)
assert(dim(y)==1 or dim(y)==2)
self.shape_X = shape(X)
self.shape_Y = shape(y)
if dim(y) == 1:
y = [[k] for k in y]
best_w = None
min_err = None
for i in range(self.max_iter):
W = self.random_w((shape(X)[1],shape(y)[1]))
y_ = matrix_matmul(X,W)
err = mean(sqrt(mean(square(minus(y,y_)),axis=1)))
if not best_w or min_err>err:
best_w = W
min_err = err
print(err)
self.W = best_w
def cross_val_score(estimator_instance,
X,
y,
is_shuffle=False,
cv='full',
scoring='score',
random_state=None,
return_mean=False,
verbose=False):
assert ((type(cv) == int and cv > 1) or cv == 'full')
assert (scoring == 'score' or scoring == 'loss')
if type(cv) == int:
assert (cv < len(X))
if is_shuffle:
X, y = shuffle(X, y=y, random_state=random_state)
N = len(X)
K = N if cv == 'full' else cv
h = len(X) / float(K)
scores = []
losses = []
for i in range(K):
s = int(round((i * h)))
e = int(round((i + 1) * h))
X_train, Y_train = [], []
X_train.extend(X[:s])
X_train.extend(X[e:])
Y_train.extend(y[:s])
Y_train.extend(y[e:])
X_val, Y_val = X[s:e], y[s:e]
estimator_instance.fit(X_train, Y_train)
p = estimator_instance.predict(X_val)
score = official_score(p, Y_val)
loss = l2_loss(p, Y_val)
# score = estimator_instance.score(X_val,Y_val)
scores.append(score)
losses.append(loss)
# print(scores)
if return_mean:
if scoring == 'score':
# print(scores)
std = sqrt(mean(square(minus(scores, mean(scores)))))
return (sorted(scores)[len(scores) / 2] + mean(scores) -
0.5 * std) / 2.0
# return (sorted(scores)[len(scores)/2] + mean(scores) - std)/2.0
# return sorted(scores)[len(scores)/2] - std
# return max(scores)
# return mean(scores[:len(scores)/2])
# return mean(sorted(scores)[::-1][:len(scores)/2])
# return (mean(scores) + max(scores))/2.0
# return mean(scores)
# return mean(scores) -0.5*std
elif scoring == 'loss':
# return mean(losses)
std = sqrt(mean(square(minus(losses, mean(losses)))))
# return mean(losses)
return ((sorted(losses)[len(losses) / 2] + mean(losses) + std) /
2.0)
else:
if scoring == 'score':
return scores
elif scoring == 'loss':
return losses
def _whatch(self, X, y):
p = self.predict(X)
loss = sum(square(minus(p, y)))
print(loss)
def _score_calc(y, y_):
y_ = [int(round(i)) for i in y_]
numerator = sqrt(mean(square(minus(y, y_))))
return numerator