def fit():
x, y = sample(5)
poly = polyreg.PolynomialRegression(4)
poly.fit(x, y)
lin = linearreg.LinearRegression()
lin.fit(x, y)
return poly, lin
def fit(self, x, y):
x_pow = []
xx = x.reshape(len(x), 1)
for i in range(1, self.degree + 1):
x_pow.append(xx ** i)
mat = np.concatenate(x_pow, axis=1)
linreg = linearreg.LinearRegression()
linreg.fit(mat, y)
self.w_ = linreg.w_
def train_lr(datasets):
model = {}
sse = {}
mse = {}
y_pred_dict = {}
for i in range(1, 7):
model[i] = {}
sse[i] = {}
mse[i] = {}
y_pred_dict[i] = {}
for j, dataset in enumerate(datasets):
if i == 1:
y_train = zip(*dataset)[1]
y_predict = [1] * len(y_train)
sse_train = calculate_sse(y_train, y_predict)
model[i][j] = None
sse[i][j] = sse_train
mse[i][j] = sse_train / len(y_predict)
y_pred_dict[i][j] = y_predict
else:
x_train = map(lambda x: gx(i, x), zip(*dataset)[0])
y_train = zip(*dataset)[1]
x_train_std = standardize_data(x_train)
linearreg = lr.LinearRegression(x_train_std, y_train)
w = linearreg.getparam()
y_predict_train = linearreg.predict(x_train_std)
y_predict = y_predict_train.getT().tolist()[0]
sse_train = calculate_sse(y_train, y_predict)
model[i][j] = copy.deepcopy(linearreg)
sse[i][j] = sse_train
mse[i][j] = sse_train / len(y_predict)
y_pred_dict[i][j] = y_predict
return {
'model': model,
'sse': sse,
'mse': mse,
'y_predict_dict': y_pred_dict,
}
def fit(self, x, y):
x_pow = []
# reshapeは第二引数の列数に配列を変換するreshape(5, 2)ならば10列のarrayは5行2列になる
xx = x.reshape(len(x), 1)
# xxは以下のような形
# [[5.48813504]
# [7.15189366]
# [6.02763376]
# [5.44883183]
# [4.23654799]
# [6.45894113]
# [4.37587211]
# [8.91773001]
# [9.63662761]
# [3.83441519]]
for i in range(1, self.degree + 1):
# ここで多項式回帰の元になるn乗の計算を実現している
x_pow.append(xx**i)
# ベクトルを横につなぐ
mat = np.concatenate(x_pow, axis=1)
linreg = linearreg.LinearRegression()
linreg.fit(mat, y)
self.w_ = linreg.w_
xmin = 0
xmax = 12
ymin = -1
ymax = 25
fig, axes = plt.subplots(nrows=2, ncols=5)
for i in range(5):
axes[0, i].set_xlim([xmin, xmax])
axes[0, i].set_ylim([ymin, ymax])
axes[1, i].set_xlim([xmin, xmax])
axes[1, i].set_ylim([ymin, ymax])
xx = x[:2 + i * 2]
yy = y[:2 + i * 2]
axes[0, i].scatter(xx, yy, color="k")
axes[1, i].scatter(xx, yy, color="k")
model = linearreg.LinearRegression()
model.fit(xx, yy)
xs = [xmin, xmax]
ys = [model.w_[0] + model.w_[1] * xmin,
model.w_[0] + model.w_[1] * xmax]
axes[0, i].plot(xs, ys, color="k")
model = ridge.RidgeRegression(10.)
model.fit(xx, yy)
xs = [xmin, xmax]
ys = [model.w_[0] + model.w_[1] * xmin,
model.w_[0] + model.w_[1] * xmax]
axes[1, i].plot(xs, ys, color="k")
plt.show()
xx = np.arange(0, 5, 0.01)
np.random.seed(0)
y_poly_sum = np.zeros(len(xx))
y_poly_sum_sq = np.zeros(len(xx))
y_lin_sum = np.zeros(len(xx))
y_lin_sum_sq = np.zeros(len(xx))
y_true = f(xx)
n = 100000
warnings.filterwarnings("ignore")
for _ in range(n):
x, y = sample(5)
poly = polyreg.PolynomialRegression(4)
poly.fit(x, y)
lin = linearreg.LinearRegression()
lin.fit(x, y)
y_poly = poly.predict(xx)
y_poly_sum += y_poly
y_poly_sum_sq += (y_poly - y_true)**2
y_lin = lin.predict(xx)
y_lin_sum += y_lin
y_lin_sum_sq += (y_lin - y_true)**2
fig = plt.figure()
ax1 = fig.add_subplot(121)
ax2 = fig.add_subplot(122)
ax1.set_title("Linear reg.")
ax2.set_title("Polynomial reg.")
ax1.set_ylim(0, 1)
ax2.set_ylim(0, 1)