def computeH(pts1, pts2):
l1, l2 = len(pts1), len(pts2)
assert l1 == l2
A = []
B = []
for j in range(l1 * 2):
i = int(j / 2)
if (j % 2 == 0):
A.append([
pts1[i][0], pts1[i][1], 1, 0, 0, 0, -pts2[i][0] * pts1[i][0],
-pts2[i][0] * pts1[i][1]
])
else:
A.append([
0, 0, 0, pts1[i][0], pts1[i][1], 1, -pts2[i][1] * pts1[i][0],
-pts2[i][1] * pts1[i][1]
])
for i in range(l2):
B.append(pts2[i][0])
B.append(pts2[i][1])
# H matrix 3*3
a, b, c, d, e, f, g, h = lstsq(A, B)[0]
return [[a, b, c], [d, e, f], [g, h, 1]]
def draw_edge(img):
univ = depth_to_edge_univ(img)
xs, ys = img_to_dots(univ)
A = np.array([xs, np.ones(xs.size)])
w = linalg.lstsq(A.T, ys)[0]
# slope, intercept, r_value, p_value, std_err = scipy.stats.linregress(xs, ys)
for y in range(0, univ.shape[1]):
x = int(y * w[0]) + w[1]
if x < univ.shape[1] and y < univ.shape[0]:
univ[x, y] = 254
return univ
def _linregression(self, x, y):
m, c = linalg.lstsq(numpy.vstack([x, numpy.ones(len(x))]).T, y)[0]
return m, c
sline = line.split(',')
aaa = []
year = np.array([[1,1980],[1,1984],[1,1988],[1,1992],[1,1996],[1,2000],[1,2004],[1,2008]])
x = np.array(float(sline[1]))
for i in range(7):
# a.append(int(sline[i+1]))
a = np.array(float(sline[2+2*i]))
x = np.vstack((x,a))
print x
'''
if cnt == 30:
plt.plot(x)
plt.show()
'''
#Do the regression here:
reg = lsq.lstsq(year,x)
rr = reg[0]
a = rr[0]
b = rr[1]
a12 = float(a + 2012*b)
print a12
for i in range(8):
aaa.append(float(x[i]))
aaa.append(a12)
writer.writerow(aaa)
cnt += 1
if cnt >50:
break
f1.close()
def _linregression(self, x, y):
m, c = linalg.lstsq(numpy.vstack([x, numpy.ones(len(x))]).T, y)[0]
return m, c
def find_coefficients(data, exponents):
X = tuple((tuple((pow(x,p) for p in exponents)) for (x,y) in data))
y = tuple(((y) for (x,y) in data))
x, resids, rank, s = lstsq(X,y)
return x
