S=at+b has a function to calculate the result directly. S is a function of y and t is a function of X.

1)

A) y = [ABC] * [x 2x 1] t matrix form

Let t0 = x2t1= x.

S=y- 1 has s =at0+bt 1, which is the least square about t0 and t 1.

b) lny = ln(a)+n ln(x)

T=ln(x) s=ln(y) b=ln(a) has the linear form of s=b+nt.

Note that if a is less than 0, you need to do a process, take the inverse of all y, and then do it according to the above form.

2)

A) ln(y) = ln(b)+nln(x) y, b and x are all positive numbers.

s=ln(y) t=ln(x) c=ln(b)

s=c+nt

b) ln(y) = ln(b)+cx

s=ln(y) a=ln(b) t=x

s=a+ct

c) y = a ln(x)+b

s=y t=ln(x)

s=at+b

d) ln(y)=ln(a)+2x

s=ln(y) b=ln(a) t=x

s=b+2t

e) ln(y)=ln(a)+3x

s=ln(y) b=ln(a) t=x

s=b+3t

After calculating the correlation coefficient between s and t, the original coefficient in the form of y x can be deduced.