E-36
maxX
1
2
(S-VAR)
2
(MINMAX)
2
Obtains the maximum value of the sample
x
-data.
minY
1
2
(S-VAR)
2
(MINMAX)
e
1
Obtains the minimum value of the sample
y
-data.
maxY
1
2
(S-VAR)
2
(MINMAX)
e
2
Obtains the maximum value of the sample
y
-data.
A
Regression Coefficient and Estimated Value Calculation
Formula Table
Linear Regression
Command Calculation Formula
Regression Formula
Constant Term a
a =
n
Σy
i
– b
.
Σx
i
Regression Coefficient b
b =
n
.
Σx
i
2
–
(
Σx
i
)
2
n
.
Σx
i
y
i
– Σx
i
.
Σy
i
Correlation Coefficient
r
r =
{n
.
Σx
i
2
–
(
Σx
i
)
2
}{n
.
Σy
i
2
–
(
Σy
i
)
2
}
n
.
Σx
i
y
i
– Σx
i
.
Σy
i
Estimated Value
m
m
y – a
b
=
Estimated Value
n = a + bx
Quadratic Regression
Command Calculation Formula
Regression Formula
Constant Term a
a = – b
(
)
– c
(
)
n
Σy
i
n
Σx
i
n
Σx
i
2
Regression Coefficient b
b =
Sxx
.
Sx
2
x
2
– (Sxx
2
)
2
Sxy
.
Sx
2
x
2
– Sx
2
y
.
Sxx
2
Regression Coefficient c
c =
Sxx
.
Sx
2
x
2
– (Sxx
2
)
2
Sx
2
y
.
Sxx
– S
xy
.
Sxx
2
However,
(Σx
i
)
2
Sxx = Σx
i
2
–
n
Sxy = Σx
i
y
i
–
n
(Σx
i
.
Σy
i
)
Sxx
2
= Σx
i
3
–
n
(Σx
i
.
Σx
i
2
)
Sx
2
x
2
= Σx
i
4
–
n
(Σx
i
2
)
2
Sx
2
y = Σx
i
2
y
i
–
n
(Σx
i
2
.
Σy
i
)