
145
Obtaining coefficients from the fit model
Coefficients can be obtained from the chosen fit model algebraically. The
function
PREDY
from
MATH
gives a predicted y value
using the last line of
best fit that was calculated
. This means that you must use the
SYMB
view to
ensure that your set of data is the only one
ed and make sure
SYMB
SETUP
is set to the correct fit model, and also use the
PLOT
screen and the
key to ensure that your set of data was the last one graphed and that it
has had its curve of best fit displayed. Until the curve has been displayed, the
coefficients are not available or, worse, might belong to another data set.
If the line of best fit is
*
y
m X
b
=
+
(see the
SYMB
screen) then the
calculations shown below and right will give the slope and y-intercept.
(0)
*0
and
(1)
(0)
( *1
) ( *0
)
PREDY
m
b
b
PREDY
PREDY
m
b
m
b
m b b
m
=
+
=
−
=
+ −
+
= + −
=
Finding Fit Coefficients
Linear
-
m*X+b
b
=
PREDY(0)
m
=
PREDY(1)-PREDY(0)
Logarithmic - m*LN(X)+b
b
=
PREDY(1)
m
=
PREDY(e)- PREDY(1)
Exponential - b*EXP(m*X)
b
=
PREDY(0)
m
=
LN(PREDY(1)/PREDY(0))
Power
-
b*X^m
b
=
PREDY(1)
m
=
LN(PREDY(e)/PREDY(1))
Quadratic
- a*X^2+b*X+c
a
=
(PREDY(2)-2*PREDY(1)
+
PREDY(0))/2
b
=
(PREDY(2)+4*PREDY(1)-3*PREDY(0))/2
c
=
PREDY(0)
or,
1
0
0 1
(0)
1
1 1
(1)
4
2 1
(2)
a
PREDY
b
PREDY
c
PREDY
−
=
×
Cubic -
a*X^3+b*X^2+cX+d
1
0
0
0 1
(0)
1
1
1 1
(1)
8
4
2 1
(2)
27
9
3 1
(3)
a
PREDY
b
PREDY
c
PREDY
d
PREDY
−
=
×