Scribd
Upload
46 views29 pages

Linear Law Linear Programming: By: Ting Hie Huong Lee Mee Ching

Linear relationships between variables can be represented using straight lines. This allows data to be more easily modeled and future values to be predicted. Non-linear relationships can be linearized by transforming the data, such as plotting y against x2 to reduce a quadratic relationship to a straight line. Common transformations include plotting the log of y against x to analyze power law relationships, or plotting the log of y against log x to study exponential forms. These transformations allow parameters like gradients and intercepts to determine the underlying relationship between variables.

Uploaded by

Anna Lee
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
46 views29 pages

Linear Law Linear Programming: By: Ting Hie Huong Lee Mee Ching

Linear relationships between variables can be represented using straight lines. This allows data to be more easily modeled and future values to be predicted. Non-linear relationships can be linearized by transforming the data, such as plotting y against x2 to reduce a quadratic relationship to a straight line. Common transformations include plotting the log of y against x to analyze power law relationships, or plotting the log of y against log x to study exponential forms. These transformations allow parameters like gradients and intercepts to determine the underlying relationship between variables.

Uploaded by

Anna Lee
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 29

Linear Law

Linear Programming
By: Ting Hie Huong
Lee Mee Ching
Linear Law
Form 5 Chapter 2
Today
Reducing data relationships to straight lines

WHY ?

One good reason is to be able to make forecasts from data

It is much easier to work and predict future results when the data
lies on a straight line

Another reason is that reducing to straight line form may help us


understand the relationship between x and y variables better

Consider the following data...............................


Non linear growth
140 Reduced to straight line form
G 120 150
r 100 v
N
o 80 a 100
e
w 60 l
w 50
t 40 u
h 20 e
Y 0
0 s
0 50 100 150
0 5 10 15 New X values
time

Forecast using straight line


250
N Convert back to Non-Linear for
e 200
w
actual predicted value
150 250
Y G Predicted value
200
100 r
Forecast
v o 150

a w 100
50
l t
50
u h
0
0
e 0 50 100 150 200 250
0 5 10 15
s
New X values
time
Results that give a straight line graph
Here’s some data that is linear

x 3.1 5.6 7.0 10.4 12.5

y 2.3 9.7 14.2 24.2 30.4


If these values are plotted on a graph they
lie on a straight line and so obey the law

y = mx + c
Experimental data 1

35

30

25

20

15

y 10

0
-2 0 2 4 6 8 10 12 14
-5

-10

-15
x
 We can calculate a and b from the graph
as follows:
The gradient, a
Choosing two points a long way apart, the
line passes through (2.3, 0) and (12.0, 28.5),
Therefore gradient a = 28.5/9.7 = 2.94 = 2.9
(to 2 s.f)
The y intercept, b
The graph cuts the y axis at (0, -7) ,
therefore y intercept b = -7

Equation is theny = 2.9x – 7.0


Equations of the form y = ax²+ b
If in another experiment the data appears to satisfy a
quadratic relationship; If we let

Y =y, X = x²

Then we can get a straight line by plotting


Y = aX + b

We can plot Y (=y) against X (= x²) we should get a


straight line and we can find a and b from our
graph.
Example
A hosepipe squirts water and the height, y
metres of the water above a fixed level at
a distance x m from the hose is measured
as
x 2 4 5 6 7 8
y 6.1 3.6 2.2 -0.1 -2.9 -5.5

This is thought to obey the law

y = ax2 + b
We need to make a table with values of x²
and plot these on a graph, (a BIG graph if by
plotted by hand!)

x²=X 4 16 25 36 49 64

y=Y 6.1 3.6 2.2 -0.1 -2.9 -5.5


Hosepipe: height (Y) against distance squared (X)

2
Y (height)

0
0 10 20 30 40 50 60 70
-2

-4

-6

-8
X (distance squared)
Plotting Y against X, gives a straight line Y=
aX + b

From the graph, choosing 2 points e.g.


(0, 6.9) and (35, 0) gives a = -0.20
(to 2 s.f.) (The gradient is -0.2)

The line cuts the Y axis where Y = 6.9 and so b = 6.9

Therefore y = -0.2X + 6.9 or

y = -0.2x2 + 6.9
Hosepipe: height (Y) against distance squared (X)

2
Y (height)

0
0 10 20 30 40 50 60 70
-2

-4

-6

-8
X (distance squared)
Equations of the form y = kxn

(Power Law growth / decay y=kx-n )


y = kx n
 Plot log y against log x

 Intercept is log k

 Gradient is n
A water pipe is being laid between two
points. The following data are being used to
show how, for a given pressure difference,
the rate of flow R litres per second, varies
with the pipe diameter d cm

d 1 2 3 5 10

R 0.02 0.32 1.62 12.53 199.8


Here we try the relationship R = kdm where k
is a constant

log R = m log d + log k

Compare with y = mx + c
log d 0 0.3 0.48 0.70 1.00 (x)

log R -1.70 -0.49 0.21 1.10 2.30 (y)


Log flow rate against log pipe diameter

2.5
y = 4.0x - 1.7
2

1.5

0.5
Log R

0
0 0.2 0.4 0.6 0.8 1 1.2
-0.5

-1

-1.5

-2
Log d
Reading from the graph we can see that
c = = - 1.70
so k = = 0.02

gradient m = (change in y)/(change in x)


= 4.0

The relationship is then

R = 0.02d4
Equations of the form y = kax

(Exponential relationships)
y = kax
Plot logy against x

intercept is logk

gradient is loga
The temperature θ in ºC of a cup of coffee
after t minutes is recorded below

t 2 4 6 8 10 12

θ 81 70 61 52 45 38
 If the relationship is of the form  = kat
where k and a are constants

 log  = (loga)t + logk

 (y = mx + c)

t 2 4 6 8 10 12

log θ 1.91 1.85 1.79 1.72 1.65 1.58


Log theta against time

2.5

1.5
Log theta

y = -0.03x + 1.98

0.5

0
0 2 4 6 8 10 12 14
t (minutes)
Reading from the graph gives

Log k = 1.98 k = 101.98 so k = = 95.5

gradient = y / x = -0.03

Log a = - 0.03 so a = 10-0.03 = 0.93

Relationship is

= 95.5 x 09.3t


Summary
y x 2
Plot Y vs X where X=x2

y  kx n
Plot Log(y) vs Log(x)

y  kb x
Plot Log(y) vs x

You might also like