What is Curve Fitting?

Curve Fitting is the process of constructing a

curve or mathematical function that best fits a
series of data points, possibly subject to
constraints.

1
Dr. Sonal Nirwal
 Why Curve Fitting?
• The main purpose of curve fitting is to theoretically
describe experimental data with a model (function or
equation) and to find the parameters associated with
the model.
• Fitted curves can be used as an aid for data
visualization, to infer of a function where no data are
available, and to summarize the relationships among
two or more variables.

2
Dr. Sonal Nirwal
 Fitting a Straight Line
The equation of straight line is

𝑦 = 𝑎 + 𝑏𝑥 (1)
Multiply by 1 on both side in Eq. (1) and take summation
𝑛 𝑛

෍ 𝑦𝑖 = 𝑛𝑎 + 𝑏 ෍ 𝑥𝑖
𝑖=1෍ 𝑦 = 𝑛𝑎 + 𝑏 ෍𝑖=1
𝑥

Or (2)
Multiply by x on both side in Eq. (1) and take summation
𝑛 𝑛 𝑛

෍ 𝑥𝑖 𝑦𝑖 = 𝑎 ෍ 𝑥𝑖 + 𝑏 ෍ 𝑥𝑖2
𝑖=1 𝑖=1 𝑖=1
෍ 𝑥𝑦 = 𝑎 ෍ 𝑥 + 𝑏 ෍ 𝑥 2
Or (3)

3
Equations (2) and (3) are called Normal Equations.
Dr. Sonal Nirwal
 Example 1. By the method of least square, find the straight line that best fits the following data:
x 1 2 3 4 5
y 14 27 40 55 68

Solution. Let the straight line of best fit be

𝑦 = 𝑎 + 𝑏𝑥 (1)
Normal Equations are
෍ 𝑦 = 𝑛𝑎 + 𝑏 ෍ 𝑥 (2)
෍ 𝑥𝑦 = 𝑎 ෍ 𝑥 + 𝑏 ෍ 𝑥 2
(3)
Note: here n=5.
Now x y xy 𝒙𝟐
1 14 14 1
2 27 54 4
3 40 120 9
4 55 220 16
5 68 340 25
σ 𝑥=15 σ 𝑦=204 σ 𝑥𝑦=748 σ 𝑥 2 =55
4
Dr. Sonal Nirwal
By substituting the values from table in Eqs. (2) and (3), we get
204 = 5𝑎 + 15𝑏
748 = 15𝑎 + 55𝑏

By solving these, we get 𝑎 = 0, 𝑏 = 13.6

Hence required straight line is 𝑦 = 13.6 x

Answer

Dr. Sonal Nirwal 5

Example 2. By the method of least square, find the straight line that best fits the following data:
x 0 1 2 3 4
y 1 1.8 3.3 4.5 6.3

𝑦 = 0.72 + 1.33𝑥
Answer

Example 3. By the method of least square, find the straight line that best fits the following data:

x 1 2 3 4 6 8
y 2.4 3.1 3.5 4.2 5.0 6.0

𝑦 = 2.0253 + 0.502𝑥
Answer

Dr. Sonal Nirwal 6

 Fitting a 2 nd degree parabola
Let us consider we have n data points
The equation of parabola is
𝑦 = 𝑎 + 𝑏𝑥 +
𝑐 𝑥2 (1)
Multiply by 1 on both side in Eq. (1) and take summation
σ 𝑦 = 𝑛𝑎 + 𝑏 σ 𝑥+cσ 𝑥 2 (2)

Multiply by x on both side in Eq. (1) and take summation

෍ 𝑥𝑦 = 𝑎 ෍ 𝑥 + b ෍ 𝑥 2 + c ෍ 𝑥 3

(3)

Multiply by x on both side in Eq. (1) and take summation

෍ 𝑥2𝑦 = 𝑎 ෍ 𝑥2 + b ෍ 𝑥3 + c ෍ 𝑥4
(4)
7
Equations (2), (3) and (4) are called Normal Equations.
Dr. Sonal Nirwal
 Example 4. By the method of least square, find the parabola that best fits the following data:
x 1 2 3 4 5 6 7
y -5 -2 5 16 31 50 73

Solution. Let the parabola of best fit be

𝑦 = 𝑎 + 𝑏𝑥 + 𝑐 𝑥 2 (1)
Normal Equations are
σ 𝑦 = 𝑛𝑎 + 𝑏 σ 𝑥+cσ 𝑥 2 (2)
σ 𝑥𝑦 = 𝑎 σ 𝑥 + b σ 𝑥 2 + c σ 𝑥 3 (3)

σ 𝑥2𝑦 = 𝑎 σ 𝑥2 + b σ 𝑥3 + c σ 𝑥4 (4)
Note: here n=7.
Now

8
r. Sonal Nirwal
 By substituting the values from table in Eqs. (2), (3) and (4), we get
168 = 7𝑎 + 28𝑏 + 140𝑐
1036 = 28𝑎 + 140𝑏 + 784𝑐
6440 = 140𝑎 + 784𝑏 + 4676𝑐

By solving these, we get 𝑎 = −4, 𝑏 = −3, 𝑐 = 2

Hence, the required parabola is 𝑦 = −4 − 3𝑥 + 2𝑥 2

Answer

9
Dr. Sonal Nirwal
 Fitting an Exponential Curve
The equation of straight line is

𝑦 = 𝑎𝑒 𝑏𝑥 (1)
Taking log on both sides
𝑙𝑜𝑔10 𝑦 = 𝑙𝑜𝑔10 𝑎 + 𝑏𝑥𝑙𝑜𝑔10 𝑒
or
𝑙𝑜𝑔10 𝑦 = 𝑙𝑜𝑔10 𝑎 + 𝑏𝑙𝑜𝑔10 𝑒 𝑥

Y =𝐴 +𝐵 𝑋 (2)
෍ 𝑌 = 𝑛𝐴line
Normal equations related to straight +𝐵෍ 𝑋 Eq. 2) are
(i.e.,
෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2

where Y= 𝑙𝑜𝑔10 𝑦, X=x, A= 𝑙𝑜𝑔10 𝑎, B= 𝑏 𝑙𝑜𝑔10 𝑒

10
Dr. Sonal Nirwal
Example 5. By the method of least square, find the curve 𝑦 = 𝑎𝑒 𝑏𝑥 that best fits the following data:
x 1 5 7 9 12
y 10 15 12 15 21

Solution. The curve to be fitted is

𝑦 = 𝑎𝑒 𝑏𝑥 (1)
Or
Y= 𝐴 + 𝐵𝑋 (2)

Where Y= 𝑙𝑜𝑔10 𝑦, X=x, A= 𝑙𝑜𝑔10 𝑎, B= 𝑏 𝑙𝑜𝑔10 𝑒

Normal Equations are

෍ 𝑌 = 𝑛𝐴 + 𝐵 ෍ 𝑋 (3)

෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2 (4)

Note: here n=5.

Dr. Sonal Nirwal 11
 X=x y Y=𝑙𝑜𝑔10 𝑦 𝑿𝟐 XY
Y= 𝑙𝑜𝑔10 𝑦,
1 10 𝑙𝑜𝑔10 10 =1.0000 1 1 X=x,
5 15 𝑙𝑜𝑔10 15=1.1761 25 5.8805 A= 𝑙𝑜𝑔10 𝑎,
7 12 𝑙𝑜𝑔10 12=1.0792 49 7.5544 B= 𝑏 𝑙𝑜𝑔10 𝑒

9 15 𝑙𝑜𝑔10 15=1.1761 81 10.5849

෍ 𝑌 = 𝑛𝐴 + 𝐵 ෍ 𝑋
12 21 𝑙𝑜𝑔10 21=1.3222 144 15.8664
σ 𝑄=34 σ 𝑦=204 σ 𝑌=5.7536 σ 𝑋 2 =300 σ 𝑋𝑌=40.8862 ෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2

By substituting the values from table in Eqs. (3) and (4), we get
5.7536 = 5𝐴 + 34𝐵
40.8862 = 34𝐴 + 300𝐵

By solving these, we get A= 0.9766; 𝐵 = 0.025611

Now, 𝑎 = 𝑎𝑛𝑡𝑖𝑙𝑜𝑔10 𝐴(= 𝑎𝑛𝑡𝑖𝑙𝑜𝑔10 0.9766) = 9.4754

𝐵 0.025611
𝑏= = = 0.059
𝑙𝑜𝑔10 𝑒 𝑙𝑜𝑔10 𝑒

Hence the required curve is

𝑦 = 9.4754𝑒 0.059𝑥Dr. Sonal Nirwal 12
Answer
 Example 6. Determine the constants a and b by the method of least square such that 𝑦 = 𝑎𝑒 𝑏𝑥 fits the
following data:
x 2 4 6 8 10
y 4.077 11.084 30.128 81.897 222.62

𝑦 = 1.49989 𝑒 0.50001𝑥
Answer

Example 7. Find the curve of the best fits 𝑦 = 𝑎𝑒 𝑏𝑥 to the following data by the method of least square.

x 1 2 3 4 5 6
y 1.6 4.5 13.8 40.2 125 300

𝑦 = 0.5580 𝑒 1.0631𝑥
Answer

Dr. Sonal Nirwal 13

• Fitting of Curve 𝑦 = 𝑎 𝑥 𝑏
𝑙𝑜𝑔10 𝑦 = 𝑙𝑜𝑔10 𝑎 + 𝑏 𝑙𝑜𝑔10 𝑥
Y=𝐴+𝐵𝑋
Normal equation are
෍ 𝑌 = 𝑛𝐴 + 𝐵 ෍ 𝑋

෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2

Where Y= 𝑙𝑜𝑔10 𝑦, X= 𝑙𝑜𝑔10 𝑥, A= 𝑙𝑜𝑔10 𝑎, B= 𝑏

• Fitting of Curve 𝑦 = 𝑎 𝑏 𝑥
𝑙𝑜𝑔10 𝑦 = 𝑙𝑜𝑔10 𝑎 + 𝑙𝑜𝑔10 𝑏 𝑥
Y= 𝐴 + 𝐵 𝑋
Normal equation are
෍ 𝑌 = 𝑛𝐴 + 𝐵 ෍ 𝑋

෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2

Where Y=𝑙𝑜𝑔10 𝑦, X= 𝑥, A=𝑙𝑜𝑔10 𝑎, B=𝑙𝑜𝑔10 𝑏

Dr. Sonal Nirwal 14
Example 8. By the method of least square, find the curve 𝑦 = 𝑎𝑏 𝑥 that best fits the following data:
x 2 3 4 5 6
y 8.3 15.4 33.1 65.2 127.4

Solution. The curve to be fitted is

𝑦 = 𝑎𝑏 𝑥 (1)
Or
Y= 𝐴 + 𝐵𝑋 (2)

Where Y=𝑙𝑜𝑔10 𝑦, X= 𝑥, A=𝑙𝑜𝑔10 𝑎, B=𝑙𝑜𝑔10 𝑏

Normal Equations are ෍ 𝑌 = 𝑛𝐴 + 𝐵 ෍ 𝑋

(3)
෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2
(4)

Dr. Sonal Nirwal 15

Note: here n=5.
 X=x y Y=𝑙𝑜𝑔10 𝑦 𝑿𝟐 XY
Y=𝑙𝑜𝑔10 𝑦,
2 8.3 𝑙𝑜𝑔10 8.3 =0.9191 4 1.8283 X= 𝑥,
3 15.4 𝑙𝑜𝑔10 15.4=1.1872 9 3.5616 A=𝑙𝑜𝑔10 𝑎,
4 33.1 𝑙𝑜𝑔10 33.1=1.5198 16 6.0792 B=𝑙𝑜𝑔10 𝑏

5 65.2 𝑙𝑜𝑔10 65.2=1.8142 25 9.0710

෍ 𝑌 = 𝑛𝐴 + 𝐵 ෍ 𝑋
6 127.4 𝑙𝑜𝑔10 127.4=2.1052 36 12.6312
σ 𝑋=20 σ 𝑌=7.5455 σ 𝑋 2 =90 σ 𝑋𝑌=33.1812 ෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2

By substituting the values from table in Eqs. (3) and (4), we get
7.5455 = 5𝐴 + 20𝐵
33.1812 = 20𝐴 + 90𝐵

By solving these, we get A= 0.31; 𝐵 = 0.3

Now, 𝑎 = 𝑎𝑛𝑡𝑖𝑙𝑜𝑔10 𝐴(= 𝑎𝑛𝑡𝑖𝑙𝑜𝑔10 0.31) = 2.04

𝑏 = 𝑎𝑛𝑡𝑖𝑙𝑜𝑔10 𝐵(= 𝑎𝑛𝑡𝑖𝑙𝑜𝑔10 0.3) = 1.995

Hence the required curve is

r. Sonal Nirwal 𝑦 = 2.04(1.995) 𝑥 16
Answer
 Example 9. Determine the constants a and b by the method of least square such that 𝑦 = 𝑎𝑥 𝑏 fits the
following data:
x 1 2 3 4 5
y 0.5 2 4.5 8 12.5

𝑦 = 0.5012 𝑥 1.9977
Answer

Dr. Sonal Nirwal 17

• Fitting of Curve 𝑃𝑉 𝛾 = 𝑘
𝑘
𝑉𝛾 =
𝑃
𝑘
𝛾 𝑙𝑜𝑔𝑉 = 𝑙𝑜𝑔
𝑃
𝑘
𝛾 𝑙𝑜𝑔𝑉 = 𝑙𝑜𝑔
𝑃
𝛾 𝑙𝑜𝑔𝑉 = 𝑙𝑜𝑔𝑘 − 𝑙𝑜𝑔𝑃
Or
1 1
𝑙𝑜𝑔𝑉 = 𝑙𝑜𝑔𝑘 − 𝑙𝑜𝑔𝑃
𝛾 𝛾

෍ 𝑌𝑌==𝑛𝐴
𝐴++ 𝐵𝑋
𝐵 ෍𝑋
Normal equation are

෍ 𝑋𝑌 = 𝐴 ෍ 𝑋 + 𝐵 ෍ 𝑋 2

1 1
Where Y=𝑙𝑜𝑔𝑉, X=𝑙𝑜𝑔𝑃, A= 𝑙𝑜𝑔𝑘, B=−
𝛾 𝛾

Dr. Sonal Nirwal 18

 Moment about Mean (Central Moments)
For individual series:
If 𝑥1 , 𝑥2 … 𝑥𝑛 are the values of the variable under consideration. Then the rth moment
(denoted by 𝜇𝑟 ) about the mean (𝑥) is given by

𝑛 𝑟 = 0,1,2, … . 𝑛
1
𝜇 𝑟 = ෍(𝑥𝑖 − 𝑥)𝑟
𝑛
𝑖=1 σ𝑛
𝑖=1 𝑥 𝑖
where mean 𝑥 = .
𝑛
For frequency distribution:
If 𝑥1 , 𝑥2 … 𝑥𝑛 are the values of the variable x with the corresponding frequency
𝑓1 , 𝑓2 … 𝑓𝑛 respectively. Then the rth moment 𝜇𝑟 about the mean 𝑥 is given by
𝑛
1 𝑟 = 0,1,2, … . 𝑛
𝜇 𝑟 = ෍ 𝑓𝑖 (𝑥𝑖 − 𝑥)𝑟
𝑁
𝑖=1
where 𝑁 = σ𝑛𝑖=1 𝑓𝑖 .
σ𝑛
𝑖=1 𝑓 𝑖 𝑥 𝑖
mean 𝑥 = 𝑁
Dr. Sonal Nirwal 19
 Moment about Mean (Central Moments)

• For r=0
𝑛 𝑛
1 1
𝜇0 = ෍ 𝑓𝑖 (𝑥𝑖 − 𝑥) 0 = ෍ 𝑓𝑖
𝑁 𝑁
𝑖=1 𝑖=1
1
= ∗𝑁 =1
𝑁
𝜇 0 = 1.
• For r=1
𝜇1 = 0
• For r=2
𝑛
1
𝜇 2 = ෍ 𝑓𝑖 (𝑥𝑖 − 𝑥)2 = Variance = (𝜎𝑥 )2
𝑁
𝑖=1
𝜎𝑥 is called standard deviation (SD).
𝑛
1
𝜎𝑥 = ෍ 𝑓𝑖 (𝑥𝑖 − 𝑥) 2
𝑁
𝑖=1

Dr. Sonal Nirwal 20

Example 1. Find the first four moments for the following individual series:
x 3 6 8 10 18

Solution. Calculation of Moments 𝑛

1
𝑥 = ෍ 𝑥𝑖
𝑛 𝑛
1 𝑖=1
𝜇 𝑟 = ෍(𝑥𝑖 − 𝑥)𝑟 45
𝑛 𝑥= =9
𝑖=1 5

5
S.No. x 𝑥𝑖 − 𝑥 (𝑥𝑖 − 𝑥) 𝟐 (𝑥𝑖 − 𝑥) 𝟑 (𝑥𝑖 − 𝑥) 𝟒 1 0
1. 3 -6 36 -216 1296 𝜇1 = ෍ 𝑥𝑖 − 𝑥 = = 0
5 5
𝑖=1
2. 6 -3 9 -27 81 5
1 2
128
3. 8 -1 1 -1 1 𝜇 2 = ෍(𝑥𝑖 − 𝑥) = = 25.6
5 5
4. 10 1 1 1 1 𝑖=1
5
5. 18 9 81 729 6561
1 3
486
𝜇 3 = ෍(𝑥𝑖 − 𝑥) = = 97.2
5 5
𝑖=1
෍𝑥 ෍ 𝑥𝑖 − 𝑥 ෍(𝑥𝑖 − 𝑥) 𝟐 ෍(𝑥𝑖 − 𝑥)𝟑 ෍(𝑥𝑖 − 𝑥)𝟒 5
1 4
7940
= 45 =0 = 128 = 486 = 7940 𝜇 4 = ෍(𝑥𝑖 − 𝑥) = = 1588
5 5
𝑖=1
Dr. Sonal Nirwal 21
Example 2. Calculate 𝝁𝟏 , 𝝁𝟐 , 𝝁𝟑 , 𝝁𝟒 for the following frequency distribution:
Marks 5-15 15-25 25-35 35-45 45-55 55-65
No. of 10 20 25 20 15 10
Solution. students

Calculation of Moments 𝑛
1
𝜇 𝑟 = ෍ 𝑓𝑖 (𝑥𝑖 − 𝑥)𝑟 1
𝑁 Mean 𝑥 = σ𝑛𝑖=1 𝑓𝑖 𝑥𝑖
𝑖=1 𝑁
6 𝑁 = σ𝑛𝑖=1 𝑓𝑖
1
𝜇1 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = 𝑓1 𝑥1 − 𝑥 + 𝑓2 𝑥2 − 𝑥 + ⋯ … 𝑓6 𝑥6 − 𝑥
100
𝑖=1
6
1 2
𝜇2 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = 𝑓1 (𝑥1 − 𝑥) 2 +𝑓2 (𝑥2 − 𝑥)2 + ⋯ … 𝑓6 (𝑥6 − 𝑥) 2
100
𝑖=1
6
1 3
𝜇3 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = 𝑓1 (𝑥1 − 𝑥) 3 +𝑓2 (𝑥2 − 𝑥)3 + ⋯ … 𝑓6 (𝑥6 − 𝑥) 3
100
𝑖=1
6
1 4
𝜇4 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = 𝑓1 (𝑥1 − 𝑥) 4 +𝑓2 (𝑥2 − 𝑥)4 + ⋯ … 𝑓6 (𝑥6 − 𝑥) 4
100
𝑖=1
Dr. Sonal Nirwal 22
Mark No. of Mid- 𝒇 𝑖 𝑥𝑖 (𝑥𝑖 − 𝑥) 𝒇𝑖 (𝑥𝑖 − 𝑥) 𝒇𝑖 (𝑥𝑖 − 𝑥) 𝟐 𝒇𝑖 (𝑥𝑖 − 𝑥)𝟑 𝒇𝑖 (𝑥𝑖 − 𝑥)𝟒
s stud. (f) point
(x)
5-15 10 10 100 -24 -240 5760 -138240 3317760
15-25 20 20 400 -14 -280 3920 -54880 768320
25-35 25 30 750 -4 -100 400 -1600 6400
35-45 20 40 800 6 120 720 4320 25920
45-55 15 50 750 16 240 3840 61440 983040
55-65 10 60 600 26 260 6760 175760 4569760
𝑁 = 100
෍ 𝒇 𝑖 𝑥𝑖 ෍ 𝒇𝑖 𝑥𝑖 − 𝑥) ෍ 𝒇𝑖 (𝑥𝑖 − 𝑥)𝟐 ෍ 𝒇𝑖 (𝑥𝑖 − 𝑥) 𝟑 ෍ 𝒇𝑖 (𝑥𝑖 − 𝑥)𝟒

= 3400 =0 = 21400 = 46800 = 9671200

6 6
1 1 2
21400
𝜇1 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = 0 , 𝜇2 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = = 214
100 100 100
𝑖=1 𝑖=1
6 6
1 3
46800 1 4
9671200
𝜇3 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 = = 468 , 𝜇4 = ෍ 𝑓𝑖 𝑥𝑖 − 𝑥 == = 96712
100 100 100 100
𝑖=1 𝑖=1
Dr. Sonal Nirwal 23
Example 3. Calculate the first four moments about mean of the following series:

(a) 𝜇1 = 0, 𝜇2 = 12, 𝜇3 = −12, 𝜇4 = 208.8

x 1 3 7 9 10
𝜇1 = 0, 𝜇2 = 36, 𝜇3 = 0, 𝜇4 = 2268
(b) x 4 7 10 13 16 19 22

(c) x 0 1 2 3 4 5 6 7 𝜇1 = 0, 𝜇2 = 2, 𝜇3 = 0, 𝜇4 = 11
f 1 8 28 56 70 56 28 8

(d) Marks 0-10 10-20 20-30 30-40 40-50

𝜇1 = 0, 𝜇2 = 125, 𝜇3 = −300, 𝜇4 = 37628
No. of 5 10 40 20 25
Students

(e) Class 0-10 10-20 20-30 30-40 40-50

interval 𝜇1 = 0, 𝜇2 = 120, 𝜇3 = 0, 𝜇4 = 36000
f 10 20 40 20 10

Dr. Sonal Nirwal 24

Moment about an Arbitrary Number (Raw Moments) (or working Mean)
If 𝑥1 , 𝑥2 … 𝑥𝑛 are the values of the variable x with the corresponding frequency
𝑓1 , 𝑓2 … 𝑓𝑛 respectively. Then the rth moment 𝜇𝑟′ about the number x=A is given by
𝑛
1 𝑟 = 0,1,2, … . 𝑛
𝜇 𝑟′ = ෍ 𝑓𝑖 (𝑥𝑖 − 𝐴)𝑟
𝑁 where 𝑁 = σ𝑛𝑖=1 𝑓𝑖 .
𝑖=1

For r=0,
1 𝑛 1 𝑛
𝜇0′ = σ 𝑓 (𝑥 − 𝐴) = 0 σ 𝑓 =1
𝑁 𝑖=1 𝑖 𝑖 𝑁 𝑖=1 𝑖
For r=1,
𝑛 𝑛 𝑛
1 1 𝐴
𝜇1′ 1
= ෍ 𝑓𝑖 (𝑥𝑖 − 𝐴) = ෍ 𝑓𝑖 𝑥𝑖 − ෍ 𝑓𝑖 = (𝑥 − 𝐴)
𝑁 𝑁 𝑁
𝑖=1 𝑖=1 𝑖=1
So on….

Dr. Sonal Nirwal 25

 Moment about the origin
If 𝑥1 , 𝑥2 … 𝑥𝑛 are the values of the variable x with the corresponding frequency
𝑓1 , 𝑓2 … 𝑓𝑛 respectively. Then the rth moment 𝜇𝑟′ about the number x=0 is given by
𝑛
1 𝑟 = 0,1,2, … . 𝑛
𝑣𝑟 = ෍ 𝑓𝑖 (𝑥𝑖 )𝑟
𝑁 where 𝑁 = σ𝑛𝑖=1 𝑓𝑖 .
𝑖=1

For r=0,
1 𝑛 0 1 𝑛
𝑣0 = σ 𝑓 (𝑥 ) = σ𝑖=1 𝑓𝑖 =1
𝑁 𝑖=1 𝑖 𝑖 𝑁
For r=1,
𝑛 𝑛
1 1
1
𝑣1 = ෍ 𝑓𝑖 (𝑥𝑖 ) = ෍ 𝑓𝑖 𝑥𝑖 = 𝑥
𝑁 𝑁
𝑖=1 𝑖=1
So on….

Dr. Sonal Nirwal 26

 Relation between 𝜇𝑟 and 𝜇𝑟′ Number
𝑛 𝑛
1 1
We know that 𝜇𝑟 = ෍ 𝑓𝑖 (𝑥𝑖 − 𝑥) 𝑟 𝜇𝑟′ = ෍ 𝑓𝑖 (𝑥𝑖 − 𝐴) 𝑟
𝑛 𝑁
𝑖=1 𝑖=1
𝑛
1 𝜇1′ = = (𝑥 −
= ෍ 𝑓𝑖 ((𝑥𝑖 − 𝐴) − (𝑥 − 𝐴)) 𝑟 𝐴)
𝑁
𝑖=1
𝑛
1
= ෍ 𝑓𝑖 ( 𝑥𝑖 − 𝐴 − 𝜇1′ )𝑟
𝑁
𝑖=1
𝑛
1 𝑟−1 𝜇 ′
= ෍ 𝑓𝑖 𝑥𝑖 − 𝐴 𝑟 − 𝑟𝐶1 𝑥𝑖 − 𝐴 1 + 𝑟𝐶2 𝑥𝑖 − 𝐴 𝑟−2 𝜇1′ 2 − ⋯ . + −1 𝑟 𝜇1′ 𝑟
𝑁
𝑖=1
′ ′
𝜇𝑟 = 𝜇𝑟′ − 𝑟𝐶1 𝜇𝑟−1 𝜇1′ + 𝑟𝐶2 𝜇𝑟−2 𝜇1′ 2 − ⋯ . + −1 𝑟 𝜇1′ 𝑟

𝜇1 = 0
𝜇2 = 𝜇2′ − 𝜇1′ 2
𝜇3 = 𝜇3′ − 3𝜇2′ 𝜇1′ + 2 𝜇1′ 3
𝜇: 4 = 𝜇4′ − 4𝜇3′ 𝜇1′ + 6𝜇2′ 𝜇1′ 2 − 3 𝜇1′ 4

Dr. Sonal Nirwal 27

 Relation between 𝜇𝑟 and 𝑣𝑟
𝑛
1
𝜇𝑟′ = ෍ 𝑓𝑖 (𝑥𝑖 − 𝐴) 𝑟
𝑁
𝑖=1
𝑣1 = 𝑥 𝜇1′ = = (𝑥 −
𝑣2 = 𝜇 2 + 𝑥 2 𝐴)
𝑣3 = 𝜇3 + 3𝜇2 𝑥 + 𝑥 3
𝑣4 = 𝜇4 + 4𝜇3 𝑥 + 6𝜇2 𝑥 2 + 𝑥 4

Karl Pearson’s 𝛽 and 𝛾 𝐂𝐨𝐞𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐭𝐬

𝜇3 2
𝛽1 = 3
𝜇2 𝛾1 = 𝛽1
𝜇4 𝛾2 = 𝛽2 − 3
𝛽2 = 2
𝜇2

Dr. Sonal Nirwal 𝛽 𝐂𝐨𝐞𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐭𝐬 𝛾 𝐂𝐨𝐞𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐭𝐬 28

Example 4. The first four moments of a distribution, about the value ‘2’ of the variable are 1, 16 and -40.
Show that the mean is 3, variance is 15 and 𝜇 3 = −86.

Solution. We have 𝐴 = 2, 𝜇1′ = 1, 𝜇 2′ = 16 and 𝜇 3′ = −40.

we have to show 𝑥 = 3, 𝜇2 = 15 and 𝜇 3 = −86.
𝜇1′ = 𝑥 − 𝐴
𝜇2 = 𝜇2′ − 𝜇1′ 2 1
𝜇3 = 𝜇3′ − 3𝜇2′ 𝜇1′ + 2 𝜇1′ 3 Mean 𝑥 = σ𝑛𝑖=1 𝑓𝑖 𝑥𝑖
𝑁
𝑁 = σ𝑛𝑖=1 𝑓𝑖

By using the above mentioned equations, we have

𝑥 =3
𝜇 2 = 15
𝜇3 = −86.

Example 5. The first four moments of a distribution, about the value ‘2’ are 1, 2.5, 5.5 and 16. Calculate
the first four moments about the mean and about origin.
𝜇1 = 0 𝑣1 = 3
𝜇2 = 1.5 𝑣2 = 10.5
𝜇3 = 0 𝑣3 = 40.5
𝑣4 = 168 29
Dr. Sonal Nirwal 𝜇4 = 6 Answer
Example 6. In the certain distribution, the first four moments about the point ‘x=4’ are -1.5, 17, -30 and
308. Find the moment about the mean and about origin. Also, calculate 𝛽1 and 𝛽2
𝜇1 = 0 𝑣1 = 2.5
𝜇2 = 14.75 𝑣2 = 21 𝛽1 = 0.492377
𝜇3 = 39.75 𝑣3 = 166 𝛽2 = 1.573398
𝜇4 = 342.3125 𝑣4 = 1332 Answer

Example 7. Calculate the first four moments about the point ‘x=15’ and hence find the moment about the
mean of the following distribution.
x 10 11 12 13 14 15 16 17 18 19 20 21 𝜇1 = 0, 𝜇2 = 5.5,
f 9 36 75 105 116 107 88 66 45 30 18 5 𝜇3 = 4.4, 𝜇4 = 77.8

Note: Moment about mean , 𝜇 𝑟

Moment about a point x=a, or moment about working mean, 𝜇 𝑟′
Moment about origin, 𝑣𝑟

30
Dr. Sonal Nirwal
 Skewness
Symmetrical
Distribution
When we do not have a perfect
Mean=Mode=Median symmetry curve, then the lack
of symmetry is called skewness.

Unsymmetrical
Distribution

Tail on the left side Tail on the right side

A.M<Mode A.M>Mode

Dr. Sonal Nirwal

Skewness implies a lack of symmetry in the
31
frequency distribution
 Method of Moments (For Measure of Skewness)

The measure of skewness by method of moments is called the moments coefficients of skewness.

Moments Coefficients of Skewness :

𝜇3
𝑆𝐾𝑀 =
𝜇23 where 𝜇2 and 𝜇3 are the
𝜇3 2 𝜇3
second and third moments
We know that 𝛽1 = ⟹ 𝛽1 = about mean of the distribution.
𝜇2 3 𝜇2 3
Also we know that 𝛾1 = 𝛽1
Thus, Moment of skewness can be defined as
𝜇3
𝛾1 =
𝜇2 3 Note: (a) If 𝛾1 (= 𝑆𝐾𝑀)>0, Positively skewed distribution.
(b) If𝛾1 (= 𝑆𝐾𝑀) =0, Symmetric distribution (No skewness)
Dr. Sonal Nirwal
(c) If𝛾1 (= 𝑆𝐾𝑀) <0,Negatively skewed distribution32.
Example 1. The first three central moment of a distribution are 0, 15 and -31. Find the moment coefficient
of skewness.
Answer: -0.53

Example 2. In the certain distribution, the first four moments about point 4 are -1.5, 17, -30 and 308.
Calculate the moment coefficients of skewness.
Answer: 0.7017

Dr. Sonal Nirwal 33

 kurtosis

The measure of peakedness or flatness of

a distribution is called kurtosis.

Measure of kurtosis
The measure of kurtosis is defined by 𝛽2 which can be defined as
𝜇4
𝛽2 =
𝜇22

where 𝜇2 and 𝜇4 are the second and fourth moments about mean of the distribution.
Note: (a) If 𝛽2 >3, the distribution is leptokurtic.
(b) If 𝛽2 =3, the distribution is mesokurtic.
(c) If 𝛽2 <3, the distribution is platykurtic.
Dr. Sonal Nirwal 34
 kurtosis

The kurtosis of a distribution can aso be measured by 𝛾2 which can be defined as

𝛾2 = 𝛽2 − 3.

Note: (a) If 𝛾2 > 0⟹ 𝛽2 − 3>0⟹ 𝛽2>3, the distribution is leptokurtic.

(b) If 𝛾2 =0⟹ 𝛽2 − 3 =0⟹ 𝛽2 = 3, the distribution is mesokurtic.
(c) If 𝛾2 < 0⟹ 𝛽2 − 3 <0⟹ 𝛽2 < 3, the distribution is platykurtic.

𝛾2 = 0 𝛾2 > 0 𝛾2 < 0
𝛽2 >3 𝛽2 >3 𝛽2 >3

Dr. Sonal Nirwal 35

Example 1. The first four moments of a distribution about x=4 are 1, 4,10 and 45. Obtained the various
characteristics of a distribution on the basis of the given information. Comment upon the nature of the
distribution.
Answer: 𝛽 = 2.89 < 3
𝜇1 = 0 Moment coefficient of skewness 𝛾 = 0
1 2
𝜇2 = 3 The distribution is symmetrical. The distribution is platykurtic.
𝜇3 = 0
𝜇4 = 26

Example 2. The first four moments of a distribution about the working mean 28.5 of a distribution are
0.294, 7.144, 42.409 and 454.98. Calculate the moments about mean. Also calculate 𝛽1 , 𝛽2 and comment
upon the skewness and kurtosis of the distribution.
Answer: 𝜇1 = 0 Skewness 𝛾1 = 1.9285
𝜇2 = 7.0576 𝛽1 = 3.7193 Positively skewed.
𝜇3 = 36.1588 𝛽2 = 8.2070
𝜇4 = 408.7896 𝛽2 = 8.2070 > 3
The distribution is leptokurtic.

36
Dr. Sonal Nirwal
Example 3. Calculate the first four moments about the mean for the following frequency distribution and
hence find the coefficient of skewness and kurtosis and comment upon the nature of the distribution.

Class-interval 5-10 10-15 15-20 20-25 25-30 30-35 35-40

Frequency 6 8 17 21 15 11 2

Answer: 𝜇1 = 0 𝛾1 = −0.4213 Negatively skewed

𝜇2 = 56 𝛽2 = 2.3925 and platykurtic.
𝜇3 = −176.5625
𝜇4 = 7502.9375

37
Dr. Sonal Nirwal
 Correlation

In a bivariate distribution, if the change in one relation affects the change in another
variable, then the variable is said to be correlated.
Scatter
Scatter
Heigh Weigh Price Quali Diagram
Diagram
t t if ty
(cm) (kg) Rice Sold
x y (Rs) (kg)
x y
50 4
10 200
60 7
20 150
70 9
30 100
80 10
40 80
90X y13 +ve or direct
50 X y
60 -ve or inverse
correlation correlation

Dr. Sonal Nirwal 38

 Correlation
Scatter
Mango Quali Diagram
Cost ty
(Rs) Sold
x (kg)
y
20 40
30 60
40 80
50 100
60 X 120
y
Perfect
correlation

Dr. Sonal Nirwal 39

 Measure of Correlation

Karl Pearson’s coefficient of correlation or Product moment

correlation coefficient
𝐶𝑜𝑣 𝑥, 𝑦 Cov stand for
𝑟 = 𝑟𝑥𝑦 = covariance
𝜎𝑥 𝜎𝑦

wher
e σ 𝑥 − 𝑥ҧ 𝑦 − 𝑦ത 1 1
𝐶𝑜𝑣 𝑥, 𝑦 = 𝜎𝑥 = ෍(𝑥 − 𝑥)ҧ 2 𝜎𝑦 = ത 2
෍(𝑦 − 𝑦)
𝑛 𝑛 𝑛
σ 𝑥 − 𝑥 ҧ 𝑦 − 𝑦ҧ
𝑟=
σ(𝑥 − 𝑥)ҧ 2 σ(𝑦 − 𝑦)ҧ 2
We can also Alternate Method
write wher
σ𝑋𝑌 e 𝑛 σ 𝑥𝑦 − σ 𝑥 σ 𝑦
𝑟= 𝑋 = 𝑥 − 𝑥ҧ
ൠ
𝑟𝑥𝑦 =
σ 𝑋2 σ𝑌 2 Y = 𝑦 − 𝑦
ത 𝑛 σ 𝑥2 − σ 𝑥 2 𝑛 σ 𝑦2 − σ 𝑦 2

Dr. Sonal Nirwal 40

Example 1. Find the correlation coefficient of following data:
x 1 3 5 7 8 10
y 8 12 15 17 18 20

Solution.
𝑛 σ 𝑥𝑦 − σ 𝑥 σ 𝑦 x y xy 𝒙2 𝒀2
r=
𝑛 σ 𝑥2 − σ 𝑥 2 𝑛 σ 𝑦2 − σ 𝑦 2
1 8 8 1 64
3 12 36 9 144
6 ∗ 582 − 34 ∗ 90 5 15 75 25 225
𝑟=
6 ∗ 248 − 34 ∗ 34 6 ∗ 1146 − 90 ∗ 90 7 17 119 49 289
8 18 144 64 324
𝑟 = 0.9879
10 20 200 100 400

෍𝑥 ෍𝑦 ෍ 𝒙𝒚 ෍ 𝑋𝟐 ෍ 𝑌𝟐

= 34 = 90 = 582 = 248 = 1446

Dr. Sonal Nirwal 41

Example 2. Find the correlation coefficient of following data:
x 100 200 300 400 500 600 700
y 30 50 60 80 100 110 130

Solution. x y xy 𝒙2 𝒚2
𝑛 σ 𝑥𝑦 − σ 𝑥 σ 𝑦
r= 100 30 3000 10000 900
𝑛 σ 𝑥2 − σ 𝑥 2 𝑛 σ 𝑦2 − σ 𝑦 2
200 50 10000 40000 2500
𝑟 = 0.997 300 60 18000 90000 3600
400 80 32000 160000 6400
500 100 50000 250000 10000
600 110 66000 360000 12100
700 130 91000 490000 16900

෍𝑥 ෍𝑦 ෍ 𝒙𝒚 ෍ 𝑥𝟐 ෍ 𝑦𝟐

= 2800 = 560 = 2700000 = 1400000 = 52400

Dr. Sonal Nirwal 42

 Rank Correlation/Spearman’s Rank Correlation Method

Sometime the given data is given in the form of merit or rank. Then we
find the rank correlation of the data by using the following formula.

6 σ 𝑑2 When the ranks are not

𝑟 = 1− tied
𝑛(𝑛2 − 1) wher 𝑑 = 𝑅𝑥 − 𝑅𝑦
e

Example 3. The marks secured by some students in Mathematics and Physics are given below.
Marks obtained in 10 15 12 17 13 16 24 14 22
maths (x)
Marks obtained in 30 42 45 46 33 34 40 35 39
physics (y)

Compute the rank correction coefficient.

Dr. Sonal Nirwal 43

x y 𝑹𝒙 𝑹𝒚 d 𝒅2

10 30 9 9 0 0
15 42 5 3 2 4
12 45 8 2 6 36
17 46 3 1 2 4
13 33 7 8 -1 1
16 34 4 7 -3 9
24 40 1 4 -3 9
14 35 6 6 0 0
22 39 2 5 -3 9

2 ෍ 𝑑2 = 72
6σ𝑑
𝑟 = 1−
𝑛(𝑛2 − 1)
6 ∗ 72
𝑟 =1− = 0.4.
9 81 − 1

Dr. Sonal Nirwal 44

Example 4. Ten competitors in a beauty contest ranked by three judges as given by
Judge 1 1 6 5 10 3 2 4 9 7 8
Judge 2 3 5 8 4 7 10 2 1 6 9
Judge 3 6 4 9 8 1 2 3 10 5 7
Use rank correlation method to declare which pair of judge have nearest approach to common taste of
beauty. Judge 1 Judge Judge 3 𝒅𝟏𝟐 𝒅𝟐𝟑 𝒅𝟏𝟑 ෍ 𝒅𝟏𝟐 2 = 200
Solution. 2
1 3 6 -2 -3 5 ෍ 𝒅𝟐𝟑 2 = 214

6 5 4 1 1 -2 ෍ 𝒅𝟏𝟑 2 = 60
5 8 9 -3 -1 4
𝑟12 = −0.21
10 4 8 6 -4 -2
3 7 1 -4 6 -2
2 10 2 -8 8 0 𝑟23 = −0.29
4 2 3 2 -1 -1
9 1 10 8 -9 1 𝑟13 = 0.63
7 6 5 1 1 -2
8 9 7 -1 2 -1
45
Since r13 is maximum so the pair of first and third judge has the nearest common approach to the common taste in beauty.
 Rank Correlation Coefficient for tied ranks
(3+4)/2=3.5
50 1 (5+6+7)/3=6
40 2
30 3.5
30 3.5
10 6
10 6 The modified formula for tie rank
correlation coefficient is
10 6
1 1
6 σ 𝑑 + 12 (𝑚1 − 𝑚1 ) + 12 (𝑚23 − 𝑚2 ) + ⋯ . .
2 3
𝑟 =1−
𝑛(𝑛2 − 1)

where m1, m2, m3 are the number of times a value is repeated.

Dr. Sonal Nirwal 46

Example 5. Obtained the rank correlation coefficient of the following data:
X 68 64 75 50 64 80 75 40 55 64
Y 62 58 68 45 81 60 68 48 50 70
Solution.
X Y RX RY d 𝒅𝟐
Here m has the values 2,3,2.
1 1
68 62 4 5 -1 1 σ 𝑑2 + (𝑚13 − 𝑚1 ) + (𝑚23 − 𝑚2 )
6 12 12
64 58 6 7 -1 1 1
+ (𝑚33 − 𝑚3 )
𝑟 =1− 12
75 68 2.5 3.5 -1 1 𝑛(𝑛2 − 1)
50 45 9 10 -1 1
𝑟 = 0.545
64 81 6 1 5 25
80 60 1 6 -5 25
75 68 2.5 3.5 -1 1
40 48 10 9 1 1
55 50 8 8 0 0
64 70 6 2 4 4

47
 Regression Analysis
• Regression Analysis is the technique for measuring or estimating the relationship among variables.
• Regression analysis provides estimates of values of the dependent variables from the values of
independent variables.
• The device used to accomplish this estimation procedure is the regression line.
• The regression line describes the average relationship between variable x and y.
Regression Lines

Regression line of x on y
Regression line of y on x Both
𝑥 − 𝑥ҧ = 𝑏𝑥𝑦 𝑦 − 𝑦ത
𝑦 − 𝑦ത = 𝑏𝑦𝑥 𝑥 − 𝑥ҧ line 𝑛 σ 𝑥𝑦−σ 𝑥 σ 𝑦 𝑟𝜎𝑥
𝑛 σ 𝑥𝑦−σ 𝑥 σ 𝑦 𝑟𝜎𝑦 interse 𝑏𝑥𝑦 = or 𝑏𝑥𝑦 =
𝑏𝑦𝑥 = or 𝑏𝑦𝑥 = 𝑛 σ 𝑦2 − σ 𝑦 2 𝜎𝑦
𝑛 σ 𝑥2− σ 𝑥 2 𝜎𝑥 ct at
bxy is called regression coefficient of x on y
byx is called regression coefficient of y on x point
𝑥ҧ ,ഥ𝑦 y: independent variable
x: independent variable
x: dependent variable
y : dependent variable
Used to calculate x for given x
Used to calculate y for given x
Slope of line: 1/bxy
Slope of line: byx
Dr. Sonal Nirwal 48
 Angle between two lines of regression
If 𝜃 is the acute angle between two regression lines in the case of two variables x and y, show that
1−𝑟 2 𝜎𝑥 𝜎𝑦
Tanθ = . 2 2
𝑟 𝜎𝑥 +𝜎𝑦
where r 𝜎𝑥 and 𝜎𝑦 have their usual meanings. Explain the significance of the formula when r=0 and r=±1

Proof: Equations to the lines of regression of y on x

𝑦 − 𝑦ത = 𝑏𝑦𝑥 𝑥 − 𝑥ҧ
𝑟𝜎𝑦
Or we can write 𝑦 − 𝑦ത = 𝑥 − 𝑥ҧ
𝜎𝑥
𝑟𝜎𝑦
Slope of the line 𝑚1 = (1)
𝜎𝑥

Equations to the lines of regression of y on x and x on y

𝑥 − 𝑥ҧ = 𝑏𝑥𝑦 𝑦 − 𝑦ത
𝑟𝜎
Or we can write 𝑥 − 𝑥 ҧ = 𝑥 𝑦 − 𝑦ҧ
𝜎𝑦
𝜎𝑦
(𝑥 − 𝑥)ҧ = 𝑦 − 𝑦ത
𝑟𝜎𝑥
𝜎𝑦
Slope of the line 𝑚2 =
𝑟𝜎𝑥
(2)
Dr. Sonal Nirwal 𝑚2 − 𝑚1 49
We know that Tanθ = ±
1+𝑚2 𝑚1
 Angle between two lines of regression
𝑚2 − 𝑚1
tanθ = ±
1 + 𝑚2 𝑚1
Substituting the values from Eq. (1) and Eq. (2), we have
1 − 𝑟 2 𝜎𝑥 𝜎𝑦
tanθ = ± . 2
𝑟 𝜎𝑥 + 𝜎𝑦2
Since −1 ≤ 𝑟 ≤ 1, thus 𝑟 2 ≤ 1 and 𝜎𝑥 𝜎𝑦 are positive.
+ve sign gives the acute angle between the lines.
1 − 𝑟 2 𝜎𝑥 𝜎𝑦
tanθ = . 2
𝑟 𝜎𝑥 + 𝜎𝑦2
𝜋
When r=0, tanθ = ∞ which gives 𝜃 = .
2
Thus the two line of regression are perpendicular of each other.
When r=±1, tanθ = 0 which gives 𝜃 = 0 or 𝜋
Hence the lines of regression coincide and there is perfect correlation between x and y.

Dr. Sonal Nirwal 50

 Properties of regression coefficient
Property 1: Correlation coefficient is the GM between the regression coefficients.
𝑟𝜎𝑦 𝑟𝜎𝑥
Proof: The coefficients of regression are 𝜎 and 𝑦 .
𝑥
𝑟𝜎𝑦 𝑟𝜎𝑥
Geometric mean between them= . = 𝑟. 𝑟=r=correlation coefficient.
𝜎𝑥 𝜎𝑦

Property 2: If one of the regression coefficient is greater than unity, then other will be less
than unity.
𝑟𝜎𝑦 𝑟𝜎𝑥
Proof: The two regression coefficients are 𝑏𝑦𝑥 = and 𝑏𝑥𝑦 =
𝜎𝑥 𝜎𝑦
Let 𝑏𝑦𝑥 > 1, then 1/𝑏𝑦𝑥 < 1
Since 𝑏𝑦𝑥 . 𝑏𝑥𝑦 = 𝑟 2 and we know that 𝑟 2 ≤ 1
Thus, 𝑏𝑦𝑥 . 𝑏𝑥𝑦 ≤ 1
𝑏𝑥𝑦 ≤ 1/𝑏𝑦𝑥 <1
𝑏𝑥𝑦 <1.
Similarly, if 𝑏𝑥𝑦 >1, the 𝑏𝑦𝑥 < 1.
Dr. Sonal Nirwal 51
Example 1. (i) Obtain the line of regression of y on x for the following data :
x 1.53 1.78 2.60 2.95 3.42
y 33.50 36.30 40.00 45.8 53.50
(ii) The following data regarding the height (y) and weights (x) of 100 college students are given:
σ 𝑥 = 15000, σ 𝑦 = 6800, σ 𝑥 2 = 2272500, σ 𝑦 2 = 463025 𝑎𝑛𝑑 σ 𝑥𝑦 = 1022250.
Find the regression line of height on weight.
Answer: (i) y=17.932+9.726x
(ii) y=0.1x+53.

Example 2. For 10 observations on price (x) and supply (y), the following data were obtained (in
appropriate units):
σ 𝑥 = 130, σ 𝑦 = 220, σ 𝑥 2 = 2288, σ 𝑦 2 = 5506 𝑎𝑛𝑑 σ 𝑥𝑦 = 3467.
Obtained the two regression and estimate the supply when the price is 16 units.
Answer: y=8.805+1.015x
x=0.911y-7.0508.

When x=16 units.

Y=1.015*16+8.805=25.045 units.

52
Example 3. In a partially destroyed laboratory record of an analysis of correlation data, the following
results only are legible.
Variance of x=9
Regression equations: 8x-10y+66=0, 40x-18y=214.
Find (a) the mean values of x and y (b) the standard deviation of y and the coefficient of correlation
between x and y.
Answer: Since both the lines of regression pass through the point 𝑥ҧ ,ഥ𝑦 , therefore we have
8 𝑥ത − 10 𝑦ത + 66 = 0
40 𝑥ത − 18 𝑦ത = 214
By solving these equation, we get 𝑥ത = 13, 𝑦ത = 17
(b) r=0.6 and 𝜎𝑦 = 4

Example 4. Two lines of regression are x=-0.4y+6.4 and y=-0.6x+4.6

Find (a) the mean values of x and y (b) the coefficient of correlation between x and y.
Answer: By solving these equation, we get 𝑥ത = 6, 𝑦ത = 1
(b) r=-0.48989

53