Practical Manual

GEO.525: QUANTITATIVE TECHNIQUES IN

GEOGRAPHY
Academic Session: 2020-2022

Name of the student - : Saurbh

Registration ID- : 20mageog13

Department : Geography

School : Environment and Earth Sciences

Course coordinator : Dr. Kiran K. Singh

Saurbh

Signature of Student Signature of Course coordinator

DEPARTMENT OF GEOGRAPHY
SCHOOL OF ENVIRONMENT AND EARTH SCIENCES
CENTRAL UNIVERSITY OF PUNJAB, BATHINDA
 Geo 525 Quantitative Techniques in Geography

Table of Contents

Contents
Exercise 1: Lorenz Curve..................................................................................................................2
Exercise 2: Ginni’s Coefficient...........................................................................................................5
Exercise 3: Location Quotients..........................................................................................................8
Properties ......................................................................................................................................8
Limitations of Location Quotient......................................................................................................9
Exercise 4: Nearest Neighbour Analysis.......................................................................................... 10
Exercise 5: Transportation Network Analysis & Connectivity Metrix .............................................. 13
Exercise 6:- Simple Linear Regression:- .......................................................................................... 23
Exercise 7 Multiply regression:- ...................................................................................................... 26
Exercise 8:- Gravity model............................................................................................................... 29
Exercise 9 Theissen polygon: - ........................................................................................................ 32

Introduction
Quantitative methods are concerned with objective measurements and statistica l, mathematical,
or numerical analysis of data collected through polls, questionnaires, and surveys, as well as the
modification of pre-existing statistical data using computational tools. Quantitative research is
gathering numerical data and interpreting it across groups of people or understanding a
phenomenon. The goal of a quantitative research study is to establish a relationship within a
population between one item [an independent variable] and another [a dependent or outcome
variable]. Quantitative research designs can be descriptive [subjects are usually only measured
once] or experimental [subjects are measured before and after a treatment].

DATE:

Exercise 1: Lorenz Curve

Lorenz Curve developed by American economist Max Lorenz in 1905. Lorenz curve are used to
measure the inequality in the distribution of wealth or income, concentration of population and
other demographic attributes etc. It is to compare cumulative percentage distribution of one
variables with another variable’s cumulative percentage distribution. If both cumulative
 Geo 525 Quantitative Techniques in Geography

percentage distribution are same, it means there has no inequality. On the other hand, if both are
no same, it means the distribution is not equal.
Graphically, cumulative distribution of the population are plotted on the horizontal axis and
cumulative distribution of the income or wealth are plotted on the vertical axis. All these point
will be pointed on the diagonal line of the square. This diagonal line are to present as line of
equal distribution. When we plot all the point according to data, the shape of the curve are
formed which is called ‘Lorenz Curve’.

Some other measures such as Coefficient of variation, Location Quotient, Sopher’s Index etc. of
inequality that are used in place of Lorenz Curve to show the level of inequality in any
distribution.

 Derive the Lorenz curve using the data given in table 1.

Data:-
Table 1: To show the Income and number of persons in a place (X)
Table 1,INCOME AND NUMBER OF PERSONS IN A PLACE (X)

Sr. No. Income ('000) Population

1 10 14
2 20 12
3 40 10
4 50 8
5 80 6

Method/formula:-

 First of all, put all the data in Ms-Excel

 Derive the cumulative frequency of income and population( by adding each other)
 After that find out the cumulative % of population and cumulative % of income (by using
this formula: - Cumulative of population or cumulative of income/total population*100).

Solution:-

Cumulative of Cumulative of C % of C%
Region income('000) Population income Population population Income
0 0 0 0
A 10 14 10 14 28 5
B 20 12 30 26 52 15
C 40 10 70 36 72 35
D 50 8 120 44 88 60
E 80 6 200 50 100 100
 Geo 525 Quantitative Techniques in Geography

 Select the data which we want to plot Lorenz Curve.

 Click on the charts commands and select the charts. For examples:- I have used “scatter
with straight lines and matters” charts for making the Lorenz Curve.

Result:-
Lorenz Curve
Lorenz Equality
100

90

80

70

60
C % Income

50

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100
C % Population

Figure 1, Lorenz Curve

 Geo 525 Quantitative Techniques in Geography

Interpretation:-
 This Lorenz curve represents income distribution.
 In this data, two percentage are not same, so we do not get a straight line, in spite of
we get a curve below the line of equal distribution.
 This Lorenz Curve given above shows deficient income inequality in the population
of ‘X’ place.
 Curve shows the inequality in the distribution of income in relation to the distribution
of population of the place.
 We compared the income distribution of ‘x’ place with a straight line which
representing perfect equality.
 At the 28th percentile of population, the value of the Lorenz curve is 5. In other
words, this Lorenz curve estimates that 28% of the population takes 5% of the total
income. If ‘X’ place have a perfect equality, 28% population would get 28% income.

DATE:

Exercise 2: Ginni’s Coefficient

Ginni’s Coefficient or Gini index measure the area of between the line of equal distribution and a
given Lorenz Curve. It developed by Italian statistician Corrado Gini in 1912. Ginni’s
Concentration ratio ‘G’ has provided a more accurate quantitative measures of this deviation. We
can use it as a gauge of economic inequality, measuring income distribution, wealth distribution.
But wealth is more difficult to measure than income, so Ginni’s coefficient always show income
distribution without any specifying. The value of Ginni’s concentration ration ‘G’ or coefficient
ranges varies from zero to 1. Higher value of coefficient is 1 that will show a higher level of
inequality in the distribution. And lower value of coefficient is zero that will show low
inequality.

 Derive the Ginni’s Coefficient using the data given in table 3 (use % area
and % population figures).

Data:-
Ta bl e 2, AREA, AND DENSITY OF POPULATION IN DISTRICTS OF KERALA 2011

Population
District % Area (sq km) % Population (2011)
density (2011)

1 Idukki 255 11.21 3.32

2 Wayanad 383 5.48 2.45
3 Pathanamthitta 451 6.82 3.58
4 Palakkad 626 11.54 8.41
5 Kasaragod 657 5.12 3.91
 Geo 525 Quantitative Techniques in Geography

6 Kannur 852 7.62 7.55

7 Kottayam 895 5.68 5.91
8 Thrissur 1031 7.79 9.34
9 Kollam 1061 6.39 7.89
10 Ernakulam 1071 7.88 9.83
11 Malappuram 1157 9.15 12.31
12 Kozhikode 1316 6.04 9.24
13 Alappuzha 1503 3.64 6.37
14 Thiruvananthapuram 1508 5.64 9.89
Method/formula:-
We can derive the value of Ginni’s coefficient by using the method of numerical analysis. By
using the numerical method we can derive Ginni’s coefficient in a small computer programme
also. Method of deriving the value of ‘G’ is as given below:-
1
G=100∗100 *( ⅀Y1*X1k+1 - ⅀X1*Y1k+1)

Here X1 and Y2 are the cumulative percentage of population and area. And 100*100 is the
length and breath. In the bracket, the value is to be taken without plus or minus sign.
Solution (Insert Excel table and figure):-
After plotting the Lorenz Curve which showing the inequality in the population distribution
over the area of the district of Kerala, by using the column 4 and column 5.
 Geo 525 Quantitative Techniques in Geography

Ta bl e 3, CUMULATIVE PERCENTAGE AREA AND POPULATION OF KERALA 2011

% Cumulati Cumulative
Population % Area
District Populatio ve % of % of A B
density (2011) (sq km)
n (2011) area Population
X Y X1 Y1 Y1*X1k+1 X1*Y1k+1
1 Idukki 255 11.21 3.32 11.21 3.32 55.4108 64.6817
2 Wayanad 383 5.48 2.45 16.69 5.77 135.6527 156.0515
3 Pathanamthitta 451 6.82 3.58 23.51 9.35 327.7175 417.5376
4 Palakkad 626 11.54 8.41 35.05 17.76 713.4192 759.5335
5 Kasaragod 657 5.12 3.91 40.17 21.67 1035.6093 1173.767
6 Kannur 852 7.62 7.55 47.79 29.22 1562.3934 1678.863
7 Kottayam 895 5.68 5.91 53.47 35.13 2152.0638 2377.811
8 Thrissur 1031 7.79 9.34 61.26 44.47 3008.3955 3207.574
9 Kollam 1061 6.39 7.89 67.65 52.36 3954.7508 4207.154
10 Ernakulam 1071 7.88 9.83 75.53 62.19 5266.2492 5626.985
11 Malappuram 1157 9.15 12.31 84.68 74.5 6758.64 7091.103
12 Kozhikode 1316 6.04 9.24 90.72 83.74 7901.7064 8174.779
13 Alappuzha 1503 3.64 6.37 94.36 90.11 9011 9436
14 Thiruvananthapuram 1508 5.64 9.89 100 100 0 0
Total 100 100 41883.009 44371.84
B-A= 2488.831
Ginni's concentration ratio G= 0.248883

Lorenz Curve 100

100
90.11
83.74

80 74.5
Cumul ative % of Population

62.19
60 52.36
44.47

40 35.13
29.22
21.67
17.76
20
10 9.35
5.77
3.32

0
0 20 40 60 80 100
Cumul ative % of a rea

Lorenze Equality
 Geo 525 Quantitative Techniques in Geography

We can also measure the inequality through Ginni’s Concentration ratio ‘G’.
Column 5 and 6 of the table 3, give the cumulative percentage of population X1 and Y1.
Column 7, give the values of Y1*X1+1, By multiplying each value of Y with the value of
following Y that is Y1 with X2, Y2 with X3 and so on. In the next column 8, give the values of
X1*Y1+1, in this value we do reserve of it that is multiply X1 with Y2, X2 with Y3 and so on.
Total of column 7 and 8 of the table 3 will give the values of ⅀ (X1Y1+1) and ⅀ (Y1X2+1).
The value of the ‘G’ can be calculated as:-
1 1
G = 100∗100 (44371.84-41883.009) =100∗100 *2488.831 =0.248883Ans.

Result:-
Here the value of “G” is 0.248883 which indicates less amount of inequality in the distribution of
population over area.

Interpretation:-
 In this example, Y-axis denotes ‘Cumulative % of population’ and X-axis denotes
‘cumulative % of area.
 In this example, straight line shows here line of equality, it means there is zero
inequality. And curve shows here inequality. Gini index which show the value of
inequality between Zero to 1. In this example, Gini Index is close to zero it means less
amount of inequality or greater equality. If everyone had exactly the same amount of
money, Gini Index would register a reading of 0. But low numbers are not always a
perfect indicator of economic health. If the no. can be multiplied by 100 in order to
express it as a percentage.

Date:-

Exercise 3: Location Quotients

Location Quotients (LQ) are a way of quantifying It are used to determine the cluster, industry,
occupation, dispersal, or demographic in a region as compared to entire region or nation. On the
other word, location quotients is a ratio that compares a region to a larger region according to
some asset. It is an index for comparing the importance of a phenomenon’s share with the
general phenomenon over a certain area. Location Quotients are frequently used in demography,
economics, and any type of location analysis.

Location quotient is useful in demographic studies because it shows what makes the region’s
demographics unique in comparison to its state and/or the nation.

Properties:
 If LQ is less than 1, then it indicates opportunity to develop business/industry in the local
area as there is less competing area
 If it is greater than 1, then the area has less opportunity to develop new project as more
industries or other economic sector is concentrated.
 If LQ = 1, the state of nation has a share of the total in accordance with its share of
 Geo 525 Quantitative Techniques in Geography

Advantages of location Quotient:-

It evaluates the strength and size of a particular Economic sector in a region.
Identifies growth opportunities and comparative regional advantages.
Limitations of Location Quotient:
It shows concentrations in a single point of time and does not account for trends over time.
Level of productivity of workers or qualitative aspects are assumed to be equal

 Derive the location quotients value by using the data which give on table
4.

Data:-
Table 4, Native speaking population and total population in different regions of the X

Region of the Native speaking Total Population

‘X’ population(thousands) (thousands)
(S= ‘Share of (T= ‘Total
Phenomenon’) phenomenon)
A 200 700
B 800 2000
C 50 300
D 2200 3000
E 3250 6000

Method/Formula:-
𝑆/⅀(𝑆)
Location Quotients =
𝑇/⅀(𝑇)

‘S’ = Share of the phenomenon observed,

⅀(S)=is the sum of the shares,
‘T’ = Total of phenomenon;
⅀ (T) = Grand total (sum of ‘T’).

Solution:-
Table 5, Native speaking population and total population in different regions of the X

Native speaking Total

Region % of native % of native
Population(thousand Population(thousands) Location
of the speaking in speaking in
s),(S= Share of ,(T=Total Quotient
'X' sub-region region
Phenomenon) phenomenon)
A 200 700 6.153846154 11.66666667 0.527473
B 800 2000 24.61538462 33.33333333 0.738462
 Geo 525 Quantitative Techniques in Geography

C 50 300 1.538461538 5 0.307692

D 2200 3000 67.69230769 50 1.353846
E 3250 6000 100 100 1

Interpretation:
 A Location Quotient less than one indicates that the percentage of native speaking in the
sub region is lower than the national average.
From above result:

A, B, C sub-regions have LQ values less than 1 it means less % of native language

speaker than entire Region.
 A Location Quotient greater than one indicates that the sub-region has a higher % of the
native language speaker compared to entire region.
From above result:

D sub- region has the LQ value greater than 1 means more % of native language
speaker than entire region.

 Location Quotient equal to one indicates that equal % of native language speaker equal to
the % of native language speaker in entire region.
From above result:
Region E have equal % of native language speaker

Date:-

Exercise 4: Nearest Neighbour Analysis

Nearest neighbor analysis shows the patterns of locating features over a geographical space. In
the other words, it describes phenomenon based on its distance from one to another in space.
This analysis is done with the help of index called Nearest Neighbour Index. It measures the
ratio of mean of the observed nearest distance with the average expected distance over space to
get a Randomness index of point distribution called ‘Rn’.
It do not show only distance between two locations over a space but also show the nearest place
to it. We can be used nearest neighbour analysis to describe both human and physical features to
get to know the proximity of points or locations. It provides us a numerical value for ‘clustering’
of a geographical phenomenon in space, with the help of this value, we can compare more
accurately with other places, and analyse the distribution of schools, hospitals, buildings,
settlement etc.
Nearest neighbour analysis provides basis of measuring point pattern of region which help in
understanding the spatial processes of the distribution of human activities.

Set of point of an area can be arranged in a number of ways but three basic patterns are
recognized:-(1) Regular or uniform: - That pattern will occur when interval between points is
similar.
 Geo 525 Quantitative Techniques in Geography

(2) Clustered: - In this arrangement, they seems in bunches separated by gaps.

(3) Random: - In this, the spacing could have been determined by using the random number
table.

 Derive the Nearest neighbour analysis of the following data.

Data:-

Table 6, Nearest neighbour and ground distance of settlements

Sr. no. Settlement Nearest neighbor Ground distance in km

1 A F 1.19
2 B A 1.44
3 C B 3.52
4 D E 2.67
5 E A 1.22
6 F A 1.19
7 G F 1.97

Method/Formula:-
Rn=2d*√ (n/A)

d = Average of ground distance

n = No. of total places
A = Total area of polygon

Solution:-
 Select the point or settlements on a map which we want to analyse, for example,
figure X show the distribution of 7 villages located in a part of various district of
Haryana. These point are collected from ‘Google Earth Pro’. Total area of this
polygon is 35 km seq. met.
 Geo 525 Quantitative Techniques in Geography

 With the help of ‘RULER’ option in Google Earth Pro, connect the all point with
their nearest neighbour village and measure their crow-fly distances to calculate the
average NND for the area under consideration (table-5)

Sr. no. Settlement Nearest neighbor Ground distance in km

1 A F 1.19
2 B A 1.44
3 C B 3.52
4 D E 2.67
5 E A 1.22
6 F A 1.19
7 G F 1.97
Total 7 13.2
Average(d)= 1.885714286

 Calculate the mean of the measure distance, by doing- total sum of ground
distance divided by total villages.
 Put all the value in NNA formula. Result will come.

Area= 35
d= 1.885714286
n= 7
Rn= 2d*√(n/A)
1.686634132

Result:-
The value of nearest neighbour analysis is 1.686634132.

Interpretation:-
 Nearest neighbour analysis value will be between 0-2.15.
 The smaller the nearest neighbour analysis value means clustered pattern. Higher the
value means more regular pattern. If value of NNA is ‘0’ it show complete
clustering that means maximum aggregation of all the point at one location.
 If value of NNA is 1, that means random distribution. Value 2.5 indicates a regular
pattern.
 Geo 525 Quantitative Techniques in Geography

 In the above exercise, value of NNA is 1.6866634132, that means a near random
situation or we can say that pattern is uniform.

Date

Exercise 5: Transportation Network Analysis & Connectivity Metrix

Transportation analysis is a graph. In the graph, symbolic representation are done of a network
and its connectivity. That graph do not represent the real value of network. When the conversion
of a real network is converted into planar graph, that time some basis rules are followed which
has been given:-
 Every terminal and intersection point become a node
 Each connected node link with a straight segment.
 Vertex: - vertex is also called node. A node or vertex is an intersection or terminal point
of a graph.
 Edge (link):- a link between two vertexes is called Edge or Link.
 Connectivity matrix:- it consist of one’s 1 and zeros(0). When two vertices are directly
connected by an edge, code with a 1, when two locations are not directly connected by a
edge, code with a 0.
 Diameter:- the diameter is the length of the shortest path between the most distanced
nodes in graph. On the other word,‘d’ measures the extent of a graph and the topological
length between two nodes. For example:- the number of edges between the furthest nodes
(G and D) of the above graph is 3. So, diameter will be 3 (d=3).

Edges: - There are 8 edges.

Vertexes: - 7 Vertexes.
Diameter:- 3
Connectivity matrix:-
Table 7, Connectivity matrix

A B C D E F G
 Geo 525 Quantitative Techniques in Geography

A 0 1 0 0 0 1 1

B 1 0 1 1 0 1 0

C 0 1 0 0 1 0 1

D 0 1 0 0 1 1 0

E 0 0 1 1 0 0 0

F 1 1 0 1 0 0 0

G 1 0 1 0 0 0 0

The least accessible node is ‘G’ only two direct connection is there. The most accessible node is
‘B’ where four direct connection is there.

1. Define Alpha index and derive index for the followings:-

 Alpha index:- it is a most useful measures of connectivity, particularly in complex
network. That evaluates the number of cycles in a graph in comparison with the
maximum number of cycles. If the higher the index, it means more a network in
connected. Alpha value have range value from 0 to 1. If alpha value is 1, that means
a completely connected network.

Data:-
Case 1:-

Method/Formula:-
 α= actual circuit/ maximum circuits
Or
 α= e-ν+1/ 2ν-5

e= edge, v= vertex
 Geo 525 Quantitative Techniques in Geography

Solution:-
E= 10
V= 7
10−7+1
= ;
2∗7−5
4
= ;
9
= 0.44 Ans.
Result:- the value of alpha index is 0.44 Ans.
INTERPRETATION:
For Alpha Index:

The higher the alpha index, the more a network is connected. It means network A (0.44) has
high network connectivity as simple networks will have a value of 0.

Case-2

Method/Formula:-
 α= actual circuit/ maximum circuits
Or
 α= e-ν+1/ 2ν-5

e= edge, v= vertex
Solution:-

E=4

V= 6

4-6+1/2*6-5

= -0.14 ans.

2. Define Beta Index and derive index for the followings

 Geo 525 Quantitative Techniques in Geography

 Beta: - it is a very measure of connectivity. It measure the level of connectivity in a graph. It

can be derived by dividing the total number of edges or arcs in a network by total number of
nodes or vertices.

Some characteristic of ᵦ index are:-

a. If beta value is less than 1 it means disconnected networks.
b. If beta value for any network connect with one circuit would always be equal to 1.
c. If beta value is exceeds from 1, it would always be for a complicated network structure
having more than one circuit.

Data:-
Case -1

Method/Formula:-
β = arcs/ nodes

Arcs: - edges
Nodes: - vertex
Solution:-
V=4
E=3
=3/4
=0.75.

Result: - Beta value is 0.75 Ans.

Case-2
 Geo 525 Quantitative Techniques in Geography

Method/Formula:-
β = arcs/ nodes

Solution:-
V= 5
E= 8
β = E/V
=8/5
= 1.6 Ans
Result:- beta value of data is 1.6 ans.
INTERPRETATION:
For Beta Index:

Here, our Beta value of data is 1.6 and hence this Network has complex network system as it
values is more than [Link] means this network system low degree of connectivity.

- And network is is neither complex nor simple as is value is 1.

3. Define Gamma Index and derive index for the followings:

 Gamma: - Gamma is a measure of connectivity in which number of edges
and vertices of a given transportation network are considered or observed. On
the other words, it is ratio between the number of observed links and the
number of possible links. The gamma values varies from 0 to 1. When gamma
value is 1 it means there is a completely connected network. It can be measure
in the form of percentage.

Data:-
 Geo 525 Quantitative Techniques in Geography

Method/Formula:-
ү=e/3(v – 2)

e= edges
v= vertex
Solution:-
E= 3
V= 4
= 3/3(4-2)
=0.50 ans.

Method/Formula:-
ү=e/3(v – 2)

e= edges
v= vertex
Solution:-
e= 4
v= 4
Y= 4/3(4-2)
= .66 ans.
Result:- the value of gamma is .66 ans.
 Geo 525 Quantitative Techniques in Geography

Case-3

Method/Formula:-
ү=e/3(v – 2)

e= edges
v= vertex
Solution:-
E= 8
V= 5
8
Ү=3(5−2)
=8/9
=0.88 Ans. Or
If we want to beta value in percent then multiply by 100
0.88*100= 88.8% Ans.

Result:- Gamma value is 0.88 or 88.8% Ans.

Interpretation:
For Gamma Index:

When a value of Gamma Index is 1, it indicates a completely connected network. Thus

network has almost complete connection network system as it value is 0.88

4. Define Cyclomatic number and derive CN for the followings

 Cyclomatic number:-

Data;-
Case 1.
 Geo 525 Quantitative Techniques in Geography

Method/Formula: - when two or more sub-region, then the formula for Cyclomatic number will
be:-
=a–n+x
a= number of arcs (edges)
n= number of nodes
x= number of subgraphs

Solution:-
a=4
n=6
= 4-6+2
=0 ans.
Result:- Cyclomatic number is 0.

Case-2

Method/Formula:-
When sub-graphs are not in data then this formula will be used for finding the Cyclomatic
Number.
Cyclomatic number = a – (n – 1)
Or
a –n+1
 Geo 525 Quantitative Techniques in Geography

a= number of arcs (edges)

n= number of nodes

Solution:-
a=10
n=7
= 10-7+1
=4 Ans.
Result:- Cyclomatic number is 4.

5. What is Diameter in Transport Network Analysis? Derive diameter

for the followings
 Diameter: - the diameter is the length of the shortest path between the most
distanced nodes in graph. On the other word, ‘d’ measures the extent of a
graph and the topological length between two nodes.
Case -1

The value of diameter is 3 Ans.

Case -2

The value of diameter is 3 Ans.

6. Define Detour index and derive index for the followings:

 Geo 525 Quantitative Techniques in Geography

 Detours: - For shortest distance, travellers use the straight route between two
places that type of route also known ‘desire line’. In the other words, the
detour index is the actual distance calculated as a percentage of the desire line
distance. In the detours index, mostly of the time actual distance is always
longer than the desire line distance. The value of detours index can never be
less than 100. If the value of detours index is lower, it means more direct is a
given route.

Uses: To indicate straight or direct road connectivity.

Limitation: mathematically or theoretically show high to low accessibility between

different two nodes, but practically sometime it may not applied such as heavy traffic, traffic congestion,
rapid deterioration of the road surface hinder the connectivity between two nodes.
 There are structural problems in the development and maintenance of road
network due to their varying width, depth of soiling, construction material,
road lack of proper bridge and non- metalled road cannot mention by Detour
Index.

Data:-

Method/Formula:-

Detour index= actual route distance/ straight line distance × 100

Solution:-
Detour Index:-
Actual route distance=22
Straight line distance=10
22
= ∗ 100
10
=220 Ans.

Result: - Detour index value is 220%.

 Geo 525 Quantitative Techniques in Geography

Interpretation:-
In fact, the actual route distance is almost always longer than the desire line distance, then the
detour index will be greater, in almost all cases, than 100 and in the nature of things can never be
less than 100. It is obvious that lower the detour index, the more direct is a given route. The
detour index is used for assessing the effects which the addition or abstraction of links produce in
a given network.

7. Derive Alpha, Beta, Gamma Index, Cyclomatic Number, and

Diameter for the followings:-
Ta bl e 8, Al pha , Beta , Ga mma i ndex, cycl oma ti c number, di a meter of res pecti ve groups .

Alpha 0 0 0 0 0.33 0.66 1.3

Index
Beta Index 0 0.5 0.66 0.75 1 1.25 1.75
Gama 0 0.33 0.66 0.5 0.66 0.83 1.16
Index
CN 0 0 0 0 1 2 4
Diameter 0 1 2 3 2 2 2

Exercise 6:- Simple Linear Regression:-

Regression is a statistic method which estimating the relationship between dependent variables
and independent variables. On the basis of variables, regression are two types, one is simple
linear regression and second is multiple linear regression. In simple linear regression, we uses
one independent variables to predict the outcome of dependent variables Y. The relationship
between the independent and dependent variable is linear: the line of best fit through the data
points is a straight line
 Dependent variables: - that variables who is being affected by other variables is called
dependent variables.
 Independent variables: - that variables who affect the outcome of the values of any
variables is called independent variables. For example, agricultural production depend on
the rainfall in any region. It means rainfall affect the agricultural production and
agricultural production will not affect the values of the rainfall. So agricultural
production is dependent variable and rainfall is independent variables.
 Geo 525 Quantitative Techniques in Geography

 Derive the simple linear regression value by using the data which give on
table 5.
Table 9, Ads cost and their sales

Ads Cost(X) Sales(Y)

45 15
55 25
47 17
75 34
90 41
100 47
95 46
88 37
50 22
45 20

Solution:-
 Summary output

Regression Statistics
Multiple R 0.985401606
R Square 0.971016325
Adjusted R Square 0.967393366
Standard Error 2.180543935
Observations 10

ANOVA
df SS MS F Significance F
Regression 1 1274.362 1274.362 268.0174 1.95241E-07
Residual 8 38.03817 4.754772
Total 9 1312.4

Coefficien Standard P- Lower Upper Lower Upper

ts Error t Stat value 95% 95% 95.0% 95.0%
- - - - - -
Interce 5.652056 2.4493 0.0399 10.97337 0.330741 10.97337 0.330741
pt 555 2.307591 3 78 172 389 172 389
Ads 0.522493 16.371 1.95E- 0.448896 0.596090 0.448896 0.596090
Cost(X) 573 0.031915 24 07 677 47 677 47
 Geo 525 Quantitative Techniques in Geography

 Residual Output:-

Observation Predicted Sales(Y) Residuals

1 17.86015424 -2.860154242
2 23.08508997 1.914910026
3 18.90514139 -1.905141388
4 33.53496144 0.46503856
5 41.37236504 -0.372365039
6 46.59730077 0.402699229
7 43.9848329 2.015167095
8 40.32737789 -3.327377892
9 20.47262211 1.527377892
10 17.86015424 2.139845758

 Line fit plot:-

Ta bl e 10, Li ne Fi t Pl ot

X Variables- Line Fit Plot

50

45

40 y = 0.5225x - 5.6521
35 R² = 1
Sales (Y)

30 Sales(Y)

25 Predicted Y

20
Linear (Predicted Y)
15

10
40 50 60 70 80 90 100 110

Ads Cost (X)

Interpretation:-
 Multiple R: - this is the correlation coefficient that tell us how strong the linear
relationship is between two variables. The range of value can be any value between -1
and 1. If value is 1 it means a perfect positive relationship.-1 value show a strong
negative relationship 0 value means no relationship at all. In the above example,
 Geo 525 Quantitative Techniques in Geography

correlation coefficient value is 0.985401606 that is near to 1. It means there is positive

relationship between two variables.
 R square: - this value show us how good your model is. The range of this value is
between0 to 1. It can be convert into percentage. Zero value of R square mean a terrible
model and if value of R square is 1 it means a perfect model. In the above example, the
value of R square is 0.971016325 that is showing extremely well model. So we are
confident in our prediction. This value has given in percentage.
 Standard error: - The standard error is an absolute metric that displays how much the
data points deviate from the regression line on average. And it is the measure that tell us
accurate mean of any given sample. Larger value of standard error means inaccurate
representation and vice-versa. In the above example. Standard error value is 2.180543935
that is smaller number, it means accurate representation.
 Observations: - it show the total number of the observations in our model. In given
example, observation value is 10.
 Significance F: - the value of Significance F indicates how reliable our results are. Our
model is acceptable if Significance F is less than 0.05 (5 percent). If it's larger than 0.05,
we have to choose a different independent variable. In the above example, standard value
is 1.95241E-07. When we convert this value then the value of Significance f is
0.0195241, it means this value is less than 0.05 which means our results is
reliable.
 P- VALUE: - If the null hypothesis of your statistical test was true, the p-value informs
you how often you'd expect to see a test statistic as severe as or more extreme than the
one generated by your statistical test. In above example, we predicted some value, from
our data, the value of p is 0.039978 that is intercept, and it means 97% possibility is true
out of true. That is good sign for result. In ads cost, we possibility’ value is .0000195, that
is also under 0.05.

Exercise 7 Multiply regression:-

Simple linear regression is expanded into multiple regression. When we wish to forecast the
value of a variable based on the values of two or more other variables, we use this method. The
dependent variable is the variable we wish to forecast (or sometimes, the outcome, target or
criterion variable). The independent variables are the ones we use to predict the value of the
dependent variable.

For example, we might use multiple regression to see if exam success can be predicted based on
revision time, test anxiety, lecture attendance, and gender. You may also use multiple regression
to see if daily cigarette use can be predicted based on smoking duration, age at which you first
started smoking, smoker type, income, and gender.

We can also use multiple regression to assess the model's overall fit (variance explained) and the
proportional contribution of each predictor to the total variance explained. You could, for
example, wish to know how much variance in exam performance can be explained by revision
 Geo 525 Quantitative Techniques in Geography

time, test anxiety, lecture attendance, and gender "as a whole," as well as the "relative
contribution" of each factor.

Data:-
Y Sex Ratio (Female per thousand male) (Y).
X1 Growth rate of population, 2001-11( in Percentage) (X1)
X2 Levels of Literacy ( in percentage) (X2)
X3 Population Density per square kilo meter of area (X3)
X4 Female work participation rate ( in percentage) (X4)

S.N. District Y X1 X2 X3 X4
1 Indore 928 32.88 80.87 841 20.9
2 Jabalpur 929 14.51 81.07 473 25.3
3 Sagar 893 17.63 76.46 232 28.9
4 Bhopal 918 28.62 80.37 855 19.6
5 Rewa 931 19.86 71.62 375 32.9
6 Satna 926 19.19 72.26 297 29.9
7 Dhar 964 25.6 59 268 40.2
8 Chhindwara 964 13.07 71.16 177 36.6
9 Gwalior 864 24.5 76.65 446 14.5
10 Ujjain 955 16.12 72.34 326 33.8
11 Morena 840 23.44 71.03 394 16.8
12 West Nimar 965 22.85 62.7 233 40.9
13 Chhattarpur 883 19.51 63.74 203 32.7
14 Shivpuri 877 22.76 62.55 171 34.5
15 Bhind 837 19.21 75.26 382 8.4
16 Balaghat 1021 13.6 77.09 184 47
17 Betul 971 12.92 68.9 157 42.9
18 Dewas 942 19.53 69.35 223 38.4
19 Rajgarh 956 23.26 61.21 251 41.8
20 Shajapur 938 17.2 69.09 244 39.1
21 Vidisha 896 20.09 70.53 286 21.6
22 Ratlam 971 19.72 66.78 255 39
23 Tikamgarh 901 20.13 61.43 157 36.9
24 Barwani 982 27.57 49.08 242 41.9
25 Seoni 982 18.22 72.12 157 42.4
26 Mandsaur 963 13.24 71.78 199 42.9
27 Raisen 901 18.35 72.98 178 23.4
28 Sehore 918 21.54 70.06 261 28.7
29 East Nimar 943 21.5 66.39 178 38.6
30 Katni 952 21.41 71.98 173 31
 Geo 525 Quantitative Techniques in Geography

31 Damoh 910 16.63 69.73 185 34.3

32 Guna 912 26.97 63.23 208 30.3
33 Hoshangabad 914 14.49 75.29 232 22.4
34 Singrauli 920 28.05 60.41 213 34.6
35 Sidhi 957 23.72 64.43 172 32.2
36 Narsimhapur 920 14.01 75.69 182 29.5
37 Shahdol 974 17.39 66.67 285 38.1
38 Mandla 1008 17.97 66.87 142 49
39 Jhabua 990 30.7 43.3 285 48.9
40 Panna 905 18.67 64.79 142 32.4
41 Ashoknagar 904 22.66 66.42 271 20.8
42 Neemuch 954 13.77 70.8 221 42.9
43 Datia 873 18.46 72.63 200 26
44 Burhanpur 951 19.37 64.36 229 31.5
45 Anuppur 976 12.3 67.88 94 37.6
46 Alirajpur 1011 19.45 36.1 104 48.6
47 Dindori 1002 21.32 63.9 158 52.9
48 Sheopur 901 22.94 57.43 171 28.7
49 Umaria 950 24.96 65.89 158 36.3
50 Harda 935 20.25 72.5 171 26.9

Solution:-

Interpretation
 Geo 525 Quantitative Techniques in Geography

The intercept value and the equation for each variable in this regression will be the same, but the
slope of each independent variable with the dependable variable will change. Now we'll see how
well the four independent factors can predict the dependent variable under consideration. Among
the four independent variables given above, we must choose the best predictor (independent).
 Multiple R: - this is the correlation coefficient that tell us how strong the linear
relationship is between two variables. The range of value can be any value
between -1 and 1. If value is 1 it means a perfect positive relationship.-1 value
show a strong negative relationship 0 value means no relationship at all. In the
above example, correlation coefficient value is 0.866906 that is near to 1. It
means there is positive relationship between two variables.
 R square: - this value show us how good your model is. The range of this value
is between0 to 1. It can be convert into percentage. Zero value of R square mean a
terrible model and if value of R square is 1 it means a perfect model. In the above
example, the value of R square is 0.751526 that is showing extremely well model.
So we are confident in our prediction. This value has given in percentage.
 Standard error: - The standard error is an absolute metric that displays how
much the data points deviate from the regression line on average. And it is the
measure that tell us accurate mean of any given sample. Larger value of standard
error means inaccurate representation and vice-versa. In the above example.
Standard error value is 21.9373 that is smaller number, it means accurate
representation.
 Observations: - it show the total number of the observations in our model. In
given example, observation value is 50.
 Significance F: - the value of Significance F indicates how reliable our results
are. Our model is acceptable if Significance F is less than 0.05 (5 percent). If it's
larger than 0.05, we have to choose a different independent variable. In the above
example, standard value is 4.43E-13. When we convert this value then the value
of Significance f is 0.0195241, it means this value is less than 0.05 which means
our results is reliable.
 P- VALUE: - If the null hypothesis of your statistical test was true, the p-value informs
you how often you'd expect to see a test statistic as severe as or more extreme than the
one generated by your statistical test. In above example, we predicted some value, for
different independent data.

Exercise 8:- Gravity model

It gets its name from the fact that it employs a formula that is comparable to Newton's law of
gravity. It is the most common example of spatial interaction. This have been used in estimating
migration movements, establishing potentials. In the gravity model, two variables are used for
predict the volume of spatial interaction between two places, they can be cities, country or
region.
 Geo 525 Quantitative Techniques in Geography

(1) The total population of the places, and (2) the distance between them, or the time or expense
of travelling that distance.

There will be a positive relationship between flow volume and population size, when two place
have large populations, we expect that a large volume of migrants, but if places are separated by
a large distance, we expect a small volume of migrants.

Newton's law is used to estimate and calculate the relationship between objects. The gravity
model uses this same idea to predict the relationship between places. Instead of gravitational
pull, however, we're interested in the degree of interaction between cities, towns, or regions.
Newton's law of gravity predicts that bodies which are larger and closer will exert more force.
Our main variables are size and distance. In the gravity model of human geography, we can use
these same variables. Size is measured in population, and distance can be measured using any
metric. The idea in the gravity model is the same as in Newton's law. The larger and closer two
places are, the more influence they'll have on each other

Data:-
Population
Sl. No. Urban Centres (‘000)
1 Duliajan 17.017
2 Naharkatia 15.052
3 Namrup 19.74
4 Dibrugarh 120.127

Road distance (in kilometres) among the towns

Urban Centres 1 2 3 4
Duliajan 1 0 14.8 43 45
Naharkatia 2 14.8 0 29.6 51.5
Namrup 3 43 29.6 0 81
Dibrugarh 4 45 51.5 81 0

Method/formula:-
Tij = Pi*Pj / (dij )2
Pi= Represent the population size of origin place

Pj= Represent the population size of distinction place.

D2ij= is the square of the distance between origin and destination.

Solution:-
 Geo 525 Quantitative Techniques in Geography

According to rule: - multiply the two place’s population and divided by the square of the distance
of the two place. For example, if we apply the gravity model on the Dibrugarh and Duliajan, then
multiply 120.127*17.017 and divided by the 45*45, answer will be 1.009482.

Gravity Model
Duliajan and Dibrugarh 120.12*17.017/45*45 1.009482
Duliajan and Naharkatia 17.017*15.052/14.8*14.8 1.16937493
Duliajan and Namrup 17.017*19.74/43*43 0.18167419
Namrup and Naharkatia 19.74*15.052/29.6*29.6 0.33912354
Namrup and Dibrugarh 19.74*120.127/81*81 0.36142463

Dibrugarh and Naharkatia 120.127*15.052/51.8*51.8 0.67386876

Result: - There are the values of gravity model of different place:- 1.009482, 1.16937493,
0.18167419, 0.33912354, 0.36142463, 0.67386876

Interpretation:-
 Geo 525 Quantitative Techniques in Geography

In the above example we can shows that the cities are separated by a broad variety of distances,
and that the spatial interaction between two places is influenced not just by distance but also by
population size.

We predict a significant volume of migrants or commuters when two regions have large
populations, but we expect the impact of distance, as a mediator, when places are separated by a
considerable distance. In the other words, we can see here population size leads to a positive
relationship, distance leads an inverse correlation.

Exercise 9 Theissen polygon: -

This method are used to convert the point rainfall values which was measured by various rain
gauge stations into an average value over a catchment area. It is also called weighted mean
method because it is given a weightage according to the area closest to that station. Its best things
is that it is based on assumption that a rain gauge station best represent the area which is close to
it.

Based on an area containing at least two points, a Thiessen Polygon is a 2-dimensional shape
whose boundaries contain all space which is closer to a point within the area than any other point
without the area. Such polygons are a critical component of such spatial analysis concepts as
nearest neighbor and proximity assessment. Named after American meteorologist Alfred H.
Thiessen, Thiessen polygons are a more specific application of Voronoi diagram to meteorology
and geophysics

Thiessen polygons are generated from a set of sample points such that each polygon defines an
area of influence around its sample point, so that any location inside the polygon is closer to that
point than any of the other sample points.

Data:-
Geo 525 Quantitative Techniques in Geography
 Geo 525 Quantitative Techniques in Geography

Method/formula:-
𝑃1𝐴1+𝑃2𝐴2+⋯+𝑃𝑛𝐴𝑛 ∑𝑛
𝑖=1 𝑃𝑖𝐴𝑖 1
P= = ∑𝑛 𝐴𝑖
= ∑𝑁
𝑖=1 𝐴𝑖𝑃𝑖
𝐴1+𝐴2+⋯+𝐴𝑛 𝑖=1 𝐴
Pi= precipitation in the catchment area

Ai= area come under the polygon

Solution:-
𝟖.𝟖∗𝟓𝟕𝟎+𝟕.𝟔∗𝟗𝟐𝟎+𝟏𝟎.𝟖∗𝟕𝟐𝟎+𝟗.𝟐∗𝟔𝟐𝟎+𝟏𝟑.𝟖∗𝟓𝟐𝟎+𝟏𝟎.𝟒∗𝟓𝟓𝟎+𝟖.𝟓∗𝟒𝟎𝟎+𝟏𝟎.𝟓∗𝟔𝟓𝟎+𝟏𝟏.𝟐∗𝟓𝟎𝟎+𝟗.𝟓∗𝟑𝟓𝟎+𝟕.𝟖∗𝟓𝟐𝟎+𝟓.𝟐∗𝟐𝟓𝟎+𝟓.𝟔∗𝟑𝟓𝟎
𝟓𝟕𝟎+𝟗𝟐𝟎+𝟕𝟐𝟎+𝟔𝟐𝟎+𝟓𝟐𝟎+𝟓𝟓𝟎+𝟒𝟎𝟎+𝟔𝟓𝟎+𝟓𝟎𝟎+𝟑𝟓𝟎+𝟓𝟐𝟎+𝟐𝟓𝟎+𝟑𝟓𝟎

64,850
= 6920

=9.37138728cm Ans.

Results: - the value of Theissen polygon is 9.37138728cm.

Interpretation:-
As we all know, the Thiessen polygon is a technique for forecasting average rainfall depths in
watersheds using a rain-gauge network that is particularly well adapted to electronic
computation. In the example data table, the measured rainfall at 13 locations is already provided
(table number). Rainfall variability was calculated using the regions, which was then multiplied
by sub-area rainfall and added. The average annual precipitation is 9.37138728cm, according to
the facts and estimates. It is also clear that the annual precipitation at various places with
recorded mean rainfall is significantly less than the annual total mean rainfall.