The Applicability of Principal Component Analysis: A Review of The Nigeria Gross Domestic Products (GDP)
The Applicability of Principal Component Analysis: A Review of The Nigeria Gross Domestic Products (GDP)
Abstract:- In 2014, Nigeria was largest and fastest reached the peak in fourth quarter of 2014 with certain
growing economy in Africa, and this is determine by amount of about N24,205,863.34 million with a nominal
year on year Gross Domestic Product (GDP), which is GDP. Based on the Balance Trade by Kimberly 2018, the
the major tools use by the rest of the world to determine Nigerian GDP was worth 375.77 billion US dollars in 2017.
the capacity of a country. Furthermore National Bureau The GDP value of the Nigerian economy represents 0.61%
of Statistics (NBS) releases both quarterly and yearly of the world economy. GDP in Nigeria averaged 97.52 USD
growth of Nigeria GDP, while the data source in this Billion from 1960 until 2017. The Gross Domestic Product
project is from NBS, starting from 1980 to 2017. per capita of the economy was previously recorded as
Moreover the Nigeria GDP comprises many components. 2412.41 US dollars in 2017.
The major PCs use are Agriculture, Building and
Construction, Industry, Wholesale and Retail Trade, and Objectives of the Study
Services. It was observed that there is yearly growth in The objectives of this study are to determine the yearly
Nigeria gross domestic products due to increase in all growth of Nigeria Gross Domestics Products.
contributors factored components from 1980 to 2017,
Agriculture had the highest number activities to this per The Specific Objectives of this Study are:
capital income followed by Industry, Services, wholesale
and retail trade, and building and construction. Also the To examine the yearly growth pattern of the Nigeria
overall accumulated variance for the GDP for the GDP
studied period is 10.7 for a 56% coefficient of variation. To determine the contribution of each sector contributing
All the contributors (Agriculture, Industry,……) to national GDP for adequate policy making
maintained a positive correlation all through the period To predict yearly growth of Nigeria GDP
of study expect for Services that affect/shirk the GDP
along the year. Wholesales experienced the strongest Principal Component Analysis (PCA)
doling of taking care of the GDP with 98%, followed by This is a dimension tool that is used to reduce large
Building and Construction, industry and agriculture sets of variables to small sets and still retain most of the
with 86%, 78% and 6% respectively. However for information in the large sets. It is used to transforms
Nigeria to maintain the largest economy in Africa It will numbers of probably correlated variables into a smaller
strongly advice that the economy team should look number of uncorrelated variables known as principal
properly on agriculture and industry and if this two components.
increases certainly wholesale and retail trade with
building and construction will be affected positively Reason for Principal Component Analysis
while extended to services, therefore the GDP will also be The main reasons for the PCA includes
stabilise.
To reduces a larger number of variables to a smaller
I. INTRODUCTION number of factors
To choose a fraction of variables from a whole, based on
Background to the Study the original variables that have the highest correlations
Nigeria become the fastest growing economy in Africa with the principal component.
in 2012 by the Rebasing (GDP) released from the bulletin
published by the National Bureau of Statistics (NBS) which
make Nigeria to become the largest economy in Africa and
Fig 3 (A Plot showing a Zigzag Direction of Building and Construction over Time)
Fig 5 (A Plot showing a Zigzag Direction of Wholesale and Retail over Time)
From the histograms of the contributors, it is note that all the contributors are factored components due to their increasing
factor from 1981 to 2017 to the Nigeria GDP as it is shown below in fig 7 to 12.
Fig 9 (Histogram showing the Contribution of the Building and Construction over Time)
Fig 11 (Histogram showing the Contribution of wholesale and Retail Trade over Time)
Q–Q (Quantile-Quantile) Plot below, shows the Contribution Factors are Normal q-q Plot.
Fig 13 Q–Q (Quantile-Quantile) Plot shows the Contribution Factors are Normal q-q Plot
Fig 14 Q–Q (Quantile-Quantile) Plot shows the Contribution Factors are Normal q-q Plot
Fig 15 Q–Q (Quantile-Quantile) Plot shows the Contribution Factors are Normal q-q Plot
Fig 16 Q–Q (Quantile-Quantile) Plot shows the Contribution Factors are Normal q-q Plot
Fig 17 Q–Q (Quantile-Quantile) Plot shows the Contribution Factors are Normal q-q Plot
Agriculture had the highest number activities to this for SERVICES that affect/shirk the GDP along the year.
per capital income followed by Industry, Services, Wholesales experienced the strongest doling of taking care
wholesales and building. It is to be noted that factored series of the GDP with 98%, followed by BC, industry and agric
are more suitable PCA than that of time series models. This with 86%, 78% and 6% respectively, and -6% shirking by
factored certainty was justified by the decreasing variance of Services.
the GDP from 1981 to 2017. It connotes that the variation at
the beginning of 1981 that the GDP and its contributors are Correlation Matrix:
lower and far apart from the immediate past one, the
deviation in variation was larger. That is why all the , 𝐴𝑔𝑟𝑖𝑐𝑢𝑙𝑡𝑢𝑟𝑒 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝑏𝑢𝑖𝑙𝑑𝑖𝑛𝑔 𝑤𝑟𝑡 𝐺𝐷𝑃
histograms and QQ plots maintained and suggested a single
. . . . . .
increasing flow of trend expect for services that experienced 𝐴𝑔𝑟𝑖𝑐𝑢𝑙𝑡𝑢𝑟𝑒 1.00 0.64 0.490.83 0. .80 0.93
a structural break towards last three years. 𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 0.76 1.00 0.760.54 −0.92 0.01
𝐵𝑢𝑖𝑙𝑑𝑖𝑛𝑔/𝐶 0.89 0.86 1.000.62 0.73 0.48
Correlation Matrix and Descriptive Interpretation 𝑊𝑅𝑇 0.40 0.92 0.841.00 0.72 0.40
The overall accumulated variance for the GDP for the 𝑆𝑒𝑟𝑣 𝑖𝑐𝑒𝑠 0.88 0.39 0.770.79 1.00 −0.49
studied period is 10.7 for a 56% coefficient of variation. All 𝐺𝐷𝑃 (0.06 0.78 0.860.98 −0.24 1.00 )
the contributors (Agriculture, Industry) maintained a
positive correlation all through the period of study expect
Rotation (n x k) = (5 x 5):
The Linear Combination for the First Principal Also the overall accumulated variance for the GDP for
Component is: the studied period is 10.7 for a 56% coefficient of variation.
All the contributors (Agriculture, Industry) maintained a
positive correlation all through the period of study expect
for SERVICES that affect/shirk the GDP along the year.
Wholesales experienced the strongest doling of taking care
Prediction of the GDP with 98%, followed by Building and
Construction, industry and agriculture with 86%, 78% and
6% respectively.
RECOMMENDATION