Download to read offline
![International Journal of Research in Advent Technology, Vol.7, No.11, November 2019
E-ISSN: 2321-9637
Available online at www.ijrat.org
46
doi: 10.32622/ijrat.711201939
Abstract— This comparative study has been carried out to
discuss all investors are excited to know about the trend of
groundnut area, yield and production in Coastal Andhra and
Ceded Regions (Seemandhra). Now it has 13 districts during
2003-2004 to 2018 (15 years of data) for present comparative
analysis of groundnut crop production in Coastal Andhra and
Ceded were collected and presented. Based on results
collected some conclusions are made about the forecast
production of groundnut crop by using ARIMA with R-
software.
Index Terms— Area, yield, production, Groundnut,
ARIMA, R software.
I. INTRODUCTION
Time is one of the most important factors on which our
business and real-life dependent situations. But
now a day’s technology has helped us manage the time with
continuous innovations taking place in all aspects
of our lives. Time series modelling is a dynamic research area
which has attracted the attention of researcher’s
community over last few decades. This model is used to
generate future values for the series, i.e to make
forecasts. Forecasting lays a ground for reducing the risk in
all decision making because many of decisions
need to be made under uncertainty.
India is an agricultural country. The Economic
progress and standard of living of people directly or
indirectly is determined by agriculture. Indian economy
mainly depends on agriculture as majority of the population
depends on agricultural and allied activities for seeking out
their livelihood. Groundnut is grown throughout the tropics
and its cultivation is extended to the subtropical countries
lying between 45°
North and 35°
South and up to an altitude
of 1,000 meters. The total amount of rainfall required for
presuming operations (preparatory) is 100 mm, for sowing it
is 150 mm and for flowering and pod development an evenly
Manuscript revised on December 25, 2019 and published on January
10, 2019
Ananda Kumar Ginka, currently working as a Lecturer in Statistics in
department of statistics, Sri Srinivasa Degree college, Madanapalle,
Chittoor District, Andhra Pradesh, India.
Dhanunjaya Sunkara, M.Sc.,in Statistics and Pursuing Ph.D, Research
Scholar, Department Of Statistics, Sri Krishnadevaraya University,
Anantapuramu, Andhra Pradesh, India.
Mohammed Akhtar, working as a Professor and Head, Department of
Statistics, Sri Krishnadevaraya university, Anantapuramu, Andhra Pradesh,
India.
distributed rainfall of 400-500 mm is required, Madhusudana
,B et al( 2013)[1]. Coastal Andhra is located in the eastern
region of the state of Andhrapradesh on Coromandel Coast
and comprises nine districts. In Andhra Pradesh groundnut is
grown majorly in Srikakulam and Vishakapatnam districts of
Coastal Andhra Region.
Ceded Districts is name of an area in the Deccan,
India that was ‘Ceded’ to the British East India Company by
the Nizam in 1800. The name was the British Raj, even
though the denomination had no official weight for legal or
administrative purposes. Rayalaseema, meaning ‘rocky
region’, Especially, groundnut is the only important
commercial crop in the drought prone district of
Anantapuramu in Rayalaseema region of Andhra Pradesh.
So, the district headquarter of Anantapuramu is called as
‘Groundnut City’. Groundnut is an important protein
supplement for cattle and poultry rations. It is also consumed
as confectionery product. The cake can be used for
manufacturing artificial fibre. The haulms are fed to
livestock.
Crop area estimation and forecasting of crop yield
are an essential procedure in supporting policy decision
regarding land use allocation, food security and
environmental issues. Statistical techniques able to provide
crop forecast with reasonable precessions well in advanced.
Various approaches have been used for forecasting such
agricultural systems. Concentration have been given on the
uni-variate time series Auto Regressive Integrated Moving
Average (ARIMA) MODELS, which are primarily due to
World of Box and Jenkins (1970). Among the stochastic time
series models ARIMA types are powerful and popular as they
can successfully describe the observed data and can make
forecast with minimum forecast error. These types of models
are very difficult to identify and estimate.
II. LITERATURE REVIEW
Muhammad Iqbal Ch et al.(2016) for forecasting of
wheat production: A comparative study of Pakistan and India
[2], Similar studies have been done by Rachana et al. (2010)
for forecasting pigeon pea production in India by using
ARIMA Modelling [3], N.M.F. Rahman et al. ( 2010)for
forecasting of Boro rice production in Bangladesh [4],
Najeeb Iqbal et al. (2005) for forecasting wheat area and
production in Pakistan [5], M.K Debnath et al. (2013)for
forecasting Area, production, and Yield of Cotton in India
using ARIMA Model [6], M. Hemavathi et al.(2018)
ARIMA Model for Forecasting of Area, Production and
productivity of Rice and Its Growth Status in Thanjavur
District of TamilNadu, India[7], P.K. Sahu et al.(2015) for
modelling and forecasting of area, production, yield and total
Forecasting Area, Yield and Production of Groundnut
Crop: A Comparative Study of Coastal Andhra and Ceded
Regions Using- R
Ananda Kumar Ginka, Dhanunjaya Sunkara, Mohammed Akhtar](https://image.slidesharecdn.com/711201939-200228055916/75/711201939-1-2048.jpg)
![International Journal of Research in Advent Technology, Vol.7, No.11, November 2019
E-ISSN: 2321-9637
Available online at www.ijrat.org
47
doi: 10.32622/ijrat.711201939
seeds of Rice and Wheat in SAARC Countries and the World
towards Food Security[8], Mohammed Amir Hamjah et
al.(2014) for Rice Production Forecasting in Bangladesh: An
Application of Box-Jenkins ARIMA Model[9], Muhammad
et al(1992) conducted an empirical study of modelling and
forecasting time series data of rice production in Pakistan
[10], Niaz Md. Farhat Rahman et al.(2013), Modelling for
Growth and Forecasting of pulse production in Bangladesh
[11], Vishwajith K..P et al.(2014), Timeseries Modeling
and forecasting of pulses production in India[12], Ashwin
Darekar et al.(2017), Forecasting oilseeds prices in India:
Case of Groundnut [13] , Bhola Nath et al.(2018) DS ,
Forecasting Wheat production in India: An ARIMA
modelling approach [14] , Pant, D.C. and Pradeep Pal, et
al.(2004), Comparative Economics of Agro-processing units
for Groundnut in Southern Rajasthan [15], Ap Patel, G.N.,
and N.L. Agarwal et al. (1993), Price Behaviour of
Groundnut in Gujarat [16], Mohammad Mayazzem
Hossain(2017), Comparision of ARIMA and Neural Net
Work Model to forecast the jute Production in Bangladesh,
Jahingir Nagar University Journal of Science, [17] , also use
the ARIMA Model . The study is to identify the best ARIMA
model, which is for fitting and forecasting of Groundnut
Area, Yield, Production in Ceded region respectively.
Conclusions are drawn and found the forecasting for the
future. The R-Software is used to analyse and graphical
representation of the results.
R- software: R is a commonly used free Statistics software.
R allows you to carry out statistical analyses in an interactive
mode, as well as allowing simple programming. The R-
language is widely used among statistician and data miners
for developing statistical software and data analysis.
Although R has a command line interface, there are several
graphical user interfaces, such as R studio, an integrated
development environment. R is a programming language and
environment commonly used in statistical computing, data
analytics and scientific research. It is one of the most popular
languages used by statisticians, data analysts, researchers and
marketers to retrieve, clean, analyze, visualize and present
data.
III. MATERIALS AND METHODS
A. Data collection:
The study has utilized secondary source of data. The time
series data on yearly kharif and Rabi seasons totals area, yield
and production of groundnut crop from 2003-2004 to
2017-2018 of 15 years data required for the study was
collected from the DIRECTORATE OF ECONOMICS AND
STATISTICS, HYDERABAD. The 15 years of comparative
data of groundnut producing ceded districts viz.,
Anantapuramu, Kurnool, cuddapah, chittoor districts of
Andhra Pradesh and Coastal Andhra Andhra districts viz.,
Srikakulam, Vizianagaram, Visakhapatnam, East Godavari,
West Godavari,Krishna, Guntur, Prakasam, and Nellore
districts. Coastal Andhra borders Rayalaseema regions of the
state and the states of Telangana, Odisha. The presence of the
Krishna River Godavari River and Penna River makes the
area fertile for irrigation.
Fig:1 Area, Yield and Production of Groundnut Crop in
Ceded and Coastal Andhra Regions
B. Auto Regressive Integrated Moving Average (ARIMA)
model (Box-Jenkins model):
One of the most popular and frequently used stochastic
time series models is the Auto Regressive Integrated Moving
Average (ARIMA) model was introduced by Box and
Jenkins. The basic assumption made to implement this
model is that considered time series is linear and follows a
particular known statistical distribution, such as the Normal
Distribution. ARIMA model has subclasses of other models,
such as Auto Regressive (AR), Moving Average (MA) and
Auto Regressive Moving Average (ARMA) models. For
seasonal time series forecasting, Box and Jenkins had
proposed a quite successful variation of ARIMA model, viz.
the Seasonal ARIMA (SARIMA). The popularity of the
ARIMA model is mainly is due to its flexibility to represent
several varieties of time series with simplicity as well as the
associated Box-Jenkins(1994) [5] methodology for the
optimal model building process.
The term ARIMA stands for "Auto-Regressive Integrated
Moving Average." Lags of the differenced series appearing in
the forecasting equation are called "auto-regressive" terms,
lags of the forecast errors are called "moving average" terms,
and a time series which needs to be differenced to be made
stationary is said to be an "integrated" version of a stationary
series. Random-walk and random-trend models,
autoregressive models, and exponential smoothing models
(i.e., exponential weighted moving averages) are all special
cases of ARIMA models. A non-seasonal ARIMA model is
classified as an "ARIMA (p, d, q)" model, where p is the
number of autoregressive terms, d is the number of
non-seasonal differences, and q is the number of lagged
forecast errors in the prediction equation. The Box-Jenkins
methodology seeks to transform any time series data to be
stationary; and then apply the stationary process for](https://image.slidesharecdn.com/711201939-200228055916/75/711201939-2-2048.jpg)
![International Journal of Research in Advent Technology, Vol.7, No.11, November 2019
E-ISSN: 2321-9637
Available online at www.ijrat.org
48
doi: 10.32622/ijrat.711201939
forecasting by using past uni-variate time series process for
future forecast with a host of selection and diagnostic tools.
1) Model Identification
This stage involves the specification of the correct order of
ARIMA model by determining the appropriate order of the
AR, MA and the integrated parts or the differencing order.
The major tools in the identification process are the (sample)
autocorrelation function and partial autocorrelation function.
The identification approach is basically designed for both
stationary and non-stationary processes. Spikes represent in
the line at various lags in the plot with length equal to
magnitude of autocorrelations and these spikes distinguish
the identification of a stationary and non-stationary process.
The main objective in fitting ARIMA model is to identify the
stochastic process of the time series and its stationarity
counterpart. The main objective in fitting ARIMA models is
to identify the stochastic process of the time series and
predict the future values accurately. Ansari and Ahmad
(2001)[18] worked with application of ARIMA modelling
and co-integration analysis on time series of tea price.
Different stages in forecasting model are given below.
Identification: A good starting point for time series analysis is
a graphical plot of the data. It helps to identify the presence
of trends. Before estimating the parameters p and q of the
model, the data are not examined to decide about the model
which best explains the data. This is done by examining the
sample ACF, and PACF. Both ACF and PACF are used as
the aid in the identification of appropriate models. There are
several ways of determining the order type of process, but
still there was no exact procedure for identifying the model.
2) Model Estimating the parameters
After tentatively identifying the suitable model is not
“estimating a second time series”, it is filtering it. The
function accuracy gives multiple measures of accuracy of the
model fit, ME(mean error), RMSE(root mean squared error),
MAE(mean absolute error), MPE(mean percentage error),
MAPE(mean absolute percentage error), MASE(mean
absolute scaled error) , And ACF (auto correlation function)
It is up to you to decide, based on the accuracy measures,
whether you consider this a good fit or not. For example,
mean percentage error of nearly -70% does not look good to
me in general, but that may depend on what your series are
and how much predictability you may realistically expect. It
is often a good idea to plot the original series and the fitted
values, and also model residuals. You may occasionally learn
more from the plot than from the few summarizing measures
such as the ones given by the accuracy` function. Depending
on the ACF and PACF of these sequence plots a model is run
with appropriate software (R-Software). The best fitting
model must also have few parameters as much as possible
alongside best statistics of the model according to the
information selection criteria.
3) Model Diagnostic Checking
After having estimated the parameters of a tentatively
identify ARIMA model, it is necessary to do diagnostic
checking to verify that the model is adequate. Examining
ACF And PACF considered random when all their ACF and
PACF considered random when all their ACF were within the
limits. Model checking in time series can be done by looking
at the residuals. Traditionally the residuals given by
Residuals = observed values – fitted values. These results
should be normally distributed with zero mean, uncorrelated,
and should have minimum variance or dispersion, if indeed a
model fits the well. That is model validation usually consist
of plotting residuals overtime to verify the validation.
4) Model Forecasting
After satisfying about the adequacy of the fitted model, it can
be used for forecasting future values. This was done with the
help of R- Software.
IV. RESULT AND DISCUSSIONS
Analysis of Time series data regarding agricultural
oriented groundnut crop area, yield and production using R
software tabulated along with necessary graphical
presentations mentioned below, Groundnut is an important
protein supplement for cattle and poultry rations. It is also
consumed as confectionery product. The cake can be used
for manufacturing artificial fibre. The haulms are fed to live
stock. Groundnut shell is used as fuel for manufacturing
coarse boards. Cork substitutes. Groundnut is also valued as
a rotation crop. Being a legume with root nodules, it can
synthesize with atmospheric nitrogen and thereby improve
soil fertility. All investors are timely cautious about crop
production with updated technology.
Table-1 Area, Yield and Production Of Groundnut Crop In Ceded And Coastal Andhra Regions
CEDED REGION COASTAL ANDHRA REGION
YEAR
Area
(in
000'ha.)
Yield
(in
Kg/ha.)
Prod.
(in
000'tones)
Area (in
000'ha.)
Yield
(in
Kg/ha.)
Prod.
(in
000'tones)
1 2003-2004 1164 2664 603 134 12122 159
2 2004-2005 1511 3455 1267 136 16249 175
3 2005-2006 1554 2800 924 123 18489 186
4 2006-2007 1041 2522 366 114 17059 157
5 2007-2008 1474 6278 2076 115 16221 183
6 2008-2009 1447 1978 471 118 17290 188](https://image.slidesharecdn.com/711201939-200228055916/75/711201939-3-2048.jpg)
![International Journal of Research in Advent Technology, Vol.7, No.11, November 2019
E-ISSN: 2321-9637
Available online at www.ijrat.org
56
doi: 10.32622/ijrat.711201939
(1,2,1) AR1 -0.1537 0.2740 160.7 -52.93 111.86
MA1 -1.0000 0.2183
(1,0,1) AR1 0.1425 0.2586 146.7956 328.6 -66.27 140.54
MA1 1.0000 02214 10.4633
In the present study, the ARIMA (2,1,1)in Ceded
Region & ARIMA(1,2,1)in Coastal Andhra Region were the
best fitted model through the minimum value of AIC, then
used for prediction up to 10 years of the production of
groundnut in ceded districts using 15 years time series data
i.e. from 2003-2004 to 2017-2018. ARIMA (2,1,1) in
Ceded Region & ARIMA(1,2,1)in Coastal Andhra Region
were used because the reason of its capability to make
prediction using time series data with any kind of patterns
and with auto correlated successive values of the time series.
The study was also validated and statistically tested that the
successive residuals in the fitted ARIMA (2,1,1) in Ceded
Region & ARIMA (1,2,1)in Coastal Andhra Region were not
correlated, and the residuals appear to be normally distributed
with the mean zero and constant variance. Hence, it can be a
satisfactory predictive model for the groundnut yield in ceded
districts in Andhra Pradesh for the period of 2018 to 2027.
The ARIMA (2, 1, 1) in Ceded Region & ARIMA(1,2,1)in
Coastal Andhra Region models projected an increment in the
production for the duration of 2018 to 2027. The prediction
of 2027 is resulted approximately 720.1359’000
tonnes(Ceded Region) & 163.78927’000 tonnes(Coastal
Andhra Region). Like any other predictive models for
forecasting, ARIMA model has also limitations on accuracy
of the predictions yet it is widely used for forecasting the
future values for time series. It is noticed that in Groundnut
production Ceded Region is better than Coastal Andhra
Region. The empirical Forecasting area, yield and
production of groundnut crop: a comparative study of
coastal andhra and ceded regions using- R findings of
study could help to forecast any such commodities. The
researchers and policy makers will thus get access for making
further extensive research work. We firmly believe that this
research has shed some important light on the subject area
encompassing time series forecasts of selected agricultural
crops in Seemandhra. These empirical findings can be an
important source of information to many researchers and
policy formulators as far as agricultural crops in Seemandhra
are concerned.
ACKNOWLEDGMENT
I would like to express my heartfelt gratitude to my guide and
Head of Dept. Prof. Mohammed Akhtar Palagiri who acted as
a source of inspiration in all spheres of my work for giving
valuable guidelines for completeing this course.
REFERENCES
[1] Madhusudana, B, “A survey on area, production and productivity of
groundnut crop in India”, IOSR Journal of Economics and Finance,
volume 1, PP 01-07, (2013).
[2] Muhammad Iqbal Ch, “Forecasting of wheat production: A
comparative study of Pakistan And India”, IJAR 4(12), 698-709 ,
(2016).
[3] Rachana Wankhade, “Use of the ARIMA Forecasting pigeon Pea
production in India “, International Review of Business and Finance,
Volume 2, Number 1, pp 97-102, ( 2010).
[4] N.M.F. Rahman , “ Forecasting of Boro rice production in
Baangladesh: An Approach”. J.Bangladesh Agril. Univ. ;
8(1):103-112p, (2010).
[5] Najeeb Iqbal , “Forecasting wheat area and production in Pakistan”,
Journal of Agriculture & Sciences, Volume.1, No.2, (2005).
[6] M.K Debnath , ” Forecasting Area, production, and Yield of Cotton in
India using ARIMA Model”, Journal of Space Science & Technology,
Volume 2, Issue 1. (2013).
[7] M. Hemavathi , “ ARIMA Model for Forecasting of Area, Production
and productivity of Rice and Its Growth Status in Thanjavur District of
TamilNadu”, India, International of Current Microbiology and
Applied Sciences, 7(2):149-156, (2018).
[8] P.K. Sahu, “Modelling and forecasting of area, production, yield and
total seeds of Rice and Wheat in SAARC Countries and the World
towards Food Security”, SCIEP, American Journal of Applied
Mathematics and Statistics, Vol.3, No.1,34-38., (2015).
[9] Mohammed Amir Hamjah, “ Rice Production Forecasting in
Bangladesh: An Application of Box-Jenkins ARIMA Model”,
Mathematical Theory and Modelling , Vol. 4, No 4, (2014).
[10] Muhammad, F., M., Siddiqqe, M. Bashir, S. Ahmad , “ Forecasting
Rice Production in Pakistan-Using ARIMA Models “, J. Of Animal
Plant Sci.; 2:27-3lp. (1992).
[11] Niaz Md. Farhat Rahman , “ Modeling for Growth and Forecasting of
pulse production in Bangladesh”, Research journal of Applied
Sciences, Engineering and Technology 5 (24):5578-5587. (2013).
[12] Vishwajith K..P, “Timeseries Modeling and forecasting of pulses
production in India”, Journal crop and weed, 10 (2):147-154, (2014).
[13] Ashwin Darekar, “Forecasting oilseeds prices in India: Case of
Groundnut”, J.Oilseeds Res..,34(4):235- 240, (2017).
[14] Bhola Nath, DS , “Forecasting Wheat production in India: An ARIMA
modelling approach”, Journal of Pharmacognosy and
Phytochemistry, 8(1):2158-2165, (2018).
[15] Pant, D.C. and Pradeep Pal, , “Comparative Economics of
Agro-processing units for Groundnut in Southern Rajasthan”, Indian
Journal of Agricultural Marketing, Vol, 18, No.1, January-(2004).
[16] Ap Patel, G.N., and N.L. Agarwal, ‘’Price Behaviour of Groundnut in
Gujarat”, Indian Journal of Marketing, Vol.7, No.2, pp.50-57, ( 1993).
[17] Mohammad Mayazzem Hossain, , “ Comparision of ARIMA and
Neural Net Work Model to forecast the jute Production in
Bangladesh”, Jahingir Nagar University Journal of Science, Vol.40,
pp11-18, (2017).
[18] Ansari M I and Ahmed SM,” Time series analysis of tea prices: an
application of ARIMA modelling and co- integration analysis”,
IJE,48; 49-54. (2001).
[19] Croxton.F.E., Cowden, D.J. and Klein, S, ”Applied General
Statistics”, New Delhi: Preentice Hall of India PVT. Ltd. (1979).
[20] Pankratz, A “Forecasting with Univariate Box – Jenkins models:
concepts and cases,” John Wiley, New York. ,( 1983).
[21] Box, G.E.P., Jenkins, G.M. and Reinsel, G.C “ Time series analysis.
Forecasting and control,” Pearson Education, Delhi. , ( 1994).
[22] Makridakis, S., Wheelwright, S.C. and Hyndman, R.J, “Forecasting
Methods and Applications”, 3rd Edition, John Wiley, New York. ,(
1998)](https://image.slidesharecdn.com/711201939-200228055916/75/711201939-11-2048.jpg)
More Related Content
711201939
- 1. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 46 doi: 10.32622/ijrat.711201939 Abstract— This comparative study has been carried out to discuss all investors are excited to know about the trend of groundnut area, yield and production in Coastal Andhra and Ceded Regions (Seemandhra). Now it has 13 districts during 2003-2004 to 2018 (15 years of data) for present comparative analysis of groundnut crop production in Coastal Andhra and Ceded were collected and presented. Based on results collected some conclusions are made about the forecast production of groundnut crop by using ARIMA with R- software. Index Terms— Area, yield, production, Groundnut, ARIMA, R software. I. INTRODUCTION Time is one of the most important factors on which our business and real-life dependent situations. But now a day’s technology has helped us manage the time with continuous innovations taking place in all aspects of our lives. Time series modelling is a dynamic research area which has attracted the attention of researcher’s community over last few decades. This model is used to generate future values for the series, i.e to make forecasts. Forecasting lays a ground for reducing the risk in all decision making because many of decisions need to be made under uncertainty. India is an agricultural country. The Economic progress and standard of living of people directly or indirectly is determined by agriculture. Indian economy mainly depends on agriculture as majority of the population depends on agricultural and allied activities for seeking out their livelihood. Groundnut is grown throughout the tropics and its cultivation is extended to the subtropical countries lying between 45° North and 35° South and up to an altitude of 1,000 meters. The total amount of rainfall required for presuming operations (preparatory) is 100 mm, for sowing it is 150 mm and for flowering and pod development an evenly Manuscript revised on December 25, 2019 and published on January 10, 2019 Ananda Kumar Ginka, currently working as a Lecturer in Statistics in department of statistics, Sri Srinivasa Degree college, Madanapalle, Chittoor District, Andhra Pradesh, India. Dhanunjaya Sunkara, M.Sc.,in Statistics and Pursuing Ph.D, Research Scholar, Department Of Statistics, Sri Krishnadevaraya University, Anantapuramu, Andhra Pradesh, India. Mohammed Akhtar, working as a Professor and Head, Department of Statistics, Sri Krishnadevaraya university, Anantapuramu, Andhra Pradesh, India. distributed rainfall of 400-500 mm is required, Madhusudana ,B et al( 2013)[1]. Coastal Andhra is located in the eastern region of the state of Andhrapradesh on Coromandel Coast and comprises nine districts. In Andhra Pradesh groundnut is grown majorly in Srikakulam and Vishakapatnam districts of Coastal Andhra Region. Ceded Districts is name of an area in the Deccan, India that was ‘Ceded’ to the British East India Company by the Nizam in 1800. The name was the British Raj, even though the denomination had no official weight for legal or administrative purposes. Rayalaseema, meaning ‘rocky region’, Especially, groundnut is the only important commercial crop in the drought prone district of Anantapuramu in Rayalaseema region of Andhra Pradesh. So, the district headquarter of Anantapuramu is called as ‘Groundnut City’. Groundnut is an important protein supplement for cattle and poultry rations. It is also consumed as confectionery product. The cake can be used for manufacturing artificial fibre. The haulms are fed to livestock. Crop area estimation and forecasting of crop yield are an essential procedure in supporting policy decision regarding land use allocation, food security and environmental issues. Statistical techniques able to provide crop forecast with reasonable precessions well in advanced. Various approaches have been used for forecasting such agricultural systems. Concentration have been given on the uni-variate time series Auto Regressive Integrated Moving Average (ARIMA) MODELS, which are primarily due to World of Box and Jenkins (1970). Among the stochastic time series models ARIMA types are powerful and popular as they can successfully describe the observed data and can make forecast with minimum forecast error. These types of models are very difficult to identify and estimate. II. LITERATURE REVIEW Muhammad Iqbal Ch et al.(2016) for forecasting of wheat production: A comparative study of Pakistan and India [2], Similar studies have been done by Rachana et al. (2010) for forecasting pigeon pea production in India by using ARIMA Modelling [3], N.M.F. Rahman et al. ( 2010)for forecasting of Boro rice production in Bangladesh [4], Najeeb Iqbal et al. (2005) for forecasting wheat area and production in Pakistan [5], M.K Debnath et al. (2013)for forecasting Area, production, and Yield of Cotton in India using ARIMA Model [6], M. Hemavathi et al.(2018) ARIMA Model for Forecasting of Area, Production and productivity of Rice and Its Growth Status in Thanjavur District of TamilNadu, India[7], P.K. Sahu et al.(2015) for modelling and forecasting of area, production, yield and total Forecasting Area, Yield and Production of Groundnut Crop: A Comparative Study of Coastal Andhra and Ceded Regions Using- R Ananda Kumar Ginka, Dhanunjaya Sunkara, Mohammed Akhtar
- 2. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 47 doi: 10.32622/ijrat.711201939 seeds of Rice and Wheat in SAARC Countries and the World towards Food Security[8], Mohammed Amir Hamjah et al.(2014) for Rice Production Forecasting in Bangladesh: An Application of Box-Jenkins ARIMA Model[9], Muhammad et al(1992) conducted an empirical study of modelling and forecasting time series data of rice production in Pakistan [10], Niaz Md. Farhat Rahman et al.(2013), Modelling for Growth and Forecasting of pulse production in Bangladesh [11], Vishwajith K..P et al.(2014), Timeseries Modeling and forecasting of pulses production in India[12], Ashwin Darekar et al.(2017), Forecasting oilseeds prices in India: Case of Groundnut [13] , Bhola Nath et al.(2018) DS , Forecasting Wheat production in India: An ARIMA modelling approach [14] , Pant, D.C. and Pradeep Pal, et al.(2004), Comparative Economics of Agro-processing units for Groundnut in Southern Rajasthan [15], Ap Patel, G.N., and N.L. Agarwal et al. (1993), Price Behaviour of Groundnut in Gujarat [16], Mohammad Mayazzem Hossain(2017), Comparision of ARIMA and Neural Net Work Model to forecast the jute Production in Bangladesh, Jahingir Nagar University Journal of Science, [17] , also use the ARIMA Model . The study is to identify the best ARIMA model, which is for fitting and forecasting of Groundnut Area, Yield, Production in Ceded region respectively. Conclusions are drawn and found the forecasting for the future. The R-Software is used to analyse and graphical representation of the results. R- software: R is a commonly used free Statistics software. R allows you to carry out statistical analyses in an interactive mode, as well as allowing simple programming. The R- language is widely used among statistician and data miners for developing statistical software and data analysis. Although R has a command line interface, there are several graphical user interfaces, such as R studio, an integrated development environment. R is a programming language and environment commonly used in statistical computing, data analytics and scientific research. It is one of the most popular languages used by statisticians, data analysts, researchers and marketers to retrieve, clean, analyze, visualize and present data. III. MATERIALS AND METHODS A. Data collection: The study has utilized secondary source of data. The time series data on yearly kharif and Rabi seasons totals area, yield and production of groundnut crop from 2003-2004 to 2017-2018 of 15 years data required for the study was collected from the DIRECTORATE OF ECONOMICS AND STATISTICS, HYDERABAD. The 15 years of comparative data of groundnut producing ceded districts viz., Anantapuramu, Kurnool, cuddapah, chittoor districts of Andhra Pradesh and Coastal Andhra Andhra districts viz., Srikakulam, Vizianagaram, Visakhapatnam, East Godavari, West Godavari,Krishna, Guntur, Prakasam, and Nellore districts. Coastal Andhra borders Rayalaseema regions of the state and the states of Telangana, Odisha. The presence of the Krishna River Godavari River and Penna River makes the area fertile for irrigation. Fig:1 Area, Yield and Production of Groundnut Crop in Ceded and Coastal Andhra Regions B. Auto Regressive Integrated Moving Average (ARIMA) model (Box-Jenkins model): One of the most popular and frequently used stochastic time series models is the Auto Regressive Integrated Moving Average (ARIMA) model was introduced by Box and Jenkins. The basic assumption made to implement this model is that considered time series is linear and follows a particular known statistical distribution, such as the Normal Distribution. ARIMA model has subclasses of other models, such as Auto Regressive (AR), Moving Average (MA) and Auto Regressive Moving Average (ARMA) models. For seasonal time series forecasting, Box and Jenkins had proposed a quite successful variation of ARIMA model, viz. the Seasonal ARIMA (SARIMA). The popularity of the ARIMA model is mainly is due to its flexibility to represent several varieties of time series with simplicity as well as the associated Box-Jenkins(1994) [5] methodology for the optimal model building process. The term ARIMA stands for "Auto-Regressive Integrated Moving Average." Lags of the differenced series appearing in the forecasting equation are called "auto-regressive" terms, lags of the forecast errors are called "moving average" terms, and a time series which needs to be differenced to be made stationary is said to be an "integrated" version of a stationary series. Random-walk and random-trend models, autoregressive models, and exponential smoothing models (i.e., exponential weighted moving averages) are all special cases of ARIMA models. A non-seasonal ARIMA model is classified as an "ARIMA (p, d, q)" model, where p is the number of autoregressive terms, d is the number of non-seasonal differences, and q is the number of lagged forecast errors in the prediction equation. The Box-Jenkins methodology seeks to transform any time series data to be stationary; and then apply the stationary process for
- 3. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 48 doi: 10.32622/ijrat.711201939 forecasting by using past uni-variate time series process for future forecast with a host of selection and diagnostic tools. 1) Model Identification This stage involves the specification of the correct order of ARIMA model by determining the appropriate order of the AR, MA and the integrated parts or the differencing order. The major tools in the identification process are the (sample) autocorrelation function and partial autocorrelation function. The identification approach is basically designed for both stationary and non-stationary processes. Spikes represent in the line at various lags in the plot with length equal to magnitude of autocorrelations and these spikes distinguish the identification of a stationary and non-stationary process. The main objective in fitting ARIMA model is to identify the stochastic process of the time series and its stationarity counterpart. The main objective in fitting ARIMA models is to identify the stochastic process of the time series and predict the future values accurately. Ansari and Ahmad (2001)[18] worked with application of ARIMA modelling and co-integration analysis on time series of tea price. Different stages in forecasting model are given below. Identification: A good starting point for time series analysis is a graphical plot of the data. It helps to identify the presence of trends. Before estimating the parameters p and q of the model, the data are not examined to decide about the model which best explains the data. This is done by examining the sample ACF, and PACF. Both ACF and PACF are used as the aid in the identification of appropriate models. There are several ways of determining the order type of process, but still there was no exact procedure for identifying the model. 2) Model Estimating the parameters After tentatively identifying the suitable model is not “estimating a second time series”, it is filtering it. The function accuracy gives multiple measures of accuracy of the model fit, ME(mean error), RMSE(root mean squared error), MAE(mean absolute error), MPE(mean percentage error), MAPE(mean absolute percentage error), MASE(mean absolute scaled error) , And ACF (auto correlation function) It is up to you to decide, based on the accuracy measures, whether you consider this a good fit or not. For example, mean percentage error of nearly -70% does not look good to me in general, but that may depend on what your series are and how much predictability you may realistically expect. It is often a good idea to plot the original series and the fitted values, and also model residuals. You may occasionally learn more from the plot than from the few summarizing measures such as the ones given by the accuracy` function. Depending on the ACF and PACF of these sequence plots a model is run with appropriate software (R-Software). The best fitting model must also have few parameters as much as possible alongside best statistics of the model according to the information selection criteria. 3) Model Diagnostic Checking After having estimated the parameters of a tentatively identify ARIMA model, it is necessary to do diagnostic checking to verify that the model is adequate. Examining ACF And PACF considered random when all their ACF and PACF considered random when all their ACF were within the limits. Model checking in time series can be done by looking at the residuals. Traditionally the residuals given by Residuals = observed values – fitted values. These results should be normally distributed with zero mean, uncorrelated, and should have minimum variance or dispersion, if indeed a model fits the well. That is model validation usually consist of plotting residuals overtime to verify the validation. 4) Model Forecasting After satisfying about the adequacy of the fitted model, it can be used for forecasting future values. This was done with the help of R- Software. IV. RESULT AND DISCUSSIONS Analysis of Time series data regarding agricultural oriented groundnut crop area, yield and production using R software tabulated along with necessary graphical presentations mentioned below, Groundnut is an important protein supplement for cattle and poultry rations. It is also consumed as confectionery product. The cake can be used for manufacturing artificial fibre. The haulms are fed to live stock. Groundnut shell is used as fuel for manufacturing coarse boards. Cork substitutes. Groundnut is also valued as a rotation crop. Being a legume with root nodules, it can synthesize with atmospheric nitrogen and thereby improve soil fertility. All investors are timely cautious about crop production with updated technology. Table-1 Area, Yield and Production Of Groundnut Crop In Ceded And Coastal Andhra Regions CEDED REGION COASTAL ANDHRA REGION YEAR Area (in 000'ha.) Yield (in Kg/ha.) Prod. (in 000'tones) Area (in 000'ha.) Yield (in Kg/ha.) Prod. (in 000'tones) 1 2003-2004 1164 2664 603 134 12122 159 2 2004-2005 1511 3455 1267 136 16249 175 3 2005-2006 1554 2800 924 123 18489 186 4 2006-2007 1041 2522 366 114 17059 157 5 2007-2008 1474 6278 2076 115 16221 183 6 2008-2009 1447 1978 471 118 17290 188
- 4. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 49 doi: 10.32622/ijrat.711201939 7 2009-2010 991 2588 493 96 18316 154 8 2010-2011 136 3466 962 91 16854 143 9 2011-2012 1058 2681 444 77 18098 138 10 2012-2013 1089 2899 635 70 18794 143 11 2013-2014 1111 3933 739 42 23234 106 12 2014-2015 832 2774 391 42 23278 102 13 2015-2016 732 4785 694 66 20216 142 14 2016-2017 968 3112 485 46 23764 117 15 2017-2018 697 6243 942 37 24519 106 TOTAL 15805 52178 11492 1307 284503 2199 Fig:2 Area, yield, Production of Ceded Region Fig:3 Area, yield, Production of Coastal andhra Region ACF, PACF plots are analysed to check stationarity of data upto15 ( 0 to 14) lags as shown below: Fig:4 AREA- ACF(CEDED REGION) Fig:5 AREA- PACF(CEDED REGION Fig:6 Yield- ACF(CEDED REGION) Fig: 7 YIELD - PACF(CEDED REGION) Fig :8 Production -ACF(CEDED REGION) Fig:9 PRODUCTION- PACF(CEDED REGION) Fig:10 Area-ACF(Coastal Region) Fig:11Area-PACF(Coastal Region) Fig:12 YIELD-ACF(Coastal Region) Fig:13 YIELD-ACF(Coastal Region) Fig:14 prod-ACF(Coastal Region) Fig:15 prod-ACF(Coastal Region) Table-2 Area, Yield And Production ACF And PACF(CEDED REGION & COASTAL ANDHRA REGION) Area, Yield and Production ACF and PACF(CEDED REGION) Area, Yield and Production ACF and PACF(COASTAL REGION) Lag ACF (area) PACF (area) ACF (YIELD) PACF (YIELD) ACF (PROD) PACF (PROD) ACF (AREA) PACF (AREA) ACF (YIELD) PACF (YIELD) ACF (PRO) PACF (PROD.) 0 1.000 0 1.000 0 1.000 0 1.000 0 1.000 0 1.000 0 1 0.299 0.299 -0.306 -0.306 -0.348 -0.348 0.784 0.784 0.546 0.546 0.634 0.634
- 5. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 50 doi: 10.32622/ijrat.711201939 Table-3 AREA, YIELD, AND PRODUCTTION POINT FORECAST (CEDED REGION) Area Point Forecast of Groundnut (CEDED REGION) Fig: 16 Area forecast Year Area Point forecast Lo 80 Hi 80 Lo95 Hi 95 2018 663.6431 74.34483 1252.941 -237.6108 1564.897 2019 630.2862 -230.43900 1491.011 -686.0794 1946.652 2020 596.9294 -489.68298 1683.542 -1064.9008 2258.760 2021 563.5725 -727.51535 1854.660 -1410.9760 2538.121 2022 530.2156 -952.81919 2013.250 -1737.8904 2798.322 2023 496.8587 -1169.92849 2163.646 -2052.2723 3045.990 2024 463.5018 -1381.29353 2308.297 -2357.8692 3284.873 2025 430.1449 -1588.43457 2448.724 -2657.0060 3517.296 2026 396.7881 -1792.35779 2585.934 -2951.2216 3744.798 2027 363.4312 -1993.76182 2720.624 -3241.5844 3968.447 Yield Point Forecast (CEDED REGION) Fig:17 Yield forecast Year Yield point forecast Lo 80 Hi 80 Lo95 Hi 95 2018 4338.339 2588.741 6087.937 1662.5598 1662.5598 2019 5078.286 3321.278 6835.294 2391.1738 7765.398 2020 5563.951 3536.040 7591.862 2462.5287 8665.374 2021 5098.539 2814.759 7382.319 1605.7985 8591.279 2022 5657.126 3256.581 8057.670 1985.8098 9328.442 2023 5666.277 3015.238 8317.315 1611.8634 9720.690 2024 5738.279 2906.801 8569.757 1407.9071 10068.650 2025 5998.293 2982.656 9013.931 1386.2738 10610.313 2026 6050.198 2823.687 9276.709 1115.6751 10984.720 2027 6216.148 2804.532 9627.765 998.5315 11433.765 Production Point Forecast (CEDED REGION) Fig:18 Production forecast Year Production Point forecast Lo 80 Hi 80 Lo95 Hi 95 2018 670.0918 -116.26186 1456.446 -532.5324 1872.716 2019 818.4277 -56.75061 1693.606 -520.0420 2156.898 2020 724.4765 -376.64553 1825.599 -959.5443 2408.497 2021 770.2133 -461.00217 2001.429 -1112.7683 2653.195 2022 735.4144 -663.80745 2134.636 -1404.5108 2875.340 2023 747.0475 -792.33296 2286.428 -1607.2318 3101.327 2024 731.9108 -957.86174 2421.683 -1852.3733 3316.195 2025 732.2079 1098.65717 2563.073 -2067.8586 3532.274 2026 723.6068 -1250.72629 2697.940 -2295.8751 3743.089 2027 720.1359 1394.13130 2834.403 -2513.3568 3953.629 2 -0.030 -0.131 0.119 0.028 -0.037 -0.180 0.578 -0.097 0.287 -0.016 0.410 0.014 3 0.006 0.062 0.081 0.138 0.449 0.437 0.477 0.148 0.369 0.311 0.493 0.381 4 0.205 0.202 -0.123 -0.073 -0. 318 -0.025 0.287 -0.314 0.242 -0.126 .256 -0.390 5 -0.057 -0.217 -0.041 -0.135 0.023 -0.104 0.038 0.248 -0.008 -0.185 -0.103 -0.355 6 -0.201 -0.100 0.033 -0.008 0.046 -0.240 -0.154 -0.175 -0.143 -0.208 -0.274 -0.437 7 0.016 0.140 -0.152 -0.115 -0.204 -0.137 -0.263 0.008 -0.120 0.007 -0.228 0.280 8 -0.051 -0.221 0.115 0.046 0.027 -0.054 -0.368 -0.077 -0.114 0.045 -0.331 0.078 9 -0.130 -0.002 -0.228 -0.206 -0.161 -0.191 -0.450 0.038 -0.225 -0.077 -0.480 0.108 10 -0.209 -0.104 0.284 0.208 0.054 0.047 -0.432 0.045 -0.396 -0.316 -0.311 -0.014 11 -0.108 -0.145 -0.077 0.061 -0.033 -0.014 -0.340 0.099 -0.275 0.048 -0.189 -0.187 12 -0.130 -0.048 -0.112 -0.181 -0.037 0.079 -0.274 -0.044 -0.146 0.042 -0.208 -0.104 13 -0.089 0.001 0.010 -0.185 0.049 -0.118 -0.250 -0.185 -0.288 -0.107 -0.127 -0.066 14 -0.020 -0.030 -0.094 -0.113 -0.010 -0.060 -0.134 0.068 -0.228 0.067 -0.042 -0.145
- 6. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 51 doi: 10.32622/ijrat.711201939 Table -4 AREA, YIELD, AND PRODUCTTION POINT FORECAST(COASTAL ANDHRA REGION) Area Point Forecast of Groundnut (COASTAL REGION) Fig: 19 Area forecast Year Area Point forecast Lo 80 Hi 80 Lo95 Hi 95 2018 30.313146 13.4849095 47.14138 47.14138 56.04971 2019 23.270686 0.4993657 46.04201 -11.555043 58.09642 2020 16.282895 -11.9052322 44.47102 44.47102 59.39292 2021 9.286700 -23.8423228 42.41572 -41.379768 59.95317 2022 2.291796 -35.5194736 40.10307 -55.535551 60.11914 2023 -4.703306 -47.0197310 37.61312 -69.420693 60.01408 2024 -11.698378 -58.3947964 34.99804 -83.114387 59.71763 2025 -18.693454 -69.6768774 32.28997 -96.665871 59.27896 2026 -25.688529 -80.8876528 29.51059 -110.108304 58.73124 2027 -32.683605 -92.0423157 26.67511 -123.464919 58.09771 Yield Point Forecast(COASTAL REGION) Fig:20 Yield forecast Year Yield point forecast Lo 80 Hi 80 Lo95 Hi 95 2018 25404.50 22497.42 28311.58 20958.51 29850.49 2019 26290.00 22043.95 30536.06 19796.22 32783.79 2020 27175.51 21815.12 32535.89 18977.50 35373.51 2021 28061.01 21691.92 34430.10 18320.33 37801.69 2022 28946.51 21630.53 36262.50 17757.68 40135.35 2023 29832.01 21609.56 38054.47 17256.85 42407.18 2024 30717.52 21616.92 39818.11 16799.36 44635.67 2025 31603.02 21645.13 41560.91 16373.74 46832.30 2026 32488.52 21689.21 43287.84 15972.39 49004.65 2027 33374.02 21745.71 45002.33 15590.06 51157.99 Production Point Forecast(COASTAL REGION) Fig:21 Production forecast Year Production Point forecast Lo 80 Hi 80 Lo95 Hi 95 2018 129.4848 105.739621 153.2300 93.16968 165.7999 2019 122.6794 90.200070 155.1586 73.00657 172.3521 2020 108.9265 71.588542 146.2645 51.82301 166.0300 2021 120.9781 70.405265 171.5510 43.63361 198.3226 2022 122.8605 58.419666 187.3014 24.30673 221.4143 2023 112.2065 37.588853 186.8241 -1.91134 226.3243 2024 116.2073 27.936700 204.4780 -18.79096 251.2057 2025 120.5945 16.284411 224.9046 -38.93402 280.1230 2026 114.2439 -4.218647 232.7065 -66.92895 295.4168 2027 113.9214 -19.846723 247.6895 -90.65927 318.5020 Fig:22. Area-Residuals(CEDED) Fig:23:Yield-Residua(Ceded) Fig:24 Prod.-Residuals(Ceded)
- 7. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 52 doi: 10.32622/ijrat.711201939 Table-5 Residuals & Predictive values of Area, Yield and Productions(CEDED REGION) year A residuals Y residuals P residuals A predective Y predictive P predective 2003 0.5205565 1.191376 0.2696697 1164.5206 266.1914 603.2697 2004 -0.7857512 -1.805380 0.6757360 1510.2142 3453.1946 1267.6757 2005 -214.9619841 -533.095212 -463.3631946 1339.0380 2266.9048 460.6368 2006 -578.0808788 -397.622113 -754.3694676 462.9191 2124.3779 -388.3695 2007 410.4950817 2709.383368 1316.2430014 1884.4951 8987.3834 3392.2430 2008 -93.4683524 -1638.368125 -797.0453050 1353.5316 339.6319 -326.0453 2009 -467.9382883 -2070.179756 -883.6285771 523.0617 517.8202 -390.6286 2010 -764.8823535 -450.995919 540.5922361 -628.8824 3015.0041 1502.5922 2011 999.8238396 370.213894 -217.0168277 2057.8238 3051.2139 226.9832 2012 41.7188844 -116.687055 -57.7366664 1130.7189 2783.3129 577.2633 2013 28.7763566 859.505100 252.7839521 1139.7764 4792.5051 991.7840 2014 -260.9630661 -144.666116 -250.2469229 571.0369 2629.3339 140.7531 2015 -66.8461418 1241.299335 145.7239208 665.1539 6026.2993 839.7239 2016 261.3288712 -442.772159 2.3333701 1229.3289 2669.2278 487.3334 2017 -246.6139287 2196.844463 358.7155252 450.3861 8439.8445 1300.7155 Table-6 Residuals & Predictive values of Area, Yield and Productions(COASTAL ANDHRA REGION) Fig:25. Area-Residuals Fig:26:Yield-Residuals Fig:27 Prod.- Residuals Table-7 Area, yield , and Production predictive (CEDED REGION & Coastal Region) CEDED REGION Coastal Region year Area predictive Yield predictive Prod. predective A REA Predictive YIELD predictive P RODUCTION predictive 2018 663.64312 4338.339 670.0918 30.313146 25404.50 18.52847 2019 630.28624 5078.286 818.4277 23.270686 26290.00 25.34372 2020 596.92936 5563.951 724.4765 16.282895 27175.51 29.13498 2021 563.57247 5098.539 770.2133 9.286700 28061.01 39.46221 2022 530.21559 5657.126 735.4144 2.291796 28946.51 50.28347 2023 496.85871 5666.277 747.0475 -4.703306 29832.01 58.22446 2024 463.50183 5738.279 731.9108 -11.698378 30717.52 68.87795 year A residuals Y residuals P residuals A predective Y predictive P redective 2003 0.05992661 5.421122 0.07110694 134.05993 12127.42 159.07111 2004 -0.17530698 -7.035089 -0.17754246 135.82469 16241.96 174.82246 2005 -9.75765955 1334.323530 -2.62360133 113.24234 17154.68 183.37640 2006 -3.70558785 -3766.917577 -24.31993545 110.29441 13292.08 132.68006 2007 6.55777815 -2150.927389 19.73043580 121.55778 14070.07 202.73044 2008 8.12492792 39.571429 -8.45449021 126.12493 17329.57 179.54551 2009 -15.87964846 -6.943517 -15.82348111 80.12035 18309.06 138.17652 2010 -1.17360226 -2309.309307 -22.50726628 89.82640 14544.69 120.49273 2011 -6.96986748 531.311061 -19.24844333 70.03013 18629.31 118.75156 2012 -0.83489672 -48.087117 20.78926920 69.16510 18745.91 163.78927 2013 -19.61687161 3508.860583 -21.85512723 22.38313 26742.86 84.14487 2014 5.89437125 -1017.538421 5.02948856 47.89437 22260.46 107.02949 2015 32.44007125 -3902.650475 35.66558397 98.44007 16313.35 177.66558 2016 -8.91152621 2760.767403 1.60879507 37.08847 26524.77 118.60880 2017 -4.15845853 -135.428640 8.16232185 32.84154 24383.57 114.16232
- 8. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 53 doi: 10.32622/ijrat.711201939 2025 430.14495 5998.293 732.2079 -18.693454 31603.02 81.39360 2026 396.78807 6050.198 723.6068 -25.68529 32488.52 92.43685 2027 363.43118 6216.148 720.1359 -32.683605 33374.02 104.37980 2028 330.07430 6367.048 713.7073 -39.678681 34259.53 118.01384 2029 296.71742 6480.838 708.9839 -46.673756 35145.03 131.27342 2030 263.36054 6636.899 703.2773 -53.668832 36030.53 144.74957 2031 230.00366 6769.308 698.1376 -60.663908 36916.03 159.43533 2032 196.64678 6905.240 692.6710 -67.658983 37801.53 174.34269 2033 163.28990 7048.485 687.3929 -74.654059 38687.04 189.32485 2034 129.93301 7183.148 682.0062 -81.649135 39572.54 205.09936 2035 96.57613 7322.712 676.6820 -88.644210 40458.04 221.33180 2036 63.21925 7461.466 671.3218 -95.639286 41343.54 237.69702 2037 29.86237 7598.782 665.9824 -102.634362 42229.05 254.58425 Table-8 Time series data values of Area, Yield and Production(CEDED REGION & COASTAL REGION) Fig: 28 Time series data values of Area, Yield and Production(CEDED REGION) Fig: 29 Time series data values of Area, Yield and Production(COASTAL REGION) Table-9 Area, yield and Production Training Set error measure CEDED REGION ARIMA Training set error measures ME RMSE MAE MPE MAPE MASE ACF AREA (1 , 2, 1) -63.45848 413.5649 295.8136 -40.37166 61.14672 0.913206 -0.2363987 YIELD ( 0, 2, 1) 105.483 1222.83 878.3086 -60277613 24.9433 0.5778618 -0.13744871 PRODUCTION ( 2, 1, 2) -53.73797 550.8458 402.7163 -34.2153 62.67673 0.7516369 -0.388102 COASTAL ANDHRA REGION AREA ( 1 , 2 ,1) -1.20709 11.80239 8.284033 -2.725881 12.83502 0.7387036 -0.1005916 YIELD ( 0,2,1) -522.2153 2040.165 1435.006 -3.393128 7.383666 0.7732607 -0.05819921 PRODUCTION ( 1, 2 ,1) -1.596859 17.24914 13.73779 -1.172856 9. 9.771143 0.7425834 -0.09666837 CEDED REGION COASTAL REGION Year TIMESERIES DATA-area TIMESERIES DATA-yield TIMESERIES DATA-prod. TIMESERIES DATA-AREA TIMESERIES DATA-YIELD TIMESERIES DATA-PROD. 2003 1164 2665.1914 603.2697 134.05993 12127.42 159.07111 2004 1511 3453.1946 1267.6757 135.82469 16241.96 174.82246 2005 1554 2266.9048 460.6368 113.24234 17154.68 183.37640 2006 1041 2124.3779 -388.3695 110.29441 13292.08 132.68006 2007 1474 8987.3834 3392.2430 121.55778 14070.07 202.73044 2008 1447 339.6319 -326.0453 126.12493 17329.57 179.54551 2009 991 517.8202 -390.6286 80.12035 18309.06 138.17652 2010 136 3015.0041 1502.5922 89.82640 14544.69 120.49273 2011 1058 3051.2139 226.9832 70.03013 18629.31 118.75156 2012 1089 27832.3129 577.2633 69.16510 18745.91 163.78927 2013 1111 4792.5051 991.7840 22.38313 26742.86 84.14487 2014 832 2629.3339 140.7531 47.89437 22260.46 107.02949 2015 732 6026.2993 839.7239 98.44007 16313.35 177.66558 2016 968 2669.2278 487.3334 37.08047 26524.77 118.60880 2017 697 8439.8445 1300.7155 32.84154 24383.57 114.16232
- 9. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 54 doi: 10.32622/ijrat.711201939 V. CONCLUSION A. Area of Groundnut Crop Conclusion Table-10 Identification of ARIMA(p,d,q) MODEL for AREA in Ceded & Coastal Andhra Regions Identification of ARIMA(p,d,q) MODEL for AREA (Ceded Region) Model Area ARIMA Coefficients SE Intercept σ^2 Log likelyhood AIC (1,0,1) AR1 0.0498 0.5624 1046.008 115069 -108.74 225.49 MA1 0.2924 0.5138 117.233 (1,1,1) AR1 0.1970 0.3672 134509 -102.82 211.65 MA1 -0.7633 0.2536 AR2 -0.3039 0.2835 MA1 -0.5565 0.3623 (0,1,1) MA1 -0.6696 0.2105 137404 -102.98 209.95 (1,2,1) AR1 -0.2117 0.2744 183257 -98.74 203.48 MA1 -1.000 0.2106 (1,1,0) AR1 -0.2581 0.2570 171256 -104.26 212.51 (1,1,2) AR1 -0.1588 0.7451 131574 -102.68 213.35 MA1 -0.3572 0.6960 MA2 -0.2946 0.4351 (1,2,0) AR1 -0.4470 0.2389 378348 -102.04 208.08 (0,2,1) MA1 -1.000 0.2025 197349 -99.02 202.04 Identification of ARIMA(p,d,q) MODEL for AREA in (Coastal Andhra Region) model Area ARIMA Coefficients SE Intercept σ^2 Log likelyhood AIC (1,0,1) AR1 0.9190 0.1088 85.7207 198.3 -62.03 132.07 MA1 0.1533 0.3450 32.5630 (1,1,1) AR1 -0.8244 0.2173 184.7 -56.82 119.65 MA1 0.9999 0.4058 (2,2,1) AR1 -0.2573 0.2572 131.5 -52.07 112.14 AR2 -0.3591 0.2522 MA1 -1.0000 0.2971 (1,2,2) AR1 0.4313 0.3430 114.8 -52.14 112.29 MA1 -1.9696 0.4247 MA2 0.9999 0.4251 (1,1,0) AR1 0.0648 0.2612 203.3 -57.07 118.14 (0,0,1) MA1 0.9157 0.2666 86.4312 431 -67.66 141.32 10.0006 (1,2,0) AR1 -0.4331 0.2426 321.5 -56.07 116.15 (1,0,0) AR1 0.9418 0.702 86.0175 200.7 -62.14 130.27 35.3574 In the present study, the ARIMA (1,2,1) in Ceded Region & ARIMA (2,2,1) in Coastal Region were the best fitted model through the minimum value of AIC, then used for prediction up to 10 years of the area of groundnut in ceded districts using 15 years time series data i.e. from 2003-2004 to2017-2018. ARIMA(1,2,1) in Ceded Region & ARIMA (2,2,1) in Coastal Region were used because the reason of its capability to make prediction using time series data with any kind of patterns and with auto correlated successive values of the time series. The study was also validated and statistically tested that the successive residuals in the fitted ARIMA (1, 2,1 ) in Ceded Region & ARIMA (2,2,1) in Coastal Region were not correlated, and the residuals appear to be normally distributed with the mean zero and constant variance. Hence, it can be a satisfactory predictive model for the groundnut area in ceded districts in Andhra Pradesh for the period of 2018 to 2027. The ARIMA (1,2,1) in Ceded Region & ARIMA (2,2,1) in Coastal Region models projected an increment in the area for the duration of 2018 to 2027. The prediction of 2027 is resulted approximately 363.4312’000 ha & -32.683605”000 ha. Like any other predictive models for forecasting , ARIMA model has also limitations on accuracy of the predictions yet it is widely used for forecasting the future values for time series. B. Yield of Groundnut Crop Conclusion Table 11 Identification of ARIMA(p,d,q) MODEL for Yield in Ceded & Coastal Andhra Regions Identification of ARIMA(p,d,q) MODEL for YIELD (CEDED REGION) Model YIELD ARIMA Coefficie nts SE Intercept σ^2 loglikely- hood AIC
- 10. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 55 doi: 10.32622/ijrat.711201939 (1,2,2) AR1 -0.4498 0.3036 1951664 -114.66 237.32 MA1 -1.4436 0.3670 MA2 0.5442 0.3058 (2,2,1) AR -0.9559 0.2468 1725361 -114.44 236.89 AR2 -0.4515 0.2410 MA1 -1.0000 .7099 (1,1,2) AR1 -0.3016 0.5638 1648515 -120.7 249.41 MA1 -0.8152 0.4808 MA2 0.2104 0.6083 (1,0,0) AR1 -0.4288 0.2773 3437.8939 1374004 -127.38 260.77 217.0428 (0,0,1) MA1 -0.4099 0.3307 3423.1423 1403675 -127.54 261.07 195.9421 (1,0,1) AR1 -0.4054 0.5620 3437.076 1373603 -127.38 262.77 MA1 -0.0303 0.6045 215.595 Identification of ARIMA(p,d,q) MODEL for YIELD(COASTAL ANDHRA REGION) Model YIELD ARIMA Coefficients SE Intercept σ^2 loglikely- hood AIC (1,1,1) AR1 0.3432 0.7610 5185326 -128.1 262.21 MA1 -0.4558 0.6856 (2,1,1) AR1 -0.3786 1.0185 4269286 -127.02 262.03 AR2 -0.4795 0.3230 MA1 0.3476 1.2693 (1,0,1) AR1 0.4863 0.4462 18661.461 4814922 -137.4 282.8 MA1 0.667 00.478 1712.654 (0,2,1) MA1 -1.0000 0.2202 4802619 -119.77 243.53 (1,2,1) AR1 -0.1537 0.2740 160.7 -52.93 111.86 MA1 -1.0000 0.2183 In the present study, ARIMA (1, 2 ,1 ) in ceded and ARIMA (2,2,1)coastal Andhra regions) were the best fitted models through the minimum value of AIC, then used for prediction up to 10 years of the yield of groundnut in ceded districts using 15 years time series data i.e. from 2003-2004 to 2017-2018. ARIMA ( 1, 2 ,1 ) , ARIMA (2,2,1)coastal Andhra regions used because the reason of its capability to make prediction using time series data with any kind of patterns and with auto correlated successive values of the time series. The study was also validated and statistically tested that the successive residuals in the fitted ARIMA (1,2 ,1 )ceded region, ARIMA (2,2,1)coastal Andhra regions were not correlated, and the residuals appear to be normally distributed with the mean zero and constant variance. Hence, it can be a satisfactory predictive model for the groundnut yield in ceded districts in Andhra Pradesh for the period of 2018 to 2027. The ARIMA ( 1,2,1) ceded region, ARIMA (2,2,1)coastal Andhra regions model projected an increment in the yield for the duration of 2018 to 2027. The prediction of 2027 is resulted approximately 6216.148 ’kg/ ha .(Ceded Region) & 33374.02 ’kg/ha(Coastal Andhra Region). Like any other predictive models for forecasting, ARIMA model has also limitations on accuracy of the predictions yet it is widely used for forecasting the future values for time series. C. Production of Groundnut Crop Conclusion TABLE-12 Identification of ARIMA(p,d,q) MODEL for Production in Ceded & Coastal Andhra Regions Identification of ARIMA(p,d,q) MODEL for PRODUCTION(CEDED REGION) Model PROD. ARIMA Coefficients SE Intercept σ^2 Log likelihood AIC (1,0,1) AR1 -0.2205 0.4445 765.2733 159820 -111.21 230.42 MA1 -0.1346 0.4025 74.9200 (2,1,1) AR1 -0.9458 0.2739 139690 -103.77 215.54 AR2 -0.6215 0.2299 MA1 -0.2511 0.4060 (1,0,0) AR1 -0.3329 0.2349 765.9133 161114 -111.27 228.53 79.0824 MA1 -0.7552 0.2130 Identification of ARIMA(p,d,q) MODEL for PRODUCTION(COASTAL ANDHRA REGION) Model PROD. ARIMA Coefficients SE Intercept σ^2 Log likelihood AIC (1,2,0) AR1 -0.4091 0.2393 1027 -63.61 131.22 (1,1,2) AR1 -0.3850 0.3058 306.1 -60.87 129.73 MA1 0.4757 0.2873 MA2 -0.5243 0.2372
- 11. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 56 doi: 10.32622/ijrat.711201939 (1,2,1) AR1 -0.1537 0.2740 160.7 -52.93 111.86 MA1 -1.0000 0.2183 (1,0,1) AR1 0.1425 0.2586 146.7956 328.6 -66.27 140.54 MA1 1.0000 02214 10.4633 In the present study, the ARIMA (2,1,1)in Ceded Region & ARIMA(1,2,1)in Coastal Andhra Region were the best fitted model through the minimum value of AIC, then used for prediction up to 10 years of the production of groundnut in ceded districts using 15 years time series data i.e. from 2003-2004 to 2017-2018. ARIMA (2,1,1) in Ceded Region & ARIMA(1,2,1)in Coastal Andhra Region were used because the reason of its capability to make prediction using time series data with any kind of patterns and with auto correlated successive values of the time series. The study was also validated and statistically tested that the successive residuals in the fitted ARIMA (2,1,1) in Ceded Region & ARIMA (1,2,1)in Coastal Andhra Region were not correlated, and the residuals appear to be normally distributed with the mean zero and constant variance. Hence, it can be a satisfactory predictive model for the groundnut yield in ceded districts in Andhra Pradesh for the period of 2018 to 2027. The ARIMA (2, 1, 1) in Ceded Region & ARIMA(1,2,1)in Coastal Andhra Region models projected an increment in the production for the duration of 2018 to 2027. The prediction of 2027 is resulted approximately 720.1359’000 tonnes(Ceded Region) & 163.78927’000 tonnes(Coastal Andhra Region). Like any other predictive models for forecasting, ARIMA model has also limitations on accuracy of the predictions yet it is widely used for forecasting the future values for time series. It is noticed that in Groundnut production Ceded Region is better than Coastal Andhra Region. The empirical Forecasting area, yield and production of groundnut crop: a comparative study of coastal andhra and ceded regions using- R findings of study could help to forecast any such commodities. The researchers and policy makers will thus get access for making further extensive research work. We firmly believe that this research has shed some important light on the subject area encompassing time series forecasts of selected agricultural crops in Seemandhra. These empirical findings can be an important source of information to many researchers and policy formulators as far as agricultural crops in Seemandhra are concerned. ACKNOWLEDGMENT I would like to express my heartfelt gratitude to my guide and Head of Dept. Prof. Mohammed Akhtar Palagiri who acted as a source of inspiration in all spheres of my work for giving valuable guidelines for completeing this course. REFERENCES [1] Madhusudana, B, “A survey on area, production and productivity of groundnut crop in India”, IOSR Journal of Economics and Finance, volume 1, PP 01-07, (2013). [2] Muhammad Iqbal Ch, “Forecasting of wheat production: A comparative study of Pakistan And India”, IJAR 4(12), 698-709 , (2016). [3] Rachana Wankhade, “Use of the ARIMA Forecasting pigeon Pea production in India “, International Review of Business and Finance, Volume 2, Number 1, pp 97-102, ( 2010). [4] N.M.F. Rahman , “ Forecasting of Boro rice production in Baangladesh: An Approach”. J.Bangladesh Agril. Univ. ; 8(1):103-112p, (2010). [5] Najeeb Iqbal , “Forecasting wheat area and production in Pakistan”, Journal of Agriculture & Sciences, Volume.1, No.2, (2005). [6] M.K Debnath , ” Forecasting Area, production, and Yield of Cotton in India using ARIMA Model”, Journal of Space Science & Technology, Volume 2, Issue 1. (2013). [7] M. Hemavathi , “ ARIMA Model for Forecasting of Area, Production and productivity of Rice and Its Growth Status in Thanjavur District of TamilNadu”, India, International of Current Microbiology and Applied Sciences, 7(2):149-156, (2018). [8] P.K. Sahu, “Modelling and forecasting of area, production, yield and total seeds of Rice and Wheat in SAARC Countries and the World towards Food Security”, SCIEP, American Journal of Applied Mathematics and Statistics, Vol.3, No.1,34-38., (2015). [9] Mohammed Amir Hamjah, “ Rice Production Forecasting in Bangladesh: An Application of Box-Jenkins ARIMA Model”, Mathematical Theory and Modelling , Vol. 4, No 4, (2014). [10] Muhammad, F., M., Siddiqqe, M. Bashir, S. Ahmad , “ Forecasting Rice Production in Pakistan-Using ARIMA Models “, J. Of Animal Plant Sci.; 2:27-3lp. (1992). [11] Niaz Md. Farhat Rahman , “ Modeling for Growth and Forecasting of pulse production in Bangladesh”, Research journal of Applied Sciences, Engineering and Technology 5 (24):5578-5587. (2013). [12] Vishwajith K..P, “Timeseries Modeling and forecasting of pulses production in India”, Journal crop and weed, 10 (2):147-154, (2014). [13] Ashwin Darekar, “Forecasting oilseeds prices in India: Case of Groundnut”, J.Oilseeds Res..,34(4):235- 240, (2017). [14] Bhola Nath, DS , “Forecasting Wheat production in India: An ARIMA modelling approach”, Journal of Pharmacognosy and Phytochemistry, 8(1):2158-2165, (2018). [15] Pant, D.C. and Pradeep Pal, , “Comparative Economics of Agro-processing units for Groundnut in Southern Rajasthan”, Indian Journal of Agricultural Marketing, Vol, 18, No.1, January-(2004). [16] Ap Patel, G.N., and N.L. Agarwal, ‘’Price Behaviour of Groundnut in Gujarat”, Indian Journal of Marketing, Vol.7, No.2, pp.50-57, ( 1993). [17] Mohammad Mayazzem Hossain, , “ Comparision of ARIMA and Neural Net Work Model to forecast the jute Production in Bangladesh”, Jahingir Nagar University Journal of Science, Vol.40, pp11-18, (2017). [18] Ansari M I and Ahmed SM,” Time series analysis of tea prices: an application of ARIMA modelling and co- integration analysis”, IJE,48; 49-54. (2001). [19] Croxton.F.E., Cowden, D.J. and Klein, S, ”Applied General Statistics”, New Delhi: Preentice Hall of India PVT. Ltd. (1979). [20] Pankratz, A “Forecasting with Univariate Box – Jenkins models: concepts and cases,” John Wiley, New York. ,( 1983). [21] Box, G.E.P., Jenkins, G.M. and Reinsel, G.C “ Time series analysis. Forecasting and control,” Pearson Education, Delhi. , ( 1994). [22] Makridakis, S., Wheelwright, S.C. and Hyndman, R.J, “Forecasting Methods and Applications”, 3rd Edition, John Wiley, New York. ,( 1998)
- 12. International Journal ofResearch in Advent Technology, Vol.7, No.11, November 2019 E-ISSN: 2321-9637 Available online at www.ijrat.org 57 doi: 10.32622/ijrat.711201939 Author-1 Photo Author-2 Photo Author-3 Photo AUTHORS PROFILE Mr. Ananda Kumar Ginka currently working as a Lecturer in Statistics in department of statistics, Sri Srinivasa Degree college, Madanapalle, Chittoor District, Andhrapradesh. He completed M.Sc., M.Ed.,M.Phil., and pursuing Ph.D , Research Scholar, Department of Statistics, Sri Krishnadevaraya University, Anantapuramu-515003, Andhra Pradesh, India. Mr. Dhanunjaya Sunkara, Has Completed M.Sc.,In Statistics And Pursuing Ph.D, Research Scholar, Department Of Statistics, Sri Krishnadevaraya University, Anantapuramu-515003, Andhra Pradesh. India. Th Dr. Mohammed Akhtar Palagiri, Working as A Professor and Head & Bos, Chairman, Department of Statistics, Sri Krishnadevaraya University, Anantapuramu-515003, Andhra Pradesh, India. Dr. Mohammed Akhtar Palagiri, Has Completed M.Sc., M.Phil., Ph.D., In Statistics. He Has Completed 38 Research Papers in His Portfolio.