R BASED ANALYSIS OF FORECASTING AREA, PRODUCTION AND PRODUCTIVITY OF CEREALS IN CHHATTISGARH PLAINS ON THE BASIS OF ARIMA MODELLING USING TIME-SERIES RELATIONAL DATABASE
Loading...
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Indira Gandhi Krishi Vishwavidhyalaya, Raipur
Abstract
The production scenario, its composition in cereals of Chhattisgarh and forecast of the same over future time series are always the matters of concern to the planners of the State of Chhattisgarh. Being predominantly tribalregion, Chhattisgarh has been drawing continuous attention of governmental agencies to improve crop production of cereals. This State comprises of three agro-climatic zones, namely, Chhattisgarh Plains, Bastar Plateau and Northern Hills Region. The Chhattisgarh Plains region comprises of Raipur, Bilaspur, Dhamtari, Janjgir, Kawardha, Korba, Mahasamund, Raigarh, Rajnandgaon districts. In view of above, the present study aims atdeveloping a relational database for assessment of distribution of area, production and productivity of major cereal crops (i.e., Rice, Wheat and Maize, Jowarand Kodo-Kutki) for different districts of Chhattisgarh Plains, to be able to undertake forecast of relevant future time series. Therefore, district wise time series data have been collected period wise from 1980-81 to 2014-15 on above mentioned major cereals and maintained in the relational database using MS-Access-2007for further study.
The Statistical Software “R: The R Project for Statistical Computing”, was used for various statistical analyses. That is, the program codes have been written and executed using “forecast” package in R for estimating necessary parameters of two forecast models, namely, double exponential and ARIMA, undertaken for forecasting area, production and productivity of cereals of different districts of Chhattisgarh Plains.
Some good models have been developed for area, production and productivity, for which forecast were made based on Good fitted models for the next 5years time series wherever model diagnostics permitted. A synopsis and evaluation of Good fitted forecast models have been summarized for (i) Model-1: Double exponential, (ii) Model-2: ARIMA for various districts and crops. For some Bad fitted models, reasons have been identified and suggestions for further improvement have also been incorporated into the above-said synopsis.
Description
R BASED ANALYSIS OF FORECASTING AREA, PRODUCTION AND PRODUCTIVITY OF CEREALS IN CHHATTISGARH PLAINS ON THE BASIS OF ARIMA MODELLING USING TIME-SERIES RELATIONAL DATABASE
Keywords
null