Strengthening Statistical Computing for NARS

Abstract

Statistical computing provides a strong base for any scientific investigation and enables to answer quantitative biological questions so that the amount of information generated is maximized. The present project on Strengthening Statistical Computing for NARS (SSCNARS) funded by NAIP of ICAR targets at providing research guidance in statistical computing and creating sound and healthy statistical computing environment through advanced, versatile, innovative and state-of the-art high end statistical packages for analysis of data so as to enable drawing meaningful and valid inferences and converting research output into knowledge. This consortium has been implemented in 3-tier structure with first tier as lead centre, 08 NARS organizations as Statistical Computing Hubs and 142 other NARS organizations as nodes; signed MOU with each NARS organization for the enhanced usage of the environment created which has resulted into effective collaboration among all NARS organizations. This effort is unique in the sense that all NARS organizations are linked with each other. This has empowered the statisticians and other researchers by providing the facilities of statistical computing to them as is available at premier Institutes. It has provided the statisticians an opportunity to interact with each and every researcher of NARS through training programmes and e-resources. Statistical Computing labs have been established at lead Centre (with 20 desktop computers) and at 08 other Statistical Computing Hubs (with 10 desktop computers at each hub). General purpose Statistical Software Package(s) SAS EAS, JMP and JMP Genomics have been procured with 151 licenses (100 Standalone and 50 Intranet Based) + SAS Enterprise Business Intelligence Server with all modules of SAS Intranet based for perpetual use with three years updates and upgrades by 151 different NARS Organizations. It can be installed on multiple official machines. The software is installed on more than 2750 computers of 151 NARS organizations (on an average more than 18 computers per NARS organization). More than 200 researchers/nodal officers from 151 NARS organizations have been trained through 35 Orientation-cum-Installation Training Programmes. This activity has ensured that trained manpower to install the software is available at all the NARS organizations. Capacity building efforts have been made through organization of training programmes including subject specific ones, preparing e-manuals and case studies made available at subproject website www.iasri.res.in/sscnars for rejuvenating statistical thinking and use of appropriate advanced statistical techniques in agricultural research and teaching. For creating a base for resource persons, 209 trainers have been trained through 30 working days training programmes on SAS: A Comprehensive Overview and SAS Genetics/JMP Genomics and 6 days training programme on Data Analysis Using SAS across 85 NARS organizations. In order to provide customized training to researchers in NARS, 2166 research personnel have been trained on Data Analysis using SAS through 104 training programmes of one week duration each including 37 trainings at different organizations and 15 subject specific customized training programmes for specific needs of different institutions. 164 researchers have been trained by nodal officers at Tamil Nadu Agricultural University, Coimbatore and NRCAF, Jhansi; IGFRI, Jhansi; Junagarh Agricultural University, Junagarh. (156 research scholars and 90 RA/SRFs) have also been trained. 2 To sensitize the scientists in NARS with the statistical computing capabilities available for enhancing their computing and research analytics skills, organized 21 sensitization workshops at different NARS organizations. For efficient and effective use of Internet Technologies for delivering lectures to save funds and resources, several Webinar sessions were organized in collaboration with CRIDA, Hyderabad where the licenses for Webinar sessions have been procured. Webinar sessions were also organized through the use of Social Networking Tools (Google Hangouts). For providing service oriented computing, developed and established Indian NARS Statistical Computing portal, which is available to NARS users through IP authentication at http://stat.iasri.res.in/sscnarsportal. It is a paradigm of computing techniques that operate on software-as-a-service. There is no need of installation of statistical package at client side. Twenty-four different modules of analysis of data are available on this portal. These modules can be used by uploading *.xls, *.xlsx, *.csv and *.txt files and results can be saved as *.RTF or *.pdf files. This has helped the researchers in NARS in analyzing their data in an efficient manner without losing any time. The portal is extensively being used throughout NARS and has helped the researchers in analyzing their data without losing any time. INDIAN NARS STATISTICAL COMPUTING PORTAL (both content and software) are copyrighted material of ICAR with copyright numbers: L-55719/2013 and SW-7397/2013 issued on October 25, 2013 by Registrar, Copyrights. Some other IP authenticated services like Web Report Studio, BI DashBoard, Web OLAP Viewer, E-Miner 6.1, E-Miner 7.1 have also been provided to the researchers of NARS. For creation of research data repository and standardization of analysis of data for All-India Coordinated Sorghum Crop Improvement Project, a prototype has been developed by NAARM, Hyderabad and IASRI, New Delhi in collaboration with Directorate of Sorghum Research Hyderabad. This system was made operational and now made available at www.aicsip.naarm.org.in. This would be shifted to IASRI, New Delhi after validation and testing. The prototype for automation of All India Coordinated Sorghum Improvement Project needs to be scaled up to cover all AICRPs through a separate mega network project for paving the way for developing research data repositories, standardizing the analytical modules and saving on time and resources. Already work in collaboration with AICRP on Vegetable Crops has been initiated. For customized analysis, macros for analysis of augmented designs, split-split plot designs, split factorial designs, strip plot designs, econometric analysis (diversity indices, instability index, compound growth rate, Garret scoring technique and demand analysis using LA-AIDS model) have been developed and made available on the project website which are being extensively by researchers. Reference manuals have been prepared for the benefit of the researchers which include live examples from different areas of agriculture, animal and fisheries research illustrating the power of newer statistical computing techniques and are made available at the website of subproject. These include Data Analysis Using SAS, Genetics/ Genomics Data Analysis Using SAS, Data Analysis in Social Sciences Research Using SAS, Data Mining Using SAS, Data Analysis of Agroforestry Experiments Using SAS, Data Analysis of Natural Resources Management Research Using SAS; Data Analysis of Farm Implements and Machinery Research; Data Analysis in Dairy Sciences Research; Data Analysis Using R. 3 The capacity building efforts have paved the way for publishing research papers in the high impact factor journals. Based on feedback received from NARS organizations, 173 research reports, 260 research papers have been published / accepted for publication (in journals like Animal Food Science and Technology, Field Crops Research, Journal of Food Engineering, Euphytica, Journal of Applied Polymer Science, etc.) by analyzing the data using high end statistical computing facility (out of these 260 research papers, 135 are in Journals with NAAS rating 6.0); more than 300 students have used this in their dissertations; 1700 students have used in their course work. The number of data sets analyzed is more than 6000 across NARS. Nodal officer from CMFRI, Kochi has reported a saving of 20 man months in compilation of data related to Marine Fish Household Census 2010 consisting of 10 lakh households with 16 attributes. Website of the project is being maintained and updated regularly. Website has been registered under google analytics on November 15, 2010. Till March 31, 2014 there were 43,239 page views across 529 cities of 82 countries. Average time on page is 3.03 minutes. The total hits to the portal are: 1,97,816 since April 01, 2011, which amounts to more than 100 hits per day. Based on the user logged information, the total number of logged in users from Indian NARS since March 2012 are 24,926. It has been included in the FOCARS training programmes conducted by NAARM, Hyderabad and 799 scientists have been sensitized about the capabilities of the software and data analysis using SAS. Also 211 scientists were sensitized through Refresher Courses conducted at NAARM, Hyderabad about data analysis using SAS. The high end Statistical Computing Package procured with perpetual license can be used till hardware and operating systems would support the versions available as on June 29, 2013. Hence, the statistical computing environment created would be useful in all future research endeavours. The capacity building efforts made are expected to have a multiplier effect. The activities of Strengthening Statistical Computing for NARS are proposed in EFC for XII Plan of IASRI and would be continued in project mode from April 01, 2014. This would facilitate implementation of updating license files, resolving maintenance issues and maintenance of service oriented computing modules. To further harness the benefits of the healthy statistical computing environment created and meeting the future requirements, the activities of enhancing Statistical Computing for NARES may be continued in the form of a network project with provision of appropriate funds. It would also help in strengthening of Indian NARS Statistical computing portal by adding more modules of statistical analysis of data to culminate into a service oriented computing resource for standardization of statistical analysis of data. Capacity building efforts in statistical concepts and statistical computing need to be pursued on continuing basis including international exposure for synthesizing various statistical packages into the basket. More statistical packages including freeware may be brought under this umbrella. In future, for procurement of statistical packages as per requirement and updates and upgrades of the existing ones, enough provision of funding need to be made.Not Available