TauferLab/Weather_Data_Analytics

This repository contains the MATLAB software for the frequency based analysis framework developed for our paper in eScience 2015 conference. The framework is an adaptation of a cluster tool previously proposed to predict idle resources in non-dedicated clusters. The framework employs empirical cumulative distribution function to benchmark and model occurrences of extreme climate events, specifically extreme temperature and precipitation. The framework is broken into two phases: learning phase and prediction phase. The learning phase uses ECDF-based analysis to generate modeling and forecasting windows. The prediction phases applies the modeling window to most recent weather data to estimate the likelihood that given proportions of the region can experience extreme temperature and precipitation events. This repository also contains a sample input in INPUT folder to test the functionality of the framework.

MATLABNOASSERTION