The goal of droughtR is to enable meteorological drought monitoring by
generating non-stationary drought indices under various distributional
assumptions (Normal, Gamma, Zero Inflated Gamma and Weibull). It
computes the stationary (SPI) and the non-stationary (NSPI) Standardized
Precipitation Indices using General Additive Models for Location Scale
and Shape (GAMLSS).
# Install the development version on Github
devtools::install_github("mammask/droughtR")The function computenspi creates stationary and non-stationary
meteorological drought indices at different time scales. By default,
droughtR uses the gamma distribution:
# Load droughtR library
library(droughtR)
#> Registered S3 method overwritten by 'quantmod':
#> method from
#> as.zoo.data.frame zoo
# Generate synthetic monthly rainfall data using the Gamma distribution
rain = dummyrainfall(startYear = 1950, endYear = 2010)
# Compute the non-stationary standardized precipitation index (NSPI) for scale 12 using GAMLSS
nonstatdrought = computenspi(x = rain, stationaryspi = FALSE, spiScale = 12, dist = 'gamma')
#> GAMLSS-RS iteration 1: Global Deviance = 3460.033
#> GAMLSS-RS iteration 2: Global Deviance = 3460.018
#> GAMLSS-RS iteration 3: Global Deviance = 3460.018
# Plot NSPI
plot(nonstatdrought)
# Compute the stationary standardized precipitation index (NSPI) for scale 12 using GAMLSS and the weibull distribution
statdrought = computenspi(x = rain, stationaryspi = TRUE, spiScale = 12, dist = 'weibull')
#> GAMLSS-RS iteration 1: Global Deviance = 3545.498
#> GAMLSS-RS iteration 2: Global Deviance = 3539.194
#> GAMLSS-RS iteration 3: Global Deviance = 3539.143
#> GAMLSS-RS iteration 4: Global Deviance = 3539.143
# Plot SPI
plot(statdrought)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent