Learn R Programming

telefit (version 1.0.3)

Estimation and Prediction for Remote Effects Spatial Process Models

Description

Implementation of the remote effects spatial process (RESP) model for teleconnection. The RESP model is a geostatistical model that allows a spatially-referenced variable (like average precipitation) to be influenced by covariates defined on a remote domain (like sea surface temperatures). The RESP model is introduced in Hewitt et al. (2018) . Sample code for working with the RESP model is available at . This material is based upon work supported by the National Science Foundation under grant number AGS 1419558. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Copy Link

Version

Install

install.packages('telefit')

Monthly Downloads

166

Version

1.0.3

License

GPL-3

Maintainer

Joshua Hewitt

Last Published

February 3rd, 2020

Functions in telefit (1.0.3)

lon_trans

Formatting for longitude scales in ggplot spatial maps
maternArray

Matern covariance
lat_trans

Formatting for longitude scales in ggplot spatial maps
kronSamp

Samples a multivariate normal with a Kronecker product covariance structure
plot.stData

Plot stData objects
plot.stFit

Plot stFit objects
eof

Performs an EOF decomposition of the data
errDump

Wrapper for a function to dump errors from C++
dgemkmm

Evaluate kron(A,B) * C without storing kron(A,B)
invWSamp

Samples an Inverse-Wishart matrix
maternCov

Matern covariance
mergeMean

Combine sample means from two samples
forwardsolve.kron

Solves a triangular system with a Kronecker product structure
maternEffectiveRange

Compute effective range for Matern correlation to drop to a specified level
mergeVar

Combine sample variances from two samples
stSimulate

Simulate responses from the spatio-temporal teleconnection model
stVIF

Computes variance inflation factors for fixed effects of the teleconnection model
plot.stPredict

Plot stPredict objects
rwishart

Random wishart matrix
extractStData

Basic extraction of SpatialGridDataFrame data for teleconnection analysis
mergeCovmat

Combine sample covariance matrices from two samples
mergeComposition

Combine results from composition sampler
stLL

Compute log likelihood for model
extractRegion

Extract region from a SpatialGridDataFrame
rmatnorm

Simulate matrices from matrix normal distributions
stPredict

Compute forecasts based on posterior samples
stEval

Basic evaluation of fit
stFit

Fit the remote effects spatial process (RESP) model
plot.teleCor

Plots teleconnection correlation maps
summary.stPredict

Plot stPredict objects
summariseEOFAlpha

Summarize eof-mapped alphas
svcPredict

Make predictions using a fitted varying coefficient model
summariseAlpha

Summarize alphas
teleCor

Pointwise correlations for an exploratory teleconnection analysis
svcFit

Fit a spatially varying coefficient model
telefit

Tools for modeling teleconnections
coprecip.fit

Sample MCMC output for the RESP model
HPDinterval.stFit

Compute Highest posterior density intervals from posterior samples
arrayToLong

Reshape array of data matrices into long format
coef.stFit

Compute point estimates for parameters from posterior samples
abind3

Convenience function for stacking matrices into an array.
cca.predict

Make predictions using canonical correlation analysis (CCA)
coef.stPredict

Compute point estimates for parameters from posterior samples
coprecip

Standardized anomalies of CO Precipitation
coprecip.predict

Sample composition sampling output for the RESP model