telefit1.0.3 package

Estimation and Prediction for Remote Effects Spatial Process Models

abind3

Convenience function for stacking matrices into an array.

arrayToLong

Reshape array of data matrices into long format

cca.predict

Make predictions using canonical correlation analysis (CCA)

coef.stFit

Compute point estimates for parameters from posterior samples

coef.stPredict

Compute point estimates for parameters from posterior samples

dgemkmm

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

eof

Performs an EOF decomposition of the data

errDump

Wrapper for a function to dump errors from C++

extractRegion

Extract region from a SpatialGridDataFrame

extractStData

Basic extraction of SpatialGridDataFrame data for teleconnection analy...

forwardsolve.kron

Solves a triangular system with a Kronecker product structure

HPDinterval.stFit

Compute Highest posterior density intervals from posterior samples

invWSamp

Samples an Inverse-Wishart matrix

kronSamp

Samples a multivariate normal with a Kronecker product covariance stru...

lat_trans

Formatting for longitude scales in ggplot spatial maps

lon_trans

Formatting for longitude scales in ggplot spatial maps

maternArray

Matern covariance

maternCov

Matern covariance

maternEffectiveRange

Compute effective range for Matern correlation to drop to a specified ...

mergeComposition

Combine results from composition sampler

mergeCovmat

Combine sample covariance matrices from two samples

mergeMean

Combine sample means from two samples

mergeVar

Combine sample variances from two samples

plot.stData

Plot stData objects

plot.stFit

Plot stFit objects

plot.stPredict

Plot stPredict objects

plot.teleCor

Plots teleconnection correlation maps

rmatnorm

Simulate matrices from matrix normal distributions

rwishart

Random wishart matrix

stEval

Basic evaluation of fit

stFit

Fit the remote effects spatial process (RESP) model

stLL

Compute log likelihood for model

stPredict

Compute forecasts based on posterior samples

stSimulate

Simulate responses from the spatio-temporal teleconnection model

stVIF

Computes variance inflation factors for fixed effects of the teleconne...

summariseAlpha

Summarize alphas

summariseEOFAlpha

Summarize eof-mapped alphas

summary.stPredict

Plot stPredict objects

svcFit

Fit a spatially varying coefficient model

svcPredict

Make predictions using a fitted varying coefficient model

teleCor

Pointwise correlations for an exploratory teleconnection analysis

telefit

Tools for modeling teleconnections

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) <doi:10.1002/env.2523>. Sample code for working with the RESP model is available at <https://jmhewitt.github.io/research/resp_example>. 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.

  • Maintainer: Joshua Hewitt
  • License: GPL-3
  • Last published: 2020-02-03