stFit object containing posterior samples for model

number of posterior samples to reject before computing estimates

The target probability content of the intervals

Any number of matrices of equal dimension to stack together into a 3d matrix

3 dimensional array of matrices to extract to long format

Spatial coordinates associated with the data (longitude in first column)

Vector with labels for the years

An (nvars x nobs) data frame or matrix in which each column contains all observations of measured (predictor) variables for a given timepoint or sample. For example, if X represents a spatial variable that was recorded at several timepoints, then each row of X should contain the variable's measurement for all timepoints at a single location.

An (nvars x nobs) data frame or matrix in which each column contains all observations of measured (response) variables for a given timepoint or sample.

An (nvars x nobs.new) data frame or matrix of values to use to predict Y.new using CCA.

An integer less than (nobs) indicating how many eigenvectors of (X) to use in the CCA.

An integer less than (nobs) indicating how many eigenvectors of (Y) to use in the CCA.

stFit object containing posterior samples for model

number of posterior samples to reject before computing estimates

function for computing point estimates

S3 generic/method consistency

stPredict object containing posterior estimates of alphas

stFit object containing posterior samples for model

stData object containing spatial information for dataset

number of posterior samples to reject before computing estimates

One of the following options to specify what point estimates to return

eof-alpha_knots

Remote coefficient estimates (alpha_knots) mapped onto the eof patterns of the remote covariates.

S3 generic/method consistency

(m x n) matrix

(p x q) matrix

(nq x r) matrix

[variable x observation] matrix of data for which to compute EOFs

TRUE/FALSE to center columns of X in call to prcomp

TRUE/FALSE to scale columns of X in call to prcomp

Data to save

Path/name to save data to

SpatialGridDataFrame containing data to extract

raster::extent object featuring region to extract, or a SpatialPolygonsXXX object used for extracting areal data

whether to return the raw data, anomalies (data minus temporal average at each location), standardized anomalies (anomalies divided by temporal standard deviation at each location), or spatially standardized data (data minus overall spatial average divided by spatial std. dev.; each year gets its own spatial standardization )

if provided, will spatially average the data

if an sgdf is provided, the data will be masked before extraction, aggregation, and anomaly computation

TRUE to return the aspect of the surface at each location instead of the value of the surface itself

if aspect==TRUE, this specifies the number of discrete categories to divide aspect numbers (0-360) into. NULL if the original scale (0-360) should be kept. By design, the aspect categories will be centered on north in the first category.

TRUE to return the slope of the surface at each location instead of the value of the surface itself

SpatialGridDataFrame with local covariates. If X is a list, each SpatialGridDataFrame will be included as one covariate.

SpatialGridDataFrame with response data

SpatialGridDataFrame with remote covariates. If Z is a list, this function assumes each element of the list contains observations for the same covariate, but from different spatial regions. If Z is a list, D.r and mask.r must also be lists so that this function can know which regions to extract from each SpatialGridDataFrame

Timepoint from which to extract data from X, Y, and Z. If NULL, then all timepoints will be used.

c(xmin, xmax, ymin, ymax) region from which to extract data from X and Y, or a SpatialPolygonsXXX object containing boundaries of regions to extract areal data from.

c(xmin, xmax, ymin, ymax) region from which to extract data from Z

SpatialGridDataFrame to be used as a mask when extracting data from X and Y. Locations in mask.s with NA values will be ignored when extracting data from X and Y.

SpatialGridDataFrame to be used as a mask when extracting data from Z. Locations in mask.s with NA values will be ignored when extracting data from Z.

If provided, will spatially average Y and X data

If provided, will spatially average Z data

If TRUE, an intercept will be added to the design matrix

'response' 'anomaly' or 'std.anomaly' or a vector of these options depending on whether data extracted from X should be the observed data, anomalies, or standardized anomalies (where the climatology is computed from the observations as the pointwise temporal average)

'response' 'anomaly' or 'std.anomaly' or a vector of these options depending on whether data extracted from Z should be the observed data, anomalies, or standardized anomalies (where the climatology is computed from the observations as the pointwise temporal average)

'response' 'anomaly' or 'std.anomaly' depending on whether data extracted from Y should be the observed data, anomalies, or standardized anomalies (where the climatology is computed from the observations as the pointwise temporal average)

name for X data (optional)

name for Y data (optional)

name for Z data (optional)

TRUE or vector of logicals (one for each X object) to return the aspect of the surface at each location instead of the value of the surface itself

if aspect==TRUE, this specifies the number of discrete categories to divide aspect numbers (0-360) into. NULL if the original scale (0-360) should be kept. By design, the aspect categories will be centered on north in the first category.

TRUE or vector of logicals (one for each X object) to return the slope of the surface at each location instead of the value of the surface itself

names of columns of X

formula object to specify how to create the design matrix

an m x n matrix

an p x q matrix

an mp x s matrix

an n x n scale matrix

degrees of freedom parameter

an m x n matrix

an p x q matrix

A numeric vector of distances at which the Matern correlation function should be evaluated.

Scales correlations to covariances.

Matern range parameter. Controls the decay of pointwise correlations as a function of distance.

Matern smoothness parameter. Controls the number of process derivatives.

Spatial covariance nugget.

A numeric vector or matrix of distances at which the Matern correlation function should be evaluated.

Scales correlations to covariances.

Matern range parameter. Controls the decay of pointwise correlations as a function of distance.

Matern smoothness parameter. Controls the number of process derivatives.

Spatial covariance nugget.

Effective correlation to check for

Matern range parameter. Controls the decay of pointwise correlations as a function of distance.

Matern smoothness parameter. Controls the number of process derivatives.

Raw output from one run of the Rcpp/Armadillo composition sampler

Raw output from another run of the Rcpp/Armadillo composition sampler

sample covariance matrix from the first sample, 'A'

sample covariance matrix from the second sample, 'B'

sample mean from the first sample, 'A'

sample mean from the first sample, 'A'

sample mean from the second sample, 'B'

sample mean from the second sample, 'B'

sample size from the first sample, 'A'

sample size from the second sample, 'B'

sample mean from the first sample, 'x'

sample mean from the second sample, 'y'

sample size from the first sample, 'x'

sample size from the second sample, 'y'

sample variance from the first sample, 'x'

sample variance from the second sample, 'y'

sample mean from the first sample, 'x'

sample mean from the second sample, 'y'

sample size from the first sample, 'x'

sample size from the second sample, 'y'

Object of class stData to plot.

One of the following options to specify what type of plot to build

response

sd.response

Plot standard deviation of response variable at each location.

cat.response

covariate

remote

teleconnection

This plot only applies if the stData object contains information about teleconnection effects, i.e., if it is a simulated dataset or otherwise modified to include estimates of teleconnection effects.

remote_cor

This plot shows pointwise correlations between a local coordinate and the remote covariates.

eof

eof_scores

eof_scree

eof_cor

This plot shows pointwise correlations with EOF patterns.

local_cor

This plot shows pointwise correlations with local covariates.

teleconnection_knot_local

timepoint to plot. Will automatically plot the first timepoint if t=NULL.

column index of local covariate to plot if type='covariate'. Will automatically assume the local covariate data includes an intercept and will plot the second column if p=NULL.

name of map provided by the maps package. These include county, france, italy, nz, state, usa, world, world2. By default, all stData plots will include us state outlines.

name of subregions to include. Defaults to . which includes all subregions. See documentation for map for more details.

if plot type is 'teleconnection', specifies the longitude and latitude of local coordinate for which to plot teleconnection effects. if NULL, the middle local coordinate will be plotted.

if plot type is 'teleconnection_local', specifes the longitude and latitude of remote coordinate for which to plot associated teleconnection effects. if NULL, the middle remote coordinate will be plotted.

c(min, max) vector that specifies the colorscale limits

Optional label to override the default fill scale labels

label used for fill scale in teleconnection plot

line width for fill scale label

[ncoords x ncats] list of breakpoints used for binning responses into categories

if plot type is 'remote', specifies the longitude and latitude of knot locations to overlay on the 'remote' plot

if TRUE, will highlight significant grid cells if the plotting data contain a signif column

additional named arguments with defaults to pass to additional functions

if type=='eof' this specifies which (remote) EOF pattern to plot or if type=='eof_scores' this (vector) specifies which (remote) EOF pattern scores to plot

line width for when plotting with signif.telecon==T

Used to denote where this proportion of variance is achieved in the eof_scree plots

significance level for eof_cor significance highlighting

the level of fading that should be applied to insignificant grid boxes when plotting significant effects

number that specifies the midpoint of the colorscale

c(TRUE, TRUE) to plot local and remote responses as contours vs. observations

additional arguments to pass to functions

Object of class stFit to plot.

One of the following options to specify what type of plot to build

traceplot

density

pairs

teleconnection

teleconnection_local

teleconnection_knot

teleconnection_knot_transect

teleconnection_knot_influence

beta

Object of class stData to provide coordinate and related information for plotting estimated teleconnection effects

if plot type is 'teleconnection', specifies the longitude and latitude of local coordinate for which to plot estimated teleconnection effects. if NULL, the middle local coordinate will be plotted.

if plot type is 'teleconnection_knot_influence' or 'teleconnection_knot_local', specifies the longitude and latitude of knot coordinate for which to plot influence of remote coefficient on remote covariates, or the teleconnection coefficients associated with coord.knot

number specifying the size of text labels

number specifying the size of axis text labels

number specifying the size of title

number of observations to exclude from graph

if TRUE, will highlight significant teleconnection effects when type=='teleconnection'

If stFit was fit with spatially varying coefficients, p specifies the index of the spatially varying coefficient to plot

data.frame with variables, 'lon.Y', 'lat.Y', 'x' that will be plotted against teleconnection effects if type=='teleconnection_knot_transect'

specifies linewidth for plots that include reference lines

number of significant figures to round facet latitudes and longitudes for if type=='teleconnection_knot_transect'

if type=='teleconnection_knot_transect' this specifies the bandwith of the non-parametric smooth of the estimates

if type=='teleconnection_knot_transect' this specifies the degree of the non-parametric smooth of the estimates

additional named arguments with defaults to pass to additional functions

additional arguments to pass to functions

Object of class stPredict to plot.

One of the following options to specify what type of plot to build

prediction

Spatial plot of predicted response variable for a given timepoint t.

residual

Spatial plot of residual for a given timepoint t. Note, this plot is only available if the model has been evaluated and the predictions have been compared to another response dataset.

observed

Spatial plot of observed response variable for a given timepoint t. Note, this plot is only available if the model has been evaluated and the predictions have been compared to another response dataset.

standard_error (or 'se')

Spatial plot of prediction standard errors for a given timepoint t.

local

Spatial plot of the local components of the response variable for a given timepoint t.

remote

Spatial plot of the remote components of the response variable for a given timepoint t.

w

Spatial plot of the spatial noise component of the reponse variable for a given timepoint t.

correlation

Scatterplot of observed vs. predicted response variables for a given timepoint t. Note, this plot is only available if the model has been evaluated and the predictions have been compared to another response dataset.

teleconnection

Spatial plot of remote coefficients associated with a location coord.s in the spatial response domain.

teleconnection_knot

Spatial plot of remote knot coefficients associated with a location coord.s in the spatial response domain.

teleconnection_knot_transect

errors

Series of plots that measure overall prediction error across prediction timepoints.

cat.prediction

Spatial plot of the predicted response variable category (i.e., above/below average) for a given timepoint t.

truth

Note, this plot is only available if the model has been evaluated and the predictions have been compared to another response dataset.

residual

Note, this plot is only available if the model has been evaluated and the predictions have been compared to another response dataset.

eof_alpha_knots

A map of the local domain where the plotted colors show the remote influence coefficients mapped onto the eof pattern specified by the "pattern" argument.

timepoint to plot. Will automatically plot the first timepoint if t=NULL.

Object of class stFit to provide related information and structures for plotting estimated teleconnection effects

Object of class stData to provide coordinate and related information for plotting estimated teleconnection effects

data.frame with Year column and a column for a variable that will be used to plot annual errors against

name of variable in err.comparison for plotting against

label for name of variable in err.comparison for plotting against

if type=='eof_alpha_knots', this specified which eof the remote coefficients should be mapped onto and then plotted over the local domain

additional named arguments with defaults to pass to additional functions

number of observations to exclude from graph

TRUE to highlight significant teleconnection effects

additional arguments to be passed to lower-level plotting functions

object of class teleCor, containing pointwise correlations

if TRUE, then teleCor must have a column labeled 'signif' that indicates which correlations are significant. These correlations will be printed in bold, and the rest will be printed more lightly

specifies the longitude and latitude of local coordinate for which to plot pointwise correlations (if type=='remote'). if NULL, the middle local coordinate will be plotted.

name of map provided by the maps package. These include county, france, italy, nz, state, usa, world, world2. By default, all stData plots will include us state outlines.

name of subregions to include. Defaults to . which includes all subregions. See documentation for map for more details.

c(min, max) vector that specifies the colorscale limits

additional named arguments with defaults to pass to additional functions

additional arguments to be passed to lower-level plotting functions

Number of random matrices to simulate

Covariance matrix defining dependence between rows

Covariance matrix defining dependence between columns

average value of each entry in the sampled matrices

symmetric positive definite p x p scale matrix

degrees of freedom (greater than p-1)

stPredict object containing predictions for Y

observed values of the response

the climatology for the location in Y

Object with class 'stData' containing data needed to fit this model. The data need only be manually entered if not using a stData object.

A list containing parameters for the prior distributions. The list needs to contain the following values

beta

list(Lambda=matrix) specifying the prior covariance matrix for the local effects if varying==F, otherwise list(Psi=matrix, nu=double) specifying the Inverse wishart prior distribution for the spatially varying coefficient process if varying==T.

cov.s

list(smoothness=double, range=c(min, max), variance=c(shape, rate), nugget=c(shape, rate))

cov.r

list(smoothness=double, range=c(min, max), variance=c(shape, rate), nugget=c(shape, rate))

number of iterations to run the MCMC chain for

[ns, p, nt] array of design matrices with local covariates

[ns, nt] matrix with response data

[nr, nt] matrix with remote covariates

matrix with coordinates where responses were observed (lon, lat)

matrix with coordinates where remote covariates were observed (lon, lat)

A list containing initial standard deviation parameters for the MCMC parameters requiring random walk updates

cov.s

list(range=double, nugget=double)

cov.r

list(range=double, variance=double, nugget=double)

TRUE to compute the model log-likelihood at each iteration

TRUE if covariance matrix distances should be in miles, FALSE for kilometers

scaling factor used in adapting random walk proposal variances.

target acceptance rate for random walk proposals.

TRUE to fit the model without the teleconnection effects (typically for evaluating impact of teleconnection effects)

(depreceated) TRUE to fit the model with spatially varying local coefficients

TRUE to fit the model without local effects. This will fit a local intercept, but will not incorporate local covariates.

matrix with coordinates where remote teleconnections will be based (lon, lat)

Object with class 'stData' containing data needed to fit this model. The data need only be manually entered if not using a stData object.

Object with class 'stFit' containing posterior parameter samples needed to composition sample the teleconnection effects and generate posterior predictions. The data needed from stFit need only be manually entered if not using a stData object.

values of \beta at which to evaluate the likelihood

values of \sigma^2_w at which to evaluate the likelihood

values of \sigma^2_\alpha at which to evaluate the likelihood

values of \sigma^2_\varepsilon at which to evaluate the likelihood

values of \rho_w at which to evaluate the likelihood

values of \rho_\alpha at which to evaluate the likelihood

[ns, p, nt] array of design matrices with local covariates

[ns, nt] matrix with response data

[nr, nt] matrix with remote covariates

matrix with coordinates where responses were observed (lon, lat)

matrix with coordinates where remote covariates were observed (lon, lat)

matrix with coordinates of knots for remote covariates (lon, lat)

TRUE if distances should be computed in miles (kilometers otherwise)

values of \sigma^2_{\alpha_\varepsilon} at which to evaluate the likelihood

Object with class 'stFit' containing posterior parameter samples needed to composition sample the teleconnection effects and generate posterior predictions. The data needed from stFit need only be manually entered if not using a stData object.

Object with class 'stData' containing data needed to fit this model. The data need only be manually entered if not using a stData object.

object of class stData that includes information needed for making forecasts. If response data is included, this function will automatically run stEval using the empirical climatology as the reference forecast

number of posterior samples to burn before drawing composition samples

confidence level for approximate confidence intervals of teleconnection effects (only needed if returnAlphas==TRUE)

Since the teleconnection effects and posterior predictions can be sampled in parallel, this parameter lets users specify the number of cores to use to draw teleconnection and prediction samples

Parameter specifying the HPD level to compute for posterior predictive samples

Forecast timepoint labels

[ns, p, nt] array of design matrices with local covariates

[ns, nt] matrix with response data

[nr, nt] matrix with remote covariates

[ns, p, nt0] array of design matrices with local covariates at forecast timepoints

[nr, nt0] matrix with remote covariates at forecast timepoints

matrix with coordinates where responses were observed (lon, lat)

matrix with coordinates where remote covariates were observed (lon, lat)

TRUE to return the teleconnection effects sampled at knot locations. Note that only basic summary information about the teleconnection effects will be returned.

vector of probabilities for also returning categorical predictions from the posterior prediction samples; NULL otherwise

TRUE to return the teleconnection effects. Note that only basic summary information about the teleconnection effects will be returned.

stData object with training data to simulate new Y values for

stData object with test data to simulate new Y values for

matrix with coordinates of knots for remote covariates (lon, lat)

A list containing model parameters for use in simulation

beta

vector with fixed effect coefficients

cov.s

list(smoothness=double, range=double, variance=double, nugget=double)

cov.r

list(smoothness=double, range=double, variance=double, nugget=double)

TRUE to compute distances for evaluating covariance functions in miles. This is important since the interpretations of the cov.r and cov.s parameters depend on the units with which distance is measured.

Object with class 'stData' containing data needed to fit this model.

Object with class 'stFit' containing posterior parameter samples needed to composition sample the teleconnection effects and generate posterior predictions.

number of posterior samples to burn before drawing composition samples

structure containing posterior inference for remote coefficients

confidence level for confidence intervals and significance

matrix with coordinates where responses were observed (lon, lat)

matrix with coordinates where remote covariates were observed (lon, lat)

structure containing posterior inference for transformed remote coefficients

confidence level for confidence intervals and significance

matrix with coordinates where responses were observed (lon, lat)

Object of class stPredict to summarise

timepoint to plot. Will automatically plot all timepoints and overall summary if NULL.

Number of digits to pass to signif, if not NULL.

S3 generic/method consistency

vector containing responses for each timepoint. vector is blocked by timepoint.

matrix containing local covariates for each timepoint. each row are the covariates for one location and timepoint. matrix is blocked by timepoint.

matrix containing remote covariates. each column has remote covariates for one timepoint.

n x 2 matrix containing lon-lat coordinates for locations.

T/F for whether to compute great circle distances in miles (T) or km (F)

A list containing parameters for the prior distributions. The list needs to contain the following values

T

list(Psi=matrix, nu=double) specifying the Inverse wishart prior distribution for the spatially varying coefficient process.

beta

list(Linv=matrix) specifying the prior precision matrix for the fixed local covariates.

sigmasq

list(a=double, b=double) specifying the prior shape and scale parameters for the covariance scale and nugget parameters.

rho

list(L=double, U=double) specifying the lower and upper bounds for the spatial range parameter.

cov

list(nu=double) specifying the smoothness for the matern covariance.

number of MCMC iterations to run

MCMC thinning; defaults to no thinning (thin=1)

Initial proposal standard deviation for RW samplers

optional list containing starting parameters for MCMC sampler

scaling factor used in adapting random walk proposal variances.

target acceptance rate for random walk proposals.

svcFit object containing posterior samples

[nr*nt, p] matrix of local covariates at new timepoint

[nr, nt] matrix of remote covariates at new timepoints

Object with class 'stData' containing data needed to fit this model. The data is used to compute empirical quantiles for making categorical predictions.

object of class stData that includes information needed for making forecasts.

number of posterior samples to burn from fit

vector of probabilities for also returning categorical predictions from the posterior prediction samples; NULL otherwise

Parameter specifying the HPD level to compute for posterior predictive samples

stData object containing data to analyze

[ny x nt] a matrix composed of ny row vectors, each of which contains nt observations fom a different spatial location. Spatial locations for Y are indexed by coords.s.

[nz x nt] a matrix composed of nz row vectors each of which contains nt observations from a different spatial location. Spatial locations for Z are indexed by coords.r.

coordinates of locations in Y

coordinates of locations in Z