Final minor changes for v0.1.2

DESCRIPTION

@@ -2,7 +2,7 @@ Package: FRK
 Type: Package
 Title: Fixed Rank Kriging
 Version: 0.1.2
-Date: 2017-03-06
+Date: 2017-04-21
 Authors@R: c(
     person("Andrew", "Zammit-Mangion", , "andrewzm@gmail.com", c("aut", "cre")),
     person("Timothy", "Davis", , "davis@tamu.edu", role = "ctb"),
@@ -13,15 +13,14 @@ Maintainer: Andrew Zammit-Mangion <andrewzm@gmail.com>
 VignetteBuilder: knitr
 Description: Fixed Rank Kriging is a tool for spatial/spatio-temporal modelling
     and prediction with large datasets. The approach, discussed in Cressie and
-    Johannesson (2008), decomposes the field, and hence the covariance function,
-    using a fixed set of n basis functions, where n is typically much smaller than
-    the number of data points (or polygons) m. The method naturally allows for
-    non-stationary, anisotropic covariance functions and the use of observations
-    with varying support (with known error variance). The projected field is a key
-    building block of the Spatial Random Effects (SRE) model, on which this package
-    is based. The package FRK provides helper functions to model, fit, and predict
-    using an SRE with relative ease. Reference: Cressie, N. and Johannesson, G.
-    (2008) <DOI:10.1111/j.1467-9868.2007.00633.x>.
+    Johannesson (2008) <DOI:10.1111/j.1467-9868.2007.00633.x>, decomposes the field, 
+    and hence the covariance function, using a fixed set of n basis functions, 
+    where n is typically much smaller than the number of data points (or polygons) m. 
+    The method naturally allows for non-stationary, anisotropic covariance functions 
+    and the use of observations with varying support (with known error variance). The 
+    projected field is a key building block of the Spatial Random Effects (SRE) model, 
+    on which this package is based. The package FRK provides helper functions to model, 
+    fit, and predict using an SRE with relative ease.
 BugReports: http://github.com/andrewzm/FRK/issues
 Depends:
     R (>= 3.1)

NAMESPACE

@@ -58,6 +58,7 @@ importFrom(plyr,rbind.fill)
 importFrom(stats,.getXlevels)
 importFrom(stats,coefficients)
 importFrom(stats,dist)
+importFrom(stats,formula)
 importFrom(stats,kmeans)
 importFrom(stats,lm)
 importFrom(stats,median)

R/FRK.R

@@ -16,6 +16,6 @@
 #' @importFrom digest digest
 #' @importFrom Rcpp cppFunction
 #' @importFrom grDevices chull
-#' @importFrom stats .getXlevels coefficients dist kmeans lm median model.extract model.frame model.matrix na.fail optim runif sd terms var time rnorm
+#' @importFrom stats .getXlevels coefficients dist kmeans lm median model.extract model.frame model.matrix na.fail optim runif sd terms var time rnorm formula
 #' @importFrom utils data
 NULL

R/geometryfns.R

@@ -288,7 +288,6 @@ setMethod("coordnames",signature(x="STIDF"),function(x) {
 #' @param tunit temporal unit when requiring space-time BAUs. Can be either "secs", "mins", "hours" or "days".
 #' @param xlims limits of the horizontal axis (overrides automatic selection).
 #' @param ylims limits of the vertical axis (overrides automatic selection).
-#' @param tunit temporal unit when requiring space-time BAUs. Can be either "secs", "mins", "hours" or "days".
 #' @param ... currently unused
 #' @details \code{auto_BAUs} constructs a set of Basic Areal Units (BAUs) used both for data pre-processing and for prediction. As such, the BAUs need to be of sufficienly fine resolution so that inferences are not affected due to binning.
 #'

cran-comments.md

@@ -1,4 +1,12 @@
-## Since first attempt to upload on CRAN
+## v0.1.2
+* Now code is heavily documented and manual has been made more precise
+* Made lik computation more efficient
+* Made interpolation near origin default variogram fitting choice (linear is second, exp is third)
+* Added option for user to specify xlims and ylims of BAUs 
+* Removed timeline object (was redundant)
+* Internally forced all time objects to be POSIXct
+
+## v0.1.1 Since first attempt to upload on CRAN
 * Native routines are now registered in FRK-init.c
 * vignettes now do not include packages that are not on CRAN and re-building of vignettes should be successful
 * all copyright is now acknowledged (checked using grep in all folders)