more fixes for check and bioccheck

DESCRIPTION

@@ -5,10 +5,15 @@ Title: Weighted Sliced Inverse Regression (wSIR) for supervised
         gene expression data
 Version: 0.1.3
 Authors@R: c(
-    person("Max", "Woollard", , "[email protected]", role = c("aut", "cre", "ctb")),
-    person("Pratibha", "Panwar", role = c("ctb")),
-    person("Linh", "Nghiem", role = c("aut","ctb")),
-    person("Shila", "Ghazanfar", , "[email protected]", role = c("aut", "ctb"))
+    person("Max", "Woollard", email = "[email protected]", 
+    role = c("aut", "cre", "ctb"),
+    comment=c(ORCID="0009-0000-6319-6926")),
+    person("Pratibha", "Panwar", role = c("ctb"), 
+    comment = c(ORCID = "0000-0002-7437-7084")),
+    person("Linh", "Nghiem", role = c("aut","ctb"),
+    comment=c(ORCID = "0000-0003-2874-9067")),
+    person("Shila", "Ghazanfar", email = "[email protected]", 
+    role = c("aut", "ctb"), comment=c(ORCID="0000-0001-7861-6997"))
     )
 Description: Weighted Sliced Inverse Regression (wSIR) is a supervised dimension 
     reduction algorithm for spatial transcriptomics gene expression data. For a 
@@ -23,10 +28,16 @@ License: GPL-2
 Encoding: UTF-8
 URL: https://sydneybiox.github.io/wSIR
 BugReports: https://github.com/sydneybiox/wSIR/issues
-biocViews: DimensionReduction, Software
-LazyData: true
+biocViews: DimensionReduction,
+    Software,
+    GeneExpression, 
+    SingleCell,
+    Spatial,
+    Transcriptomics,
+    Regression
+LazyData: false
 Roxygen: list(markdown = TRUE)
-Depends: R (>= 4.3.0),
+Depends: R (>= 4.4.0),
 Imports: Rfast,
     magrittr,
     ggplot2,

NAMESPACE

@@ -10,6 +10,7 @@ export(wSIR)
 importFrom(BiocParallel,MulticoreParam)
 importFrom(BiocParallel,bplapply)
 importFrom(Rcpp,evalCpp)
+importFrom(Rfast,dcor)
 importFrom(distances,distances)
 importFrom(doBy,which.maxn)
 importFrom(ggplot2,aes)

NEWS.md

@@ -0,0 +1,2 @@
+# wSIR v0.1.4 (2024-09-19)
++ Submitted to Bioconductor.

R/createWeightMatrix.R

@@ -1,20 +1,30 @@
 #' createWeightMatrix
 #'
 #' @description
-#' A function to create the weight matrix given the location of the cells, tile allocations and desired spatial weighting strength. Weight matrix entries represent level of spatial correlation between all pairs of tiles.
+#' A function to create the weight matrix given the location of the cells,
+#' tile allocations and desired spatial weighting strength. Weight matrix
+#' entries represent level of spatial correlation between all pairs of tiles.
 #'
-#' @param coords dataframe of dimension n * 2. Column names c("x", "y"). Spatial position of each cell.
-#' @param labels dataframe of dimension n * 1, column name c("coordinate"). Tile allocation of each cell. This is automatically created in the wSIR function.
-#' @param alpha numeric value giving strength of spatial weighting matrix. alpha = 0 returns identity matrix and equals SIR. Large alpha values tend all entries towards 1. Default is 4.
+#' @param coords dataframe of dimension n * 2. Column names c("x", "y").
+#' Spatial position of each cell.
+#' @param labels dataframe of dimension n * 1, column name c("coordinate").
+#' Tile allocation of each cell. This is automatically created in the wSIR
+#' function.
+#' @param alpha numeric value giving strength of spatial weighting matrix.
+#' alpha = 0 returns identity matrix and equals SIR. Large alpha values tend
+#' all entries towards 1. Default is 4.
 #'
-#' @return matrix containing the weight value for all pairs of tiles. Each value is between 0 and 1, with 1 always on the diagonal.
+#' @return matrix containing the weight value for all pairs of tiles. Each
+#' value is between 0 and 1, with 1 always on the diagonal.
 #'
 #'
 #'
 #' @keywords internal
 
 createWeightMatrix <- function(coords, labels, alpha = 4) {
-  alpha = 4/alpha # alpha_old = 0 (function argument = 0) gives alpha_new = Inf (rather than undefined) which equals SIR
+  alpha <- 4/alpha
+  # alpha_old = 0 (function argument = 0) gives
+  # alpha_new = Inf (rather than undefined) which equals SIR
 
   avg_coords <- slicerCategorical(coords, labels)
 
@@ -23,10 +33,10 @@ createWeightMatrix <- function(coords, labels, alpha = 4) {
   dist_mat <- as.matrix(stats::dist(avg_coords, diag = TRUE, upper = TRUE))
 
   # mask out the slices that are far away from each other
-  D2 = as.matrix(stats::dist(avg_coords_groups, method = "manhattan"))
+  D2 <- as.matrix(stats::dist(avg_coords_groups, method = "manhattan"))
   D2[D2 > 0] <- Inf
 
-  dist_mat_new = dist_mat + D2
+  dist_mat_new <- dist_mat + D2
 
   dist_norm <- (1 - dist_mat_new / max(dist_mat, na.rm = TRUE))^alpha
   weight_mat <- dist_norm

R/data.R

@@ -1,7 +1,8 @@
 #' MouseGastrulationData
 #'
 #' Data set consists of spatial transcriptomics data from a mouse embryo.
-#' There are three samples, for each we have gene expression data (351 genes), spatial
+#' There are three samples, for each we have gene expression data (351 genes),
+#' spatial
 #' coordinates on the two-dimensional spatial plane, and cell type labels.
 #' Sample 1 contains 19451 cells, sample 2 contains 14891 cells and sample 3
 #' contains 23194 cells. We have randomly sampled 20% of the cells from
@@ -15,19 +16,31 @@
 #' sample2_exprs sample2_coords sample2_cell_types
 #' sample3_exprs sample3_coords sample3_cell_types
 #' @docType data
-#' @format \code{sample1_exprs} has a row for each cell in sample 1 and a column for the
-#' expression level of each gene. \code{sample1_coords} has a row for each cell in sample 1
-#' and a column for its position in each of the two spatial axes. \code{sample1_cell_types}
-#' a vector whose i'th entry contains the cell type of the i'th cell of sample 1.
-#' \code{sample2_exprs} has a row for each cell in sample 2 and a column for the expression
-#' level of each gene. \code{sample2_coords} has a row for each cell in sample 2 and a column
-#' for its position in each of the two spatial axes. \code{sample2_cell_types} a vector whose
-#' i'th entry contains the cell type of the i'th cell of sample 2.  \code{sample3_exprs} has
-#' a row for each cell in sample 3 and a column for the expression level of each gene.
-#' \code{sample3_coords} has a row for each cell in sample 3 and a column for its position
-#' in each of the two spatial axes. \code{sample3_cell_types} a vector, whose i'th entry
+#' @format \code{sample1_exprs} has a row for each cell in sample 1 and a
+#' column for the
+#' expression level of each gene. \code{sample1_coords} has a row for each
+#' cell in sample 1
+#' and a column for its position in each of the two spatial axes.
+#' \code{sample1_cell_types}
+#' a vector whose i'th entry contains the cell type of the i'th cell of sample
+#' 1.
+#' \code{sample2_exprs} has a row for each cell in sample 2 and a column
+#' for the expression
+#' level of each gene. \code{sample2_coords} has a row for each cell in
+#' sample 2 and a column
+#' for its position in each of the two spatial axes. \code{sample2_cell_types}
+#' a vector whose
+#' i'th entry contains the cell type of the i'th cell of sample 2.
+#' \code{sample3_exprs} has
+#' a row for each cell in sample 3 and a column for the expression level of
+#' each gene.
+#' \code{sample3_coords} has a row for each cell in sample 3 and a column
+#' for its position
+#' in each of the two spatial axes. \code{sample3_cell_types} a vector,
+#' whose i'th entry
 #' contains the cell type of the i'th cell of sample 3.
-#' @source Integration of spatial and single-cell transcriptomic data elucidates mouse
+#' @source Integration of spatial and single-cell transcriptomic data
+#' elucidates mouse
 #' organogenesis, \emph{Nature Biotechnology}, 2022.  Webpage:
 #' \url{https://www.nature.com/articles/s41587-021-01006-2}
 #' @keywords datasets