Switch to Rmd :( (#57)

.github/workflows/check-bioc.yml

@@ -83,12 +83,10 @@ jobs:
       - name: Checkout Repository
         uses: actions/checkout@v2
 
-      ## Setup Python and nbconvert
-      - name: Setup Python for nbconvertR
-        uses: actions/setup-python@v2
-        with:
-          python-version: '3.x'
-      - run: pip install nbconvert
+      ## Necessary to do plotting with the default quartz device on macOS
+      - name: Install XQuartz
+        if: runner.os == 'macOS'
+        run: brew install --cask xquartz
 
       ## R is already included in the Bioconductor docker images
       - name: Setup R from r-lib

.gitignore

@@ -4,7 +4,6 @@
 *.dll
 *.pdf
 *.tar.gz
-.ipynb_checkpoints
 .Rproj.user
 *.Rcheck
 __pycache__/
@@ -15,25 +14,6 @@ __pycache__/
 /update_remotes.sh
 /build/
 /playground/
-/inst/doc/.build.timestamp
-/vignettes/.build.timestamp
-/inst/doc/*.aux
-/vignettes/*.aux
-/inst/doc/*.bbl
-/vignettes/*.bbl
-/inst/doc/*.blg
-/vignettes/*.blg
-/inst/doc/*.log
-/vignettes/*.log
-/inst/doc/*.out
-/vignettes/*.out
-/inst/doc/*.tex
-/vignettes/*.tex
-/inst/doc/*.toc
-/vignettes/*.toc
-/inst/doc/*.fls
-/vignettes/*.fls
-/inst/doc/*.html
 /vignettes/*.html
-/inst/doc/*_files/
+/vignettes/*.R
 /vignettes/*_files/

DESCRIPTION

@@ -52,16 +52,26 @@ Imports:
     scales,
     scatterplot3d
 LinkingTo: Rcpp, RcppEigen, grDevices
-SystemRequirements: C++11, jupyter nbconvert (see nbconvertR’s INSTALL file)
+SystemRequirements: C++11
 NeedsCompilation: yes
 Enhances: rgl, SingleCellExperiment
 Suggests:
-    nbconvertR (>= 1.3.2),
+    knitr,
+    rmarkdown,
     igraph,
     testthat,
     FNN,
-    tidyr
-VignetteBuilder: nbconvertR
+    tidyverse,
+    gridExtra,
+    cowplot,
+    conflicted,
+    viridis,
+    rgl,
+    scRNAseq,
+    org.Mm.eg.db,
+    scran,
+    repr
+VignetteBuilder: knitr
 biocViews: CellBiology,
     CellBasedAssays,
     Clustering,

NAMESPACE

@@ -1,5 +1,7 @@
 # Generated by roxygen2: do not edit by hand
 
+S3method(fortify,DPT)
+S3method(fortify,DiffusionMap)
 export("dataset<-")
 export("distance<-")
 export("eigenvalues<-")
@@ -144,6 +146,7 @@ importFrom(ggplot2,element_line)
 importFrom(ggplot2,element_rect)
 importFrom(ggplot2,element_text)
 importFrom(ggplot2,facet_wrap)
+importFrom(ggplot2,fortify)
 importFrom(ggplot2,geom_path)
 importFrom(ggplot2,geom_point)
 importFrom(ggplot2,geom_spoke)

R/dpt-plotting.r

@@ -109,7 +109,7 @@ plot.DPT <- function(
 		dpt@dm, dcs,
 		plot_more = plot_paths,
 		legend_main = legend_main,
-		col = col, legend_name = col_by,
+		col = col,
 		...)
 
 	if (!identical(Sys.getenv('LOG_LEVEL'), '')) message('Args:\n', paste(capture.output(print(args)), collapse = '\n'))

R/methods-coercion.r

@@ -58,7 +58,9 @@ setMethod('as.data.frame', 'DiffusionMap', function(x, row.names = NULL, optiona
 #' 
 #' @importFrom BiocGenerics as.data.frame
 #' @importFrom Biobase as.data.frame.ExpressionSet
+#' @importFrom ggplot2 fortify
 #' @rdname coercions
+#' @exportS3Method fortify
 #' @export fortify.DiffusionMap
 fortify.DiffusionMap <- function(model, data, ...) as.data.frame(model, ...)
 setAs('DiffusionMap', 'data.frame', function(from) as.data.frame(from))
@@ -79,7 +81,9 @@ setMethod('as.data.frame', 'DPT', function(x, row.names = NULL, optional = FALSE
 #' 
 #' @importFrom BiocGenerics as.data.frame
 #' @importFrom Biobase as.data.frame.ExpressionSet
+#' @importFrom ggplot2 fortify
 #' @rdname coercions
+#' @exportS3Method fortify
 #' @export fortify.DPT
 fortify.DPT <- function(model, data, ...) as.data.frame(model, ...)
 setAs('DPT', 'data.frame', function(from) as.data.frame(from))

tests/testthat/test_dataset_types.r

@@ -9,8 +9,19 @@ test_nfeat <- 3L
 test_obss <- paste0('c', seq_len(test_nobss))
 test_feat <- paste0('g', seq_len(test_nfeat))
 
-test_matrix <- matrix(runif(test_nobss * test_nfeat), test_nobss, test_nfeat, dimnames = list(test_obss, test_feat))
-test_matrix[test_matrix < mean(test_matrix)] <- 0
+mk_test_matrix <- function() {
+	m <- matrix(runif(test_nobss * test_nfeat), test_nobss, test_nfeat, dimnames = list(test_obss, test_feat))
+	m[m < mean(m)] <- 0
+	m
+}
+
+test_matrix <- mk_test_matrix()
+for (attempt in seq_len(100L)) {
+	if (!any(duplicated(test_matrix))) break
+	test_matrix <- mk_test_matrix()
+}
+if (any(duplicated(test_matrix))) stop('Could not create sensible test matrix after 100 attempts')
+
 test_df     <- data.frame(test_matrix, cm1 = LETTERS[seq_len(test_nobss)], stringsAsFactors = FALSE)
 test_es     <- ExpressionSet(
 	t(test_matrix),

vignettes/DPT.Rmd

@@ -0,0 +1,90 @@
+---
+title: destiny 2.0 brought the Diffusion Pseudo Time (DPT) class
+output: rmarkdown::html_vignette
+bibliography: bibliography.bib
+vignette: >
+  %\VignetteIndexEntry{destiny 2.0 brought the Diffusion Pseudo Time (DPT) class}
+  %\VignetteEngine{knitr::rmarkdown}
+  %\VignetteEncoding{UTF-8}
+---
+
+```{r}
+set.seed(1)
+```
+
+Diffusion Pseudo Time (DPT) is a pseudo time metric based on the transition probability of a diffusion process [@haghverdi_diffusion_2016].
+
+*destiny*  supports `DPT` in addition to its primary function of creating `DiffusionMap`s from data.
+
+```{r}
+library(destiny)    # load destiny…
+data(guo)           # …and sample data
+library(gridExtra)  # Also we need grid.arrange
+```
+
+`DPT` is in practice independent of Diffusion Maps:
+
+```{r}
+par(mar = rep(0, 4))
+graph <- igraph::graph_from_literal(
+    data -+ 'transition probabilities' -+ DiffusionMap,
+    'transition probabilities' -+ DPT)
+plot(
+    graph, layout = igraph::layout_as_tree,
+    vertex.size = 50,
+    vertex.color = 'transparent',
+    vertex.frame.color = 'transparent',
+    vertex.label.color = 'black')
+```
+
+However in order not to overcomplicate things, in *destiny*, you have to create `DPT` objects from `DiffusionMap` objects.  
+<small>(If you really only need the DPT, skip Diffusion Component creation by specifying `n_eigs = 0`)</small>
+
+```{r}
+dm <- DiffusionMap(guo)
+dpt <- DPT(dm)
+```
+
+The resulting object of a call like this will have three automatically chosen tip cells. You can also specify tip cells:
+
+```{r}
+set.seed(4)
+dpt_random <- DPT(dm, tips = sample(ncol(guo), 3L))
+```
+
+Plotting without parameters results in the DPT of the first root cell:
+
+TODO: wide plot
+
+```{r}
+grid.arrange(plot(dpt), plot(dpt_random), ncol = 2)
+```
+
+Other possibilities include the DPT from the other tips or everything supported by `plot.DiffusionMap`:
+
+TODO: wide plot
+
+```{r}
+grid.arrange(
+    plot(dpt, col_by = 'DPT3'),
+    plot(dpt, col_by = 'Gata4', pal = viridis::magma),
+    ncol = 2
+)
+```
+
+The `DPT` object also contains a clustering based on the tip cells and DPT, and you can specify where to draw paths from and to:
+
+```{r}
+plot(dpt, root = 2, paths_to = c(1,3), col_by = 'branch')
+```
+
+You can further divide branches. First simply plot branch colors like we did above, then identify the number of the branch you intend to plot, and then specify it in a subsequent `plot` call. In order to see the new branches best, we specify a `dcs` argument that visually spreads out out all four branches.
+
+```{r}
+plot(dpt, col_by = 'branch', divide = 3, dcs = c(-1,-3,2), pch = 20)
+```
+
+References
+==========
+
+<div id="refs"></div>

vignettes/Diffusion-Map-recap.Rmd

@@ -0,0 +1,43 @@
+---
+title: Reproduce the Diffusion Map vignette with the supplied data()
+output: rmarkdown::html_vignette
+vignette: >
+  %\VignetteIndexEntry{Reproduce the Diffusion Map vignette with the supplied data()}
+  %\VignetteEngine{knitr::rmarkdown}
+  %\VignetteEncoding{UTF-8}
+---
+
+Quickstart
+==========
+
+A short version to achive the above is by using the preprocessed version of the dataset provided with this package. `data(guo)` is already preprocessed (using the method first mentioned), has its threshold set to a constant 15 and is ready to use. Since the platform’s maximum amplification cycles are 40, that number can be used as upper border of the uncertainty range.
+
+```{r}
+library(destiny)
+data(guo)
+```
+
+It can be used directly for diffusion map creation:
+
+```{r}
+dm_guo <- DiffusionMap(guo, verbose = FALSE,
+                       censor_val = 15, censor_range = c(15, 40))
+dm_guo
+```
+
+```{r}
+plot(dm_guo)
+```
+
+using the annotation shows that the approximation worked
+
+```{r}
+palette(cube_helix(6))
+plot(dm_guo, col_by = 'num_cells',
+     legend_main = 'Cell stage')
+```
+
+References
+==========
+
+<div id="refs"></div>