campbio / celda / 758

pending completion

Build # 758

Build Type

Pull #291

travis-ci

Committed by

web-flow

Commit Message

<a href="https://github.com/campbio/celda/commit/<a class=hub.com/campbio/celda/commit/<a class="double-link" href="https://git"><a class=hub.com/campbio/celda/commit/<a class="double-link" href="https://git"><a class=hub.com/campbio/celda/commit/<a class="double-link" href="https://git"><a class=hub.com/campbio/celda/commit/bf0394aed8d7dc22f16f9569fed81507df300de7">bf0394aed">&lt;a href=&quot;https://github.com/campbio/celda/commit/</a><a class="double-link" href="https://github.com/campbio/celda/commit/&lt;a class=&quot;double-link&quot; href=&quot;https://git">&lt;a class=</a>hub.com/campbio/celda/commit/&lt;a class=&quot;double-link&quot; href=&quot;https://git">&lt;a class=</a>hub.com/campbio/celda/commit/&lt;a class=&quot;double-link&quot; href=&quot;https://git">&lt;a class=</a>hub.com/campbio/celda/commit/&lt;a class=&quot;double-link&quot; href=&quot;https://git">&lt;a class=</a>hub.com/campbio/celda/commit/&lt;a class=&quot;double-link&quot; href=&quot;https://git">&lt;a class=</a>hub.com/campbio/celda/commit/bf0394aed8d7dc22f16f9569fed81507df300de7">bf0394aed</a><a href="https://github.com/campbio/celda/commit/bf0394aed8d7dc22f16f9569fed81507df300de7">&quot;&gt;&amp;lt;a href=&amp;quot;https://github.com/campbio/celda/commit/&lt;/a&gt;&lt;a class=&quot;double-link&quot; href=&quot;https://github.com/campbio/celda/commit/&amp;lt;a class=&amp;quot;double-link&amp;quot; href=&amp;quot;https://git&quot;&gt;&amp;lt;a class=&lt;/a&gt;hub.com/campbio/celda/commit/&amp;lt;a class=&amp;quot;double-link&amp;quot; href=&amp;quot;https://git&quot;&gt;&amp;lt;a class=&lt;/a&gt;hub.com/campbio/celda/commit/&amp;lt;a class=&amp;quot;double-link&amp;quot; href=&amp;quot;https://git&quot;&gt;&amp;lt;a class=&lt;/a&gt;hub.com/campbio/celda/commit/&amp;lt;a class=&amp;quot;double-link&amp;quot; href=&amp;quot;https://git&quot;&gt;&amp;lt;a class=&lt;/a&gt;hub.com/campbio/celda/commit/&amp;lt;a class=&amp;quot;double-link&amp;quot; href=&amp;quot;https://git&quot;&gt;&amp;lt;a class=&lt;/a&gt;hub.com/campbio/celda/commit/bf0394aed8d7dc22f16f9569fed81507df300de7&quot;&gt;bf0394aed&lt;/a&gt;&lt;a href=&quot;https://github.com/campbio/celda/commit/bf0394aed8d7dc22f16f9569fed81507df300de7&quot;&gt;&amp;quot;&amp;gt;&amp;amp;lt;a href=&amp;amp;quot... (continued)

Pull Request Pull Request #291: sce

Run Details

168 of 168 new or added lines in 9 files covered. (100.0%)

2418 of 11069 relevant lines covered (21.84%)

3806.1 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

0.0

/R/splitModule.R

#' @title Split celda feature module
#' @description Manually select a celda feature module to split into 2 or
#'  more modules. Useful for splitting up modules that show divergent
#'  expression of features in multiple cell clusters.
#' @param x A \linkS4class{SingleCellExperiment} object
#'  with the matrix located in the assay slot under \code{useAssay}.
#'  Rows represent features and columns represent cells.
#' @param useAssay A string specifying which \link[SummarizedExperiment]{assay}
#'  slot to use for \code{x}. Default "counts".
#' @param module Integer. The module to be split.
#' @param n Integer. How many modules should \code{module} be split into.
#'  Default 2.
#' @param seed Integer. Passed to \link[withr]{with_seed}. For reproducibility,
#'  a default value of 12345 is used. If NULL, no calls to
#'  \link[withr]{with_seed} are made.
#' @return A updated \linkS4class{SingleCellExperiment} object with new
#'  feature modules stored in column \code{celda_feature_module} in
#'  \code{\link[SummarizedExperiment]{rowData}(x)}.
#' @export
setGeneric("splitModule",
    function(x, ...) {
        standardGeneric("splitModule")
    })


#' @rdname splitModule
#' @examples
#' data(sceCeldaCG)
#' # Split module 5 into 2 new modules.
#' sce <- splitModule(sceCeldaCG, module = 5)
#' @export
setMethod("splitModule", signature(x = "SingleCellExperiment"),
    function(x,
        useAssay = "counts",
        module,
        n = 2,
        seed = 12345) {

        if (!module %in% celdaModules(x)) {
            stop("Module ", module, " is not found in celdaModules(x).",
                " Please specify a valid module.")
        }

        celdaGMod <- .splitModuleWithSeed(x = x,
            useAssay = useAssay,
            module = module,
            n = n,
            seed = seed)

        S4Vectors::metadata(x)[["celda_parameters"]]$L <- params(celdaGMod)$L
        S4Vectors::metadata(x)[["celda_parameters"]]$finalLogLik <-
            celdaGMod@finalLogLik
        S4Vectors::metadata(x)[["celda_parameters"]]$featureModuleLevels <-
            sort(unique(celdaClusters(celdaGMod)$y))
        SummarizedExperiment::rowData(x)["celda_feature_module"] <-
            celdaClusters(celdaGMod)$y
        return(x)
    }
)


.splitModuleWithSeed <- function(x,
    useAssay,
    module,
    n,
    seed) {

    if (is.null(seed)) {
        celdaGMod <- .splitModule(x, useAssay, module, n)
    } else {
        with_seed(seed, celdaGMod <- .splitModule(x, useAssay, module, n))
    }
    return(celdaGMod)
}


.splitModule <- function(x, useAssay, module, n) {
    counts <- SummarizedExperiment::assay(x, i = useAssay)
    .validateCounts(counts)
    counts <- as.matrix(counts)
    ix <- celdaModules(x) == module

    if (sum(ix) > 1) {
        tempModel <- .celda_G(
            counts = counts[ix, , drop = FALSE],
            L = n,
            yInitialize = "random",
            splitOnIter = -1,
            splitOnLast = FALSE,
            nchains = 1,
            verbose = FALSE)

        splitY <- celdaClusters(tempModel)$y
        splitIx <- celdaClusters(tempModel)$y > 1
        splitY[splitIx] <- S4Vectors::metadata(x)$celda_parameters$L +
            splitY[splitIx] - 1
        splitY[!splitIx] <- module

        newY <- celdaModules(x)
        newY[ix] <- splitY
        newL <- max(newY)

        newLl <- .logLikelihoodcelda_G(
            counts = counts,
            y = newY,
            L = newL,
            beta = S4Vectors::metadata(x)$celda_parameters$beta,
            delta = S4Vectors::metadata(x)$celda_parameters$delta,
            gamma = S4Vectors::metadata(x)$celda_parameters$gamma)
        model <- methods::new(
            "celda_G",
            clusters = list(y = newY),
            params = list(
                L = newL,
                beta = S4Vectors::metadata(x)$celda_parameters$beta,
                delta = S4Vectors::metadata(x)$celda_parameters$delta,
                gamma = S4Vectors::metadata(x)$celda_parameters$gamma,
                countChecksum = .createCountChecksum(counts)
            ),
            names = list(row = rownames(x),
                column = colnames(x),
                sample = x@metadata$celda_parameters$sampleLevels),
            finalLogLik = newLl
        )
    } else {
        stop("Module ", module, "contains <= 1 feature. No additional",
            " splitting was able to be performed.")
    }
    return(model)
}

1	#' @title Split celda feature module
2	#' @description Manually select a celda feature module to split into 2 or
3	#' more modules. Useful for splitting up modules that show divergent
4	#' expression of features in multiple cell clusters.
5	#' @param x A \linkS4class{SingleCellExperiment} object
6	#' with the matrix located in the assay slot under \code{useAssay}.
7	#' Rows represent features and columns represent cells.
8	#' @param useAssay A string specifying which \link[SummarizedExperiment]{assay}
9	#' slot to use for \code{x}. Default "counts".
10	#' @param module Integer. The module to be split.
11	#' @param n Integer. How many modules should \code{module} be split into.
12	#' Default 2.
13	#' @param seed Integer. Passed to \link[withr]{with_seed}. For reproducibility,
14	#' a default value of 12345 is used. If NULL, no calls to
15	#' \link[withr]{with_seed} are made.
16	#' @return A updated \linkS4class{SingleCellExperiment} object with new
17	#' feature modules stored in column \code{celda_feature_module} in
18	#' \code{\link[SummarizedExperiment]{rowData}(x)}.
19	#' @export
20	setGeneric("splitModule",
21	function(x, ...) {
22	standardGeneric("splitModule")	×
23	})
24
25
26	#' @rdname splitModule
27	#' @examples
28	#' data(sceCeldaCG)
29	#' # Split module 5 into 2 new modules.
30	#' sce <- splitModule(sceCeldaCG, module = 5)
31	#' @export
32	setMethod("splitModule", signature(x = "SingleCellExperiment"),
33	function(x,
34	useAssay = "counts",
35	module,
36	n = 2,
37	seed = 12345) {
38
39	if (!module %in% celdaModules(x)) {	×
40	stop("Module ", module, " is not found in celdaModules(x).",	×
41	" Please specify a valid module.")	×
42	}
43
44	celdaGMod <- .splitModuleWithSeed(x = x,	×
45	useAssay = useAssay,	×
46	module = module,	×
47	n = n,	×
48	seed = seed)	×
49
50	S4Vectors::metadata(x)[["celda_parameters"]]$L <- params(celdaGMod)$L	×
51	S4Vectors::metadata(x)[["celda_parameters"]]$finalLogLik <-	×
52	celdaGMod@finalLogLik	×
53	S4Vectors::metadata(x)[["celda_parameters"]]$featureModuleLevels <-	×
54	sort(unique(celdaClusters(celdaGMod)$y))	×
55	SummarizedExperiment::rowData(x)["celda_feature_module"] <-	×
56	celdaClusters(celdaGMod)$y	×
57	return(x)	×
58	}
59	)
60
61
62	.splitModuleWithSeed <- function(x,
63	useAssay,
64	module,
65	n,
66	seed) {
67
68	if (is.null(seed)) {	×
69	celdaGMod <- .splitModule(x, useAssay, module, n)	×
70	} else {
71	with_seed(seed, celdaGMod <- .splitModule(x, useAssay, module, n))	×
72	}
73	return(celdaGMod)	×
74	}
75
76
77	.splitModule <- function(x, useAssay, module, n) {
78	counts <- SummarizedExperiment::assay(x, i = useAssay)	×
79	.validateCounts(counts)	×
80	counts <- as.matrix(counts)	×
81	ix <- celdaModules(x) == module	×
82
83	if (sum(ix) > 1) {	×
84	tempModel <- .celda_G(	×
85	counts = counts[ix, , drop = FALSE],	×
86	L = n,	×
87	yInitialize = "random",	×
88	splitOnIter = -1,	×
89	splitOnLast = FALSE,	×
90	nchains = 1,	×
91	verbose = FALSE)	×
92
93	splitY <- celdaClusters(tempModel)$y	×
94	splitIx <- celdaClusters(tempModel)$y > 1	×
95	splitY[splitIx] <- S4Vectors::metadata(x)$celda_parameters$L +	×
96	splitY[splitIx] - 1	×
97	splitY[!splitIx] <- module	×
98
99	newY <- celdaModules(x)	×
100	newY[ix] <- splitY	×
101	newL <- max(newY)	×
102
103	newLl <- .logLikelihoodcelda_G(	×
104	counts = counts,	×
105	y = newY,	×
106	L = newL,	×
107	beta = S4Vectors::metadata(x)$celda_parameters$beta,	×
108	delta = S4Vectors::metadata(x)$celda_parameters$delta,	×
109	gamma = S4Vectors::metadata(x)$celda_parameters$gamma)	×
110	model <- methods::new(	×
111	"celda_G",	×
112	clusters = list(y = newY),	×
113	params = list(	×
114	L = newL,	×
115	beta = S4Vectors::metadata(x)$celda_parameters$beta,	×
116	delta = S4Vectors::metadata(x)$celda_parameters$delta,	×
117	gamma = S4Vectors::metadata(x)$celda_parameters$gamma,	×
118	countChecksum = .createCountChecksum(counts)	×
119	),
120	names = list(row = rownames(x),	×
121	column = colnames(x),	×
122	sample = x@metadata$celda_parameters$sampleLevels),	×
123	finalLogLik = newLl	×
124	)
125	} else {
126	stop("Module ", module, "contains <= 1 feature. No additional",	×
127	" splitting was able to be performed.")	×
128	}
129	return(model)	×
130	}

campbio / celda / 758

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous