sherman5 authored on 30/06/2020 05:53:52 | ... | ... |
@@ -4,12 +4,24 @@ |
| 4 | 4 |
\alias{CoGAPS}
|
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 |
-CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
|
| 8 |
- messages = TRUE, outputFrequency = 1000, uncertainty = NULL, |
|
| 9 |
- checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 0, |
|
| 10 |
- checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
|
| 11 |
- workerID = 1, asynchronousUpdates = TRUE, nSnapshots = 0, |
|
| 12 |
- snapshotPhase = "sampling", ...) |
|
| 7 |
+CoGAPS( |
|
| 8 |
+ data, |
|
| 9 |
+ params = new("CogapsParams"),
|
|
| 10 |
+ nThreads = 1, |
|
| 11 |
+ messages = TRUE, |
|
| 12 |
+ outputFrequency = 1000, |
|
| 13 |
+ uncertainty = NULL, |
|
| 14 |
+ checkpointOutFile = "gaps_checkpoint.out", |
|
| 15 |
+ checkpointInterval = 0, |
|
| 16 |
+ checkpointInFile = NULL, |
|
| 17 |
+ transposeData = FALSE, |
|
| 18 |
+ BPPARAM = NULL, |
|
| 19 |
+ workerID = 1, |
|
| 20 |
+ asynchronousUpdates = TRUE, |
|
| 21 |
+ nSnapshots = 0, |
|
| 22 |
+ snapshotPhase = "sampling", |
|
| 23 |
+ ... |
|
| 24 |
+) |
|
| 13 | 25 |
} |
| 14 | 26 |
\arguments{
|
| 15 | 27 |
\item{data}{File name or R object (see details for supported types)}
|
Tom Sherman authored on 09/09/2019 23:42:33 | ... | ... |
@@ -47,6 +47,12 @@ CoGAPS) but only when the user is manually calling CoGAPS in parallel} |
| 47 | 47 |
|
| 48 | 48 |
\item{asynchronousUpdates}{enable asynchronous updating which allows for multi-threaded runs}
|
| 49 | 49 |
|
| 50 |
+\item{nSnapshots}{how many snapshots to take in each phase, setting this to 0 disables
|
|
| 51 |
+snapshots} |
|
| 52 |
+ |
|
| 53 |
+\item{snapshotPhase}{which phase to take snapsjots in e.g. "equilibration", "sampling",
|
|
| 54 |
+"all"} |
|
| 55 |
+ |
|
| 50 | 56 |
\item{...}{allows for overwriting parameters in params}
|
| 51 | 57 |
} |
| 52 | 58 |
\value{
|
Tom Sherman authored on 05/09/2019 20:47:09 | ... | ... |
@@ -5,7 +5,7 @@ |
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 | 7 |
CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 |
- messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
|
| 8 |
+ messages = TRUE, outputFrequency = 1000, uncertainty = NULL, |
|
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 0, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 | 11 |
workerID = 1, asynchronousUpdates = TRUE, nSnapshots = 0, |
Tom Sherman authored on 05/09/2019 18:14:53 | ... | ... |
@@ -8,7 +8,8 @@ CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 0, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 |
- workerID = 1, asynchronousUpdates = TRUE, ...) |
|
| 11 |
+ workerID = 1, asynchronousUpdates = TRUE, nSnapshots = 0, |
|
| 12 |
+ snapshotPhase = "sampling", ...) |
|
| 12 | 13 |
} |
| 13 | 14 |
\arguments{
|
| 14 | 15 |
\item{data}{File name or R object (see details for supported types)}
|
Tom Sherman authored on 24/06/2019 15:32:30 | ... | ... |
@@ -5,7 +5,7 @@ |
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 | 7 |
CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 |
- messages = TRUE, outputFrequency = 2500, uncertainty = NULL, |
|
| 8 |
+ messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
|
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 0, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 | 11 |
workerID = 1, asynchronousUpdates = TRUE, ...) |
| ... | ... |
@@ -44,6 +44,8 @@ only worker 1 prints output and each worker outputs when it finishes, this |
| 44 | 44 |
is not neccesary when using the default parallel methods (i.e. distributed |
| 45 | 45 |
CoGAPS) but only when the user is manually calling CoGAPS in parallel} |
| 46 | 46 |
|
| 47 |
+\item{asynchronousUpdates}{enable asynchronous updating which allows for multi-threaded runs}
|
|
| 48 |
+ |
|
| 47 | 49 |
\item{...}{allows for overwriting parameters in params}
|
| 48 | 50 |
} |
| 49 | 51 |
\value{
|
sherman5 authored on 19/06/2019 22:02:20 | ... | ... |
@@ -8,7 +8,7 @@ CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 2500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 0, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 |
- workerID = 1, ...) |
|
| 11 |
+ workerID = 1, asynchronousUpdates = TRUE, ...) |
|
| 12 | 12 |
} |
| 13 | 13 |
\arguments{
|
| 14 | 14 |
\item{data}{File name or R object (see details for supported types)}
|
Tom Sherman authored on 22/02/2019 16:39:40 | ... | ... |
@@ -5,8 +5,8 @@ |
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 | 7 |
CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 |
- messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
|
| 9 |
- checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
|
| 8 |
+ messages = TRUE, outputFrequency = 2500, uncertainty = NULL, |
|
| 9 |
+ checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 0, |
|
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 | 11 |
workerID = 1, ...) |
| 12 | 12 |
} |
Tom Sherman authored on 18/02/2019 19:33:02 | ... | ... |
@@ -7,11 +7,8 @@ |
| 7 | 7 |
CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 |
- checkpointInFile = NULL, transposeData = FALSE, |
|
| 11 |
- subsetIndices = NULL, subsetDim = 0, BPPARAM = NULL, |
|
| 12 |
- geneNames = NULL, sampleNames = NULL, fixedPatterns = NULL, |
|
| 13 |
- whichMatrixFixed = "N", takePumpSamples = FALSE, |
|
| 14 |
- outputToFile = NULL, workerID = 1, ...) |
|
| 10 |
+ checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
|
| 11 |
+ workerID = 1, ...) |
|
| 15 | 12 |
} |
| 16 | 13 |
\arguments{
|
| 17 | 14 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -40,28 +37,8 @@ contained in this file} |
| 40 | 37 |
for data that is stored as samples x genes since CoGAPS requires data to be |
| 41 | 38 |
genes x samples} |
| 42 | 39 |
|
| 43 |
-\item{subsetIndices}{set of indices to use from the data}
|
|
| 44 |
- |
|
| 45 |
-\item{subsetDim}{which dimension (1=rows, 2=cols) to subset}
|
|
| 46 |
- |
|
| 47 | 40 |
\item{BPPARAM}{BiocParallel backend}
|
| 48 | 41 |
|
| 49 |
-\item{geneNames}{vector of names of genes in data}
|
|
| 50 |
- |
|
| 51 |
-\item{sampleNames}{vector of names of samples in data}
|
|
| 52 |
- |
|
| 53 |
-\item{fixedPatterns}{fix either 'A' or 'P' matrix to these values, in the
|
|
| 54 |
-context of distributed CoGAPS (GWCoGAPS/scCoGAPS), the first phase is |
|
| 55 |
-skipped and fixedPatterns is used for all sets - allowing manual pattern |
|
| 56 |
-matching, as well as fixed runs of standard CoGAPS} |
|
| 57 |
- |
|
| 58 |
-\item{whichMatrixFixed}{either 'A' or 'P', indicating which matrix is fixed}
|
|
| 59 |
- |
|
| 60 |
-\item{takePumpSamples}{whether or not to take PUMP samples}
|
|
| 61 |
- |
|
| 62 |
-\item{outputToFile}{name of a file to save the output to, will create 4 files
|
|
| 63 |
-of the form "filename_nPatterns_[Amean, Asd, Pmean, Psd].extension"} |
|
| 64 |
- |
|
| 65 | 42 |
\item{workerID}{if calling CoGAPS in parallel the worker ID can be specified,
|
| 66 | 43 |
only worker 1 prints output and each worker outputs when it finishes, this |
| 67 | 44 |
is not neccesary when using the default parallel methods (i.e. distributed |
Tom Sherman authored on 03/01/2019 16:44:24 | ... | ... |
@@ -84,14 +84,14 @@ SingleCellExperiment. The supported file types are csv, tsv, and mtx. |
| 84 | 84 |
\examples{
|
| 85 | 85 |
# Running from R object |
| 86 | 86 |
data(GIST) |
| 87 |
-resultA <- CoGAPS(GIST.data_frame, nIterations=100) |
|
| 87 |
+resultA <- CoGAPS(GIST.data_frame, nIterations=25) |
|
| 88 | 88 |
|
| 89 | 89 |
# Running from file name |
| 90 | 90 |
gist_path <- system.file("extdata/GIST.mtx", package="CoGAPS")
|
| 91 |
-resultB <- CoGAPS(gist_path, nIterations=100) |
|
| 91 |
+resultB <- CoGAPS(gist_path, nIterations=25) |
|
| 92 | 92 |
|
| 93 | 93 |
# Setting Parameters |
| 94 | 94 |
params <- new("CogapsParams")
|
| 95 |
-params <- setParam(params, "nPatterns", 5) |
|
| 96 |
-resultC <- CoGAPS(GIST.data_frame, params, nIterations=100) |
|
| 95 |
+params <- setParam(params, "nPatterns", 3) |
|
| 96 |
+resultC <- CoGAPS(GIST.data_frame, params, nIterations=25) |
|
| 97 | 97 |
} |
Tom Sherman authored on 21/12/2018 00:52:37 | ... | ... |
@@ -10,7 +10,8 @@ CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, |
| 11 | 11 |
subsetIndices = NULL, subsetDim = 0, BPPARAM = NULL, |
| 12 | 12 |
geneNames = NULL, sampleNames = NULL, fixedPatterns = NULL, |
| 13 |
- whichMatrixFixed = "N", outputToFile = NULL, workerID = 1, ...) |
|
| 13 |
+ whichMatrixFixed = "N", takePumpSamples = FALSE, |
|
| 14 |
+ outputToFile = NULL, workerID = 1, ...) |
|
| 14 | 15 |
} |
| 15 | 16 |
\arguments{
|
| 16 | 17 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -49,10 +50,15 @@ genes x samples} |
| 49 | 50 |
|
| 50 | 51 |
\item{sampleNames}{vector of names of samples in data}
|
| 51 | 52 |
|
| 52 |
-\item{fixedPatterns}{fix either 'A' or 'P' matrix to these values}
|
|
| 53 |
+\item{fixedPatterns}{fix either 'A' or 'P' matrix to these values, in the
|
|
| 54 |
+context of distributed CoGAPS (GWCoGAPS/scCoGAPS), the first phase is |
|
| 55 |
+skipped and fixedPatterns is used for all sets - allowing manual pattern |
|
| 56 |
+matching, as well as fixed runs of standard CoGAPS} |
|
| 53 | 57 |
|
| 54 | 58 |
\item{whichMatrixFixed}{either 'A' or 'P', indicating which matrix is fixed}
|
| 55 | 59 |
|
| 60 |
+\item{takePumpSamples}{whether or not to take PUMP samples}
|
|
| 61 |
+ |
|
| 56 | 62 |
\item{outputToFile}{name of a file to save the output to, will create 4 files
|
| 57 | 63 |
of the form "filename_nPatterns_[Amean, Asd, Pmean, Psd].extension"} |
| 58 | 64 |
|
Tom Sherman authored on 19/12/2018 14:59:41 | ... | ... |
@@ -7,9 +7,10 @@ |
| 7 | 7 |
CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 |
- checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
|
| 11 |
- geneNames = NULL, sampleNames = NULL, matchedPatterns = NULL, |
|
| 12 |
- outputToFile = NULL, ...) |
|
| 10 |
+ checkpointInFile = NULL, transposeData = FALSE, |
|
| 11 |
+ subsetIndices = NULL, subsetDim = 0, BPPARAM = NULL, |
|
| 12 |
+ geneNames = NULL, sampleNames = NULL, fixedPatterns = NULL, |
|
| 13 |
+ whichMatrixFixed = "N", outputToFile = NULL, workerID = 1, ...) |
|
| 13 | 14 |
} |
| 14 | 15 |
\arguments{
|
| 15 | 16 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -38,17 +39,28 @@ contained in this file} |
| 38 | 39 |
for data that is stored as samples x genes since CoGAPS requires data to be |
| 39 | 40 |
genes x samples} |
| 40 | 41 |
|
| 42 |
+\item{subsetIndices}{set of indices to use from the data}
|
|
| 43 |
+ |
|
| 44 |
+\item{subsetDim}{which dimension (1=rows, 2=cols) to subset}
|
|
| 45 |
+ |
|
| 41 | 46 |
\item{BPPARAM}{BiocParallel backend}
|
| 42 | 47 |
|
| 43 | 48 |
\item{geneNames}{vector of names of genes in data}
|
| 44 | 49 |
|
| 45 | 50 |
\item{sampleNames}{vector of names of samples in data}
|
| 46 | 51 |
|
| 47 |
-\item{matchedPatterns}{manually matched patterns for distributed CoGAPS}
|
|
| 52 |
+\item{fixedPatterns}{fix either 'A' or 'P' matrix to these values}
|
|
| 53 |
+ |
|
| 54 |
+\item{whichMatrixFixed}{either 'A' or 'P', indicating which matrix is fixed}
|
|
| 48 | 55 |
|
| 49 | 56 |
\item{outputToFile}{name of a file to save the output to, will create 4 files
|
| 50 | 57 |
of the form "filename_nPatterns_[Amean, Asd, Pmean, Psd].extension"} |
| 51 | 58 |
|
| 59 |
+\item{workerID}{if calling CoGAPS in parallel the worker ID can be specified,
|
|
| 60 |
+only worker 1 prints output and each worker outputs when it finishes, this |
|
| 61 |
+is not neccesary when using the default parallel methods (i.e. distributed |
|
| 62 |
+CoGAPS) but only when the user is manually calling CoGAPS in parallel} |
|
| 63 |
+ |
|
| 52 | 64 |
\item{...}{allows for overwriting parameters in params}
|
| 53 | 65 |
} |
| 54 | 66 |
\value{
|
Tom Sherman authored on 12/11/2018 19:57:23 | ... | ... |
@@ -66,14 +66,14 @@ SingleCellExperiment. The supported file types are csv, tsv, and mtx. |
| 66 | 66 |
\examples{
|
| 67 | 67 |
# Running from R object |
| 68 | 68 |
data(GIST) |
| 69 |
-resultA <- CoGAPS(GIST.data_frame, nIterations=250) |
|
| 69 |
+resultA <- CoGAPS(GIST.data_frame, nIterations=100) |
|
| 70 | 70 |
|
| 71 | 71 |
# Running from file name |
| 72 | 72 |
gist_path <- system.file("extdata/GIST.mtx", package="CoGAPS")
|
| 73 |
-resultB <- CoGAPS(gist_path, nIterations=250) |
|
| 73 |
+resultB <- CoGAPS(gist_path, nIterations=100) |
|
| 74 | 74 |
|
| 75 | 75 |
# Setting Parameters |
| 76 | 76 |
params <- new("CogapsParams")
|
| 77 | 77 |
params <- setParam(params, "nPatterns", 5) |
| 78 |
-resultC <- CoGAPS(GIST.data_frame, params, nIterations=250) |
|
| 78 |
+resultC <- CoGAPS(GIST.data_frame, params, nIterations=100) |
|
| 79 | 79 |
} |
Tom Sherman authored on 02/11/2018 20:05:05 | ... | ... |
@@ -66,14 +66,14 @@ SingleCellExperiment. The supported file types are csv, tsv, and mtx. |
| 66 | 66 |
\examples{
|
| 67 | 67 |
# Running from R object |
| 68 | 68 |
data(GIST) |
| 69 |
-resultA <- CoGAPS(GIST.data_frame) |
|
| 69 |
+resultA <- CoGAPS(GIST.data_frame, nIterations=250) |
|
| 70 | 70 |
|
| 71 | 71 |
# Running from file name |
| 72 | 72 |
gist_path <- system.file("extdata/GIST.mtx", package="CoGAPS")
|
| 73 |
-resultB <- CoGAPS(gist_path) |
|
| 73 |
+resultB <- CoGAPS(gist_path, nIterations=250) |
|
| 74 | 74 |
|
| 75 | 75 |
# Setting Parameters |
| 76 | 76 |
params <- new("CogapsParams")
|
| 77 | 77 |
params <- setParam(params, "nPatterns", 5) |
| 78 |
-resultC <- CoGAPS(GIST.data_frame, params) |
|
| 78 |
+resultC <- CoGAPS(GIST.data_frame, params, nIterations=250) |
|
| 79 | 79 |
} |
Tom Sherman authored on 29/10/2018 19:56:14
Tom Sherman authored on 08/08/2018 22:34:56 | ... | ... |
@@ -8,7 +8,8 @@ CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 |
- ...) |
|
| 11 |
+ geneNames = NULL, sampleNames = NULL, matchedPatterns = NULL, |
|
| 12 |
+ outputToFile = NULL, ...) |
|
| 12 | 13 |
} |
| 13 | 14 |
\arguments{
|
| 14 | 15 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -39,6 +40,15 @@ genes x samples} |
| 39 | 40 |
|
| 40 | 41 |
\item{BPPARAM}{BiocParallel backend}
|
| 41 | 42 |
|
| 43 |
+\item{geneNames}{vector of names of genes in data}
|
|
| 44 |
+ |
|
| 45 |
+\item{sampleNames}{vector of names of samples in data}
|
|
| 46 |
+ |
|
| 47 |
+\item{matchedPatterns}{manually matched patterns for distributed CoGAPS}
|
|
| 48 |
+ |
|
| 49 |
+\item{outputToFile}{name of a file to save the output to, will create 4 files
|
|
| 50 |
+of the form "filename_nPatterns_[Amean, Asd, Pmean, Psd].extension"} |
|
| 51 |
+ |
|
| 42 | 52 |
\item{...}{allows for overwriting parameters in params}
|
| 43 | 53 |
} |
| 44 | 54 |
\value{
|
Tom Sherman authored on 06/08/2018 20:43:24 | ... | ... |
@@ -8,7 +8,7 @@ CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 |
- saveUnmatchedPatterns = FALSE, ...) |
|
| 11 |
+ ...) |
|
| 12 | 12 |
} |
| 13 | 13 |
\arguments{
|
| 14 | 14 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -39,9 +39,6 @@ genes x samples} |
| 39 | 39 |
|
| 40 | 40 |
\item{BPPARAM}{BiocParallel backend}
|
| 41 | 41 |
|
| 42 |
-\item{saveUnmatchedPatterns}{when running distributed cogaps, save the
|
|
| 43 |
-intermediate result from each subset of the data} |
|
| 44 |
- |
|
| 45 | 42 |
\item{...}{allows for overwriting parameters in params}
|
| 46 | 43 |
} |
| 47 | 44 |
\value{
|
Tom Sherman authored on 06/08/2018 19:12:48 | ... | ... |
@@ -8,7 +8,7 @@ CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
| 11 |
- ...) |
|
| 11 |
+ saveUnmatchedPatterns = FALSE, ...) |
|
| 12 | 12 |
} |
| 13 | 13 |
\arguments{
|
| 14 | 14 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -39,6 +39,9 @@ genes x samples} |
| 39 | 39 |
|
| 40 | 40 |
\item{BPPARAM}{BiocParallel backend}
|
| 41 | 41 |
|
| 42 |
+\item{saveUnmatchedPatterns}{when running distributed cogaps, save the
|
|
| 43 |
+intermediate result from each subset of the data} |
|
| 44 |
+ |
|
| 42 | 45 |
\item{...}{allows for overwriting parameters in params}
|
| 43 | 46 |
} |
| 44 | 47 |
\value{
|
Tom Sherman authored on 02/08/2018 16:52:31 | ... | ... |
@@ -7,7 +7,8 @@ |
| 7 | 7 |
CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 |
- checkpointInFile = NULL, transposeData = FALSE, ...) |
|
| 10 |
+ checkpointInFile = NULL, transposeData = FALSE, BPPARAM = NULL, |
|
| 11 |
+ ...) |
|
| 11 | 12 |
} |
| 12 | 13 |
\arguments{
|
| 13 | 14 |
\item{data}{File name or R object (see details for supported types)}
|
| ... | ... |
@@ -36,6 +37,8 @@ contained in this file} |
| 36 | 37 |
for data that is stored as samples x genes since CoGAPS requires data to be |
| 37 | 38 |
genes x samples} |
| 38 | 39 |
|
| 40 |
+\item{BPPARAM}{BiocParallel backend}
|
|
| 41 |
+ |
|
| 39 | 42 |
\item{...}{allows for overwriting parameters in params}
|
| 40 | 43 |
} |
| 41 | 44 |
\value{
|
| ... | ... |
@@ -53,14 +56,14 @@ SingleCellExperiment. The supported file types are csv, tsv, and mtx. |
| 53 | 56 |
\examples{
|
| 54 | 57 |
# Running from R object |
| 55 | 58 |
data(GIST) |
| 56 |
-resultA <- CoGAPS(GIST.D) |
|
| 59 |
+resultA <- CoGAPS(GIST.data_frame) |
|
| 57 | 60 |
|
| 58 | 61 |
# Running from file name |
| 59 | 62 |
gist_path <- system.file("extdata/GIST.mtx", package="CoGAPS")
|
| 60 | 63 |
resultB <- CoGAPS(gist_path) |
| 61 | 64 |
|
| 62 |
-Setting Parameters |
|
| 65 |
+# Setting Parameters |
|
| 63 | 66 |
params <- new("CogapsParams")
|
| 64 | 67 |
params <- setParam(params, "nPatterns", 5) |
| 65 |
-resultC <- CoGAPS(GIST.D, params) |
|
| 68 |
+resultC <- CoGAPS(GIST.data_frame, params) |
|
| 66 | 69 |
} |
Tom Sherman authored on 31/07/2018 18:45:03 | ... | ... |
@@ -4,7 +4,7 @@ |
| 4 | 4 |
\alias{CoGAPS}
|
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 |
-CoGAPS(data, params = new("CogapsParams"), nThreads = NULL,
|
|
| 7 |
+CoGAPS(data, params = new("CogapsParams"), nThreads = 1,
|
|
| 8 | 8 |
messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
| 9 | 9 |
checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
| 10 | 10 |
checkpointInFile = NULL, transposeData = FALSE, ...) |
Tom Sherman authored on 30/07/2018 02:34:58 | 1 | 1 |
new file mode 100644 |
| ... | ... |
@@ -0,0 +1,66 @@ |
| 1 |
+% Generated by roxygen2: do not edit by hand |
|
| 2 |
+% Please edit documentation in R/CoGAPS.R |
|
| 3 |
+\name{CoGAPS}
|
|
| 4 |
+\alias{CoGAPS}
|
|
| 5 |
+\title{CoGAPS Matrix Factorization Algorithm}
|
|
| 6 |
+\usage{
|
|
| 7 |
+CoGAPS(data, params = new("CogapsParams"), nThreads = NULL,
|
|
| 8 |
+ messages = TRUE, outputFrequency = 500, uncertainty = NULL, |
|
| 9 |
+ checkpointOutFile = "gaps_checkpoint.out", checkpointInterval = 1000, |
|
| 10 |
+ checkpointInFile = NULL, transposeData = FALSE, ...) |
|
| 11 |
+} |
|
| 12 |
+\arguments{
|
|
| 13 |
+\item{data}{File name or R object (see details for supported types)}
|
|
| 14 |
+ |
|
| 15 |
+\item{params}{CogapsParams object}
|
|
| 16 |
+ |
|
| 17 |
+\item{nThreads}{maximum number of threads to run on}
|
|
| 18 |
+ |
|
| 19 |
+\item{messages}{T/F for displaying output}
|
|
| 20 |
+ |
|
| 21 |
+\item{outputFrequency}{number of iterations between each output (set to 0 to
|
|
| 22 |
+disable status updates, other output is controlled by @code messages)} |
|
| 23 |
+ |
|
| 24 |
+\item{uncertainty}{uncertainty matrix - either a matrix or a supported
|
|
| 25 |
+file type} |
|
| 26 |
+ |
|
| 27 |
+\item{checkpointOutFile}{name of the checkpoint file to create}
|
|
| 28 |
+ |
|
| 29 |
+\item{checkpointInterval}{number of iterations between each checkpoint (set
|
|
| 30 |
+to 0 to disable checkpoints)} |
|
| 31 |
+ |
|
| 32 |
+\item{checkpointInFile}{if this is provided, CoGAPS runs from the checkpoint
|
|
| 33 |
+contained in this file} |
|
| 34 |
+ |
|
| 35 |
+\item{transposeData}{T/F for transposing data while reading it in - useful
|
|
| 36 |
+for data that is stored as samples x genes since CoGAPS requires data to be |
|
| 37 |
+genes x samples} |
|
| 38 |
+ |
|
| 39 |
+\item{...}{allows for overwriting parameters in params}
|
|
| 40 |
+} |
|
| 41 |
+\value{
|
|
| 42 |
+CogapsResult object |
|
| 43 |
+} |
|
| 44 |
+\description{
|
|
| 45 |
+calls the C++ MCMC code and performs Bayesian |
|
| 46 |
+matrix factorization returning the two matrices that reconstruct |
|
| 47 |
+the data matrix |
|
| 48 |
+} |
|
| 49 |
+\details{
|
|
| 50 |
+The supported R types are: matrix, data.frame, SummarizedExperiment, |
|
| 51 |
+SingleCellExperiment. The supported file types are csv, tsv, and mtx. |
|
| 52 |
+} |
|
| 53 |
+\examples{
|
|
| 54 |
+# Running from R object |
|
| 55 |
+data(GIST) |
|
| 56 |
+resultA <- CoGAPS(GIST.D) |
|
| 57 |
+ |
|
| 58 |
+# Running from file name |
|
| 59 |
+gist_path <- system.file("extdata/GIST.mtx", package="CoGAPS")
|
|
| 60 |
+resultB <- CoGAPS(gist_path) |
|
| 61 |
+ |
|
| 62 |
+Setting Parameters |
|
| 63 |
+params <- new("CogapsParams")
|
|
| 64 |
+params <- setParam(params, "nPatterns", 5) |
|
| 65 |
+resultC <- CoGAPS(GIST.D, params) |
|
| 66 |
+} |
Tom Sherman authored on 25/06/2018 20:35:29 | 1 | 1 |
deleted file mode 100644 |
| ... | ... |
@@ -1,70 +0,0 @@ |
| 1 |
-% Generated by roxygen2: do not edit by hand |
|
| 2 |
-% Please edit documentation in R/CoGAPS.R |
|
| 3 |
-\name{CoGAPS}
|
|
| 4 |
-\alias{CoGAPS}
|
|
| 5 |
-\title{CoGAPS Matrix Factorization Algorithm}
|
|
| 6 |
-\usage{
|
|
| 7 |
-CoGAPS(D, S, nFactor = 7, nIter = 1000, nOutputs = 250, alphaA = 0.01, |
|
| 8 |
- alphaP = 0.01, maxGibbmassA = 100, maxGibbmassP = 100, seed = NA, |
|
| 9 |
- messages = TRUE, singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 10 |
- fixedPatterns = matrix(0), checkpointInterval = 0, |
|
| 11 |
- checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
|
| 12 |
-} |
|
| 13 |
-\arguments{
|
|
| 14 |
-\item{D}{data matrix}
|
|
| 15 |
- |
|
| 16 |
-\item{S}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 17 |
- |
|
| 18 |
-\item{nFactor}{number of patterns (basis vectors, metagenes), which must be
|
|
| 19 |
-greater than or equal to the number of rows of FP} |
|
| 20 |
- |
|
| 21 |
-\item{nOutputs}{how often to print status into R by iterations}
|
|
| 22 |
- |
|
| 23 |
-\item{alphaA}{sparsity parameter for A domain}
|
|
| 24 |
- |
|
| 25 |
-\item{alphaP}{sparsity parameter for P domain}
|
|
| 26 |
- |
|
| 27 |
-\item{maxGibbmassA}{limit truncated normal to max size}
|
|
| 28 |
- |
|
| 29 |
-\item{maxGibbmassP}{limit truncated normal to max size}
|
|
| 30 |
- |
|
| 31 |
-\item{seed}{a positive seed is used as-is, while any negative seed tells
|
|
| 32 |
-the algorithm to pick a seed based on the current time} |
|
| 33 |
- |
|
| 34 |
-\item{messages}{display progress messages}
|
|
| 35 |
- |
|
| 36 |
-\item{singleCellRNASeq}{indicates if the data is single cell RNA-seq data}
|
|
| 37 |
- |
|
| 38 |
-\item{whichMatrixFixed}{character to indicate whether A or P matric contains
|
|
| 39 |
-the fixed patterns} |
|
| 40 |
- |
|
| 41 |
-\item{fixedPatterns}{matrix of fixed values in either A or P matrix}
|
|
| 42 |
- |
|
| 43 |
-\item{checkpointInterval}{time (in seconds) between creating a checkpoint}
|
|
| 44 |
- |
|
| 45 |
-\item{checkpointFile}{name of the checkpoint file}
|
|
| 46 |
- |
|
| 47 |
-\item{nCores}{number of cpu cores to run in parallel over}
|
|
| 48 |
- |
|
| 49 |
-\item{...}{keeps backwards compatibility with arguments from older versions}
|
|
| 50 |
- |
|
| 51 |
-\item{nEquil}{number of iterations for burn-in}
|
|
| 52 |
- |
|
| 53 |
-\item{nSample}{number of iterations for sampling}
|
|
| 54 |
-} |
|
| 55 |
-\value{
|
|
| 56 |
-list with A and P matrix estimates |
|
| 57 |
-} |
|
| 58 |
-\description{
|
|
| 59 |
-CoGAPS Matrix Factorization Algorithm |
|
| 60 |
-} |
|
| 61 |
-\details{
|
|
| 62 |
-calls the C++ MCMC code and performs Bayesian |
|
| 63 |
-matrix factorization returning the two matrices that reconstruct |
|
| 64 |
-the data matrix |
|
| 65 |
-} |
|
| 66 |
-\examples{
|
|
| 67 |
-data(SimpSim) |
|
| 68 |
-result <- CoGAPS(SimpSim.D, SimpSim.S, nFactor=3, nOutputs=250) |
|
| 69 |
-} |
|
| 70 |
- |
Tom Sherman authored on 19/06/2018 19:15:53 | ... | ... |
@@ -4,10 +4,9 @@ |
| 4 | 4 |
\alias{CoGAPS}
|
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 |
-CoGAPS(D, S, nFactor = 7, nEquil = 250, nSample = 250, nOutputs = 1000, |
|
| 8 |
- nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, maxGibbmassA = 100, |
|
| 9 |
- maxGibbmassP = 100, seed = NA, messages = TRUE, |
|
| 10 |
- singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 7 |
+CoGAPS(D, S, nFactor = 7, nIter = 1000, nOutputs = 250, alphaA = 0.01, |
|
| 8 |
+ alphaP = 0.01, maxGibbmassA = 100, maxGibbmassP = 100, seed = NA, |
|
| 9 |
+ messages = TRUE, singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 11 | 10 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 | 11 |
checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
| 13 | 12 |
} |
| ... | ... |
@@ -19,10 +18,6 @@ CoGAPS(D, S, nFactor = 7, nEquil = 250, nSample = 250, nOutputs = 1000, |
| 19 | 18 |
\item{nFactor}{number of patterns (basis vectors, metagenes), which must be
|
| 20 | 19 |
greater than or equal to the number of rows of FP} |
| 21 | 20 |
|
| 22 |
-\item{nEquil}{number of iterations for burn-in}
|
|
| 23 |
- |
|
| 24 |
-\item{nSample}{number of iterations for sampling}
|
|
| 25 |
- |
|
| 26 | 21 |
\item{nOutputs}{how often to print status into R by iterations}
|
| 27 | 22 |
|
| 28 | 23 |
\item{alphaA}{sparsity parameter for A domain}
|
| ... | ... |
@@ -52,6 +47,10 @@ the fixed patterns} |
| 52 | 47 |
\item{nCores}{number of cpu cores to run in parallel over}
|
| 53 | 48 |
|
| 54 | 49 |
\item{...}{keeps backwards compatibility with arguments from older versions}
|
| 50 |
+ |
|
| 51 |
+\item{nEquil}{number of iterations for burn-in}
|
|
| 52 |
+ |
|
| 53 |
+\item{nSample}{number of iterations for sampling}
|
|
| 55 | 54 |
} |
| 56 | 55 |
\value{
|
| 57 | 56 |
list with A and P matrix estimates |
Tom Sherman authored on 18/06/2018 22:20:06 | ... | ... |
@@ -5,19 +5,11 @@ |
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 | 7 |
CoGAPS(D, S, nFactor = 7, nEquil = 250, nSample = 250, nOutputs = 1000, |
| 8 |
-<<<<<<< HEAD |
|
| 9 | 8 |
nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, maxGibbmassA = 100, |
| 10 | 9 |
maxGibbmassP = 100, seed = NA, messages = TRUE, |
| 11 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 12 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 13 | 12 |
checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
| 14 |
-======= |
|
| 15 |
- alphaA = 0.01, alphaP = 0.01, maxGibbmassA = 100, maxGibbmassP = 100, |
|
| 16 |
- seed = -1, messages = TRUE, singleCellRNASeq = FALSE, |
|
| 17 |
- whichMatrixFixed = "N", fixedPatterns = matrix(0), |
|
| 18 |
- checkpointInterval = 0, checkpointFile = "gaps_checkpoint.out", |
|
| 19 |
- nCores = 1, ...) |
|
| 20 |
->>>>>>> develop |
|
| 21 | 13 |
} |
| 22 | 14 |
\arguments{
|
| 23 | 15 |
\item{D}{data matrix}
|
Tom Sherman authored on 18/06/2018 21:11:30
Tom Sherman authored on 18/06/2018 14:20:08 | ... | ... |
@@ -6,7 +6,7 @@ |
| 6 | 6 |
\usage{
|
| 7 | 7 |
CoGAPS(D, S, nFactor = 7, nEquil = 250, nSample = 250, nOutputs = 1000, |
| 8 | 8 |
nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, maxGibbmassA = 100, |
| 9 |
- maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 9 |
+ maxGibbmassP = 100, seed = NA, messages = TRUE, |
|
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 | 12 |
checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
Tom Sherman authored on 18/06/2018 13:49:06 | ... | ... |
@@ -4,9 +4,9 @@ |
| 4 | 4 |
\alias{CoGAPS}
|
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 |
-CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
|
| 8 |
- nOutputs = 1000, nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, |
|
| 9 |
- maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 7 |
+CoGAPS(D, S, nFactor = 7, nEquil = 250, nSample = 250, nOutputs = 1000, |
|
| 8 |
+ nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, maxGibbmassA = 100, |
|
| 9 |
+ maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 | 12 |
checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
Tom Sherman authored on 13/06/2018 21:35:14 | ... | ... |
@@ -4,12 +4,12 @@ |
| 4 | 4 |
\alias{CoGAPS}
|
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 |
-CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
|
| 8 |
- nOutputs = 1000, nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, |
|
| 9 |
- maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 10 |
- singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 11 |
- fixedPatterns = matrix(0), checkpointInterval = 0, |
|
| 12 |
- checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
|
| 7 |
+CoGAPS(D, S, nFactor = 7, nEquil = 250, nSample = 250, nOutputs = 1000, |
|
| 8 |
+ alphaA = 0.01, alphaP = 0.01, maxGibbmassA = 100, maxGibbmassP = 100, |
|
| 9 |
+ seed = -1, messages = TRUE, singleCellRNASeq = FALSE, |
|
| 10 |
+ whichMatrixFixed = "N", fixedPatterns = matrix(0), |
|
| 11 |
+ checkpointInterval = 0, checkpointFile = "gaps_checkpoint.out", |
|
| 12 |
+ nCores = 1, ...) |
|
| 13 | 13 |
} |
| 14 | 14 |
\arguments{
|
| 15 | 15 |
\item{D}{data matrix}
|
| ... | ... |
@@ -25,8 +25,6 @@ greater than or equal to the number of rows of FP} |
| 25 | 25 |
|
| 26 | 26 |
\item{nOutputs}{how often to print status into R by iterations}
|
| 27 | 27 |
|
| 28 |
-\item{nSnapshots}{the number of individual samples to capture}
|
|
| 29 |
- |
|
| 30 | 28 |
\item{alphaA}{sparsity parameter for A domain}
|
| 31 | 29 |
|
| 32 | 30 |
\item{alphaP}{sparsity parameter for P domain}
|
Tom Sherman authored on 05/06/2018 22:04:02
Tom Sherman authored on 31/05/2018 21:34:33 | 1 | 1 |
new file mode 100644 |
| ... | ... |
@@ -0,0 +1,71 @@ |
| 1 |
+% Generated by roxygen2: do not edit by hand |
|
| 2 |
+% Please edit documentation in R/CoGAPS.R |
|
| 3 |
+\name{CoGAPS}
|
|
| 4 |
+\alias{CoGAPS}
|
|
| 5 |
+\title{CoGAPS Matrix Factorization Algorithm}
|
|
| 6 |
+\usage{
|
|
| 7 |
+CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
|
| 8 |
+ nOutputs = 1000, nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, |
|
| 9 |
+ maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 10 |
+ singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 11 |
+ fixedPatterns = matrix(0), checkpointInterval = 0, |
|
| 12 |
+ checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
|
| 13 |
+} |
|
| 14 |
+\arguments{
|
|
| 15 |
+\item{D}{data matrix}
|
|
| 16 |
+ |
|
| 17 |
+\item{S}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 18 |
+ |
|
| 19 |
+\item{nFactor}{number of patterns (basis vectors, metagenes), which must be
|
|
| 20 |
+greater than or equal to the number of rows of FP} |
|
| 21 |
+ |
|
| 22 |
+\item{nEquil}{number of iterations for burn-in}
|
|
| 23 |
+ |
|
| 24 |
+\item{nSample}{number of iterations for sampling}
|
|
| 25 |
+ |
|
| 26 |
+\item{nOutputs}{how often to print status into R by iterations}
|
|
| 27 |
+ |
|
| 28 |
+\item{nSnapshots}{the number of individual samples to capture}
|
|
| 29 |
+ |
|
| 30 |
+\item{alphaA}{sparsity parameter for A domain}
|
|
| 31 |
+ |
|
| 32 |
+\item{alphaP}{sparsity parameter for P domain}
|
|
| 33 |
+ |
|
| 34 |
+\item{maxGibbmassA}{limit truncated normal to max size}
|
|
| 35 |
+ |
|
| 36 |
+\item{maxGibbmassP}{limit truncated normal to max size}
|
|
| 37 |
+ |
|
| 38 |
+\item{seed}{a positive seed is used as-is, while any negative seed tells
|
|
| 39 |
+the algorithm to pick a seed based on the current time} |
|
| 40 |
+ |
|
| 41 |
+\item{messages}{display progress messages}
|
|
| 42 |
+ |
|
| 43 |
+\item{singleCellRNASeq}{indicates if the data is single cell RNA-seq data}
|
|
| 44 |
+ |
|
| 45 |
+\item{whichMatrixFixed}{character to indicate whether A or P matric contains
|
|
| 46 |
+the fixed patterns} |
|
| 47 |
+ |
|
| 48 |
+\item{fixedPatterns}{matrix of fixed values in either A or P matrix}
|
|
| 49 |
+ |
|
| 50 |
+\item{checkpointInterval}{time (in seconds) between creating a checkpoint}
|
|
| 51 |
+ |
|
| 52 |
+\item{checkpointFile}{name of the checkpoint file}
|
|
| 53 |
+ |
|
| 54 |
+\item{...}{keeps backwards compatibility with arguments from older versions}
|
|
| 55 |
+} |
|
| 56 |
+\value{
|
|
| 57 |
+list with A and P matrix estimates |
|
| 58 |
+} |
|
| 59 |
+\description{
|
|
| 60 |
+CoGAPS Matrix Factorization Algorithm |
|
| 61 |
+} |
|
| 62 |
+\details{
|
|
| 63 |
+calls the C++ MCMC code and performs Bayesian |
|
| 64 |
+matrix factorization returning the two matrices that reconstruct |
|
| 65 |
+the data matrix |
|
| 66 |
+} |
|
| 67 |
+\examples{
|
|
| 68 |
+data(SimpSim) |
|
| 69 |
+result <- CoGAPS(SimpSim.D, SimpSim.S, nFactor=3, nOutputs=250) |
|
| 70 |
+} |
|
| 71 |
+ |
Tom Sherman authored on 31/05/2018 21:28:52 | 1 | 1 |
deleted file mode 100644 |
| ... | ... |
@@ -1,71 +0,0 @@ |
| 1 |
-% Generated by roxygen2: do not edit by hand |
|
| 2 |
-% Please edit documentation in R/CoGAPS.R |
|
| 3 |
-\name{CoGAPS}
|
|
| 4 |
-\alias{CoGAPS}
|
|
| 5 |
-\title{CoGAPS Matrix Factorization Algorithm}
|
|
| 6 |
-\usage{
|
|
| 7 |
-CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
|
| 8 |
- nOutputs = 1000, nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, |
|
| 9 |
- maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 10 |
- singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 11 |
- fixedPatterns = matrix(0), checkpointInterval = 0, |
|
| 12 |
- checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
|
| 13 |
-} |
|
| 14 |
-\arguments{
|
|
| 15 |
-\item{D}{data matrix}
|
|
| 16 |
- |
|
| 17 |
-\item{S}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 18 |
- |
|
| 19 |
-\item{nFactor}{number of patterns (basis vectors, metagenes), which must be
|
|
| 20 |
-greater than or equal to the number of rows of FP} |
|
| 21 |
- |
|
| 22 |
-\item{nEquil}{number of iterations for burn-in}
|
|
| 23 |
- |
|
| 24 |
-\item{nSample}{number of iterations for sampling}
|
|
| 25 |
- |
|
| 26 |
-\item{nOutputs}{how often to print status into R by iterations}
|
|
| 27 |
- |
|
| 28 |
-\item{nSnapshots}{the number of individual samples to capture}
|
|
| 29 |
- |
|
| 30 |
-\item{alphaA}{sparsity parameter for A domain}
|
|
| 31 |
- |
|
| 32 |
-\item{alphaP}{sparsity parameter for P domain}
|
|
| 33 |
- |
|
| 34 |
-\item{maxGibbmassA}{limit truncated normal to max size}
|
|
| 35 |
- |
|
| 36 |
-\item{maxGibbmassP}{limit truncated normal to max size}
|
|
| 37 |
- |
|
| 38 |
-\item{seed}{a positive seed is used as-is, while any negative seed tells
|
|
| 39 |
-the algorithm to pick a seed based on the current time} |
|
| 40 |
- |
|
| 41 |
-\item{messages}{display progress messages}
|
|
| 42 |
- |
|
| 43 |
-\item{singleCellRNASeq}{indicates if the data is single cell RNA-seq data}
|
|
| 44 |
- |
|
| 45 |
-\item{whichMatrixFixed}{character to indicate whether A or P matric contains
|
|
| 46 |
-the fixed patterns} |
|
| 47 |
- |
|
| 48 |
-\item{fixedPatterns}{matrix of fixed values in either A or P matrix}
|
|
| 49 |
- |
|
| 50 |
-\item{checkpointInterval}{time (in seconds) between creating a checkpoint}
|
|
| 51 |
- |
|
| 52 |
-\item{checkpointFile}{name of the checkpoint file}
|
|
| 53 |
- |
|
| 54 |
-\item{...}{keeps backwards compatibility with arguments from older versions}
|
|
| 55 |
-} |
|
| 56 |
-\value{
|
|
| 57 |
-list with A and P matrix estimates |
|
| 58 |
-} |
|
| 59 |
-\description{
|
|
| 60 |
-CoGAPS Matrix Factorization Algorithm |
|
| 61 |
-} |
|
| 62 |
-\details{
|
|
| 63 |
-calls the C++ MCMC code and performs Bayesian |
|
| 64 |
-matrix factorization returning the two matrices that reconstruct |
|
| 65 |
-the data matrix |
|
| 66 |
-} |
|
| 67 |
-\examples{
|
|
| 68 |
-data(SimpSim) |
|
| 69 |
-result <- CoGAPS(SimpSim.D, SimpSim.S, nFactor=3, nOutputs=250) |
|
| 70 |
-} |
|
| 71 |
- |
Tom Sherman authored on 09/05/2018 21:27:17 | ... | ... |
@@ -9,7 +9,7 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 9 | 9 |
maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 |
- checkpointFile = "gaps_checkpoint.out", ...) |
|
| 12 |
+ checkpointFile = "gaps_checkpoint.out", nCores = 1, ...) |
|
| 13 | 13 |
} |
| 14 | 14 |
\arguments{
|
| 15 | 15 |
\item{D}{data matrix}
|
sherman5 authored on 19/04/2018 20:02:35 | ... | ... |
@@ -9,8 +9,7 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 9 | 9 |
maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 |
- checkpointFile = "gaps_checkpoint.out", pumpThreshold = "unique", |
|
| 13 |
- nPumpSamples = 0, ...) |
|
| 12 |
+ checkpointFile = "gaps_checkpoint.out", ...) |
|
| 14 | 13 |
} |
| 15 | 14 |
\arguments{
|
| 16 | 15 |
\item{D}{data matrix}
|
| ... | ... |
@@ -52,10 +51,6 @@ the fixed patterns} |
| 52 | 51 |
|
| 53 | 52 |
\item{checkpointFile}{name of the checkpoint file}
|
| 54 | 53 |
|
| 55 |
-\item{pumpThreshold}{type of threshold for pump statistic}
|
|
| 56 |
- |
|
| 57 |
-\item{nPumpSamples}{number of samples used in pump statistic}
|
|
| 58 |
- |
|
| 59 | 54 |
\item{...}{keeps backwards compatibility with arguments from older versions}
|
| 60 | 55 |
} |
| 61 | 56 |
\value{
|
| ... | ... |
@@ -73,3 +68,4 @@ the data matrix |
| 73 | 68 |
data(SimpSim) |
| 74 | 69 |
result <- CoGAPS(SimpSim.D, SimpSim.S, nFactor=3, nOutputs=250) |
| 75 | 70 |
} |
| 71 |
+ |
Genevieve Stein-O'Brien authored on 26/02/2018 17:11:49
sherman5 authored on 07/02/2018 18:08:11 | ... | ... |
@@ -10,7 +10,7 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 | 12 |
checkpointFile = "gaps_checkpoint.out", pumpThreshold = "unique", |
| 13 |
- nPumpSamples = 100, ...) |
|
| 13 |
+ nPumpSamples = 0, ...) |
|
| 14 | 14 |
} |
| 15 | 15 |
\arguments{
|
| 16 | 16 |
\item{D}{data matrix}
|
| ... | ... |
@@ -52,6 +52,10 @@ the fixed patterns} |
| 52 | 52 |
|
| 53 | 53 |
\item{checkpointFile}{name of the checkpoint file}
|
| 54 | 54 |
|
| 55 |
+\item{pumpThreshold}{type of threshold for pump statistic}
|
|
| 56 |
+ |
|
| 57 |
+\item{nPumpSamples}{number of samples used in pump statistic}
|
|
| 58 |
+ |
|
| 55 | 59 |
\item{...}{keeps backwards compatibility with arguments from older versions}
|
| 56 | 60 |
} |
| 57 | 61 |
\value{
|
sherman5 authored on 06/02/2018 22:36:35 | ... | ... |
@@ -9,7 +9,8 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 9 | 9 |
maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 |
- checkpointFile = "gaps_checkpoint.out", pumpThreshold = "unique", ...) |
|
| 12 |
+ checkpointFile = "gaps_checkpoint.out", pumpThreshold = "unique", |
|
| 13 |
+ nPumpSamples = 100, ...) |
|
| 13 | 14 |
} |
| 14 | 15 |
\arguments{
|
| 15 | 16 |
\item{D}{data matrix}
|
sherman5 authored on 06/02/2018 19:49:56 | ... | ... |
@@ -9,7 +9,7 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 9 | 9 |
maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 | 11 |
fixedPatterns = matrix(0), checkpointInterval = 0, |
| 12 |
- checkpointFile = "gaps_checkpoint.out", ...) |
|
| 12 |
+ checkpointFile = "gaps_checkpoint.out", pumpThreshold = "unique", ...) |
|
| 13 | 13 |
} |
| 14 | 14 |
\arguments{
|
| 15 | 15 |
\item{D}{data matrix}
|
sherman5 authored on 30/01/2018 20:50:37 | ... | ... |
@@ -8,7 +8,8 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 8 | 8 |
nOutputs = 1000, nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, |
| 9 | 9 |
maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
| 10 | 10 |
singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
| 11 |
- fixedPatterns = matrix(0), checkpointInterval = 0, ...) |
|
| 11 |
+ fixedPatterns = matrix(0), checkpointInterval = 0, |
|
| 12 |
+ checkpointFile = "gaps_checkpoint.out", ...) |
|
| 12 | 13 |
} |
| 13 | 14 |
\arguments{
|
| 14 | 15 |
\item{D}{data matrix}
|
| ... | ... |
@@ -48,6 +49,8 @@ the fixed patterns} |
| 48 | 49 |
|
| 49 | 50 |
\item{checkpointInterval}{time (in seconds) between creating a checkpoint}
|
| 50 | 51 |
|
| 52 |
+\item{checkpointFile}{name of the checkpoint file}
|
|
| 53 |
+ |
|
| 51 | 54 |
\item{...}{keeps backwards compatibility with arguments from older versions}
|
| 52 | 55 |
} |
| 53 | 56 |
\value{
|
sherman5 authored on 26/01/2018 20:18:09 | ... | ... |
@@ -58,7 +58,11 @@ CoGAPS Matrix Factorization Algorithm |
| 58 | 58 |
} |
| 59 | 59 |
\details{
|
| 60 | 60 |
calls the C++ MCMC code and performs Bayesian |
| 61 |
- matrix factorization returning the two matrices that reconstruct |
|
| 62 |
- the data matrix |
|
| 61 |
+matrix factorization returning the two matrices that reconstruct |
|
| 62 |
+the data matrix |
|
| 63 |
+} |
|
| 64 |
+\examples{
|
|
| 65 |
+data(SimpSim) |
|
| 66 |
+result <- CoGAPS(SimpSim.D, SimpSim.S, nFactor=3, nOutputs=250) |
|
| 63 | 67 |
} |
| 64 | 68 |
|
sherman5 authored on 26/01/2018 17:42:54 | ... | ... |
@@ -47,6 +47,8 @@ the fixed patterns} |
| 47 | 47 |
\item{fixedPatterns}{matrix of fixed values in either A or P matrix}
|
| 48 | 48 |
|
| 49 | 49 |
\item{checkpointInterval}{time (in seconds) between creating a checkpoint}
|
| 50 |
+ |
|
| 51 |
+\item{...}{keeps backwards compatibility with arguments from older versions}
|
|
| 50 | 52 |
} |
| 51 | 53 |
\value{
|
| 52 | 54 |
list with A and P matrix estimates |
sherman5 authored on 26/01/2018 00:56:10 | ... | ... |
@@ -14,6 +14,39 @@ CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
| 14 | 14 |
\item{D}{data matrix}
|
| 15 | 15 |
|
| 16 | 16 |
\item{S}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
| 17 |
+ |
|
| 18 |
+\item{nFactor}{number of patterns (basis vectors, metagenes), which must be
|
|
| 19 |
+greater than or equal to the number of rows of FP} |
|
| 20 |
+ |
|
| 21 |
+\item{nEquil}{number of iterations for burn-in}
|
|
| 22 |
+ |
|
| 23 |
+\item{nSample}{number of iterations for sampling}
|
|
| 24 |
+ |
|
| 25 |
+\item{nOutputs}{how often to print status into R by iterations}
|
|
| 26 |
+ |
|
| 27 |
+\item{nSnapshots}{the number of individual samples to capture}
|
|
| 28 |
+ |
|
| 29 |
+\item{alphaA}{sparsity parameter for A domain}
|
|
| 30 |
+ |
|
| 31 |
+\item{alphaP}{sparsity parameter for P domain}
|
|
| 32 |
+ |
|
| 33 |
+\item{maxGibbmassA}{limit truncated normal to max size}
|
|
| 34 |
+ |
|
| 35 |
+\item{maxGibbmassP}{limit truncated normal to max size}
|
|
| 36 |
+ |
|
| 37 |
+\item{seed}{a positive seed is used as-is, while any negative seed tells
|
|
| 38 |
+the algorithm to pick a seed based on the current time} |
|
| 39 |
+ |
|
| 40 |
+\item{messages}{display progress messages}
|
|
| 41 |
+ |
|
| 42 |
+\item{singleCellRNASeq}{indicates if the data is single cell RNA-seq data}
|
|
| 43 |
+ |
|
| 44 |
+\item{whichMatrixFixed}{character to indicate whether A or P matric contains
|
|
| 45 |
+the fixed patterns} |
|
| 46 |
+ |
|
| 47 |
+\item{fixedPatterns}{matrix of fixed values in either A or P matrix}
|
|
| 48 |
+ |
|
| 49 |
+\item{checkpointInterval}{time (in seconds) between creating a checkpoint}
|
|
| 17 | 50 |
} |
| 18 | 51 |
\value{
|
| 19 | 52 |
list with A and P matrix estimates |
| ... | ... |
@@ -23,7 +56,7 @@ CoGAPS Matrix Factorization Algorithm |
| 23 | 56 |
} |
| 24 | 57 |
\details{
|
| 25 | 58 |
calls the C++ MCMC code and performs Bayesian |
| 26 |
-matrix factorization returning the two matrices that reconstruct |
|
| 27 |
-the data matrix |
|
| 59 |
+ matrix factorization returning the two matrices that reconstruct |
|
| 60 |
+ the data matrix |
|
| 28 | 61 |
} |
| 29 | 62 |
|
sherman5 authored on 25/01/2018 21:13:26 | ... | ... |
@@ -4,7 +4,11 @@ |
| 4 | 4 |
\alias{CoGAPS}
|
| 5 | 5 |
\title{CoGAPS Matrix Factorization Algorithm}
|
| 6 | 6 |
\usage{
|
| 7 |
-CoGAPS(D, S, ...) |
|
| 7 |
+CoGAPS(D, S, nFactor = 7, nEquil = 1000, nSample = 1000, |
|
| 8 |
+ nOutputs = 1000, nSnapshots = 0, alphaA = 0.01, alphaP = 0.01, |
|
| 9 |
+ maxGibbmassA = 100, maxGibbmassP = 100, seed = -1, messages = TRUE, |
|
| 10 |
+ singleCellRNASeq = FALSE, whichMatrixFixed = "N", |
|
| 11 |
+ fixedPatterns = matrix(0), checkpointInterval = 0, ...) |
|
| 8 | 12 |
} |
| 9 | 13 |
\arguments{
|
| 10 | 14 |
\item{D}{data matrix}
|
sherman5 authored on 24/01/2018 23:38:02 | ... | ... |
@@ -2,75 +2,24 @@ |
| 2 | 2 |
% Please edit documentation in R/CoGAPS.R |
| 3 | 3 |
\name{CoGAPS}
|
| 4 | 4 |
\alias{CoGAPS}
|
| 5 |
-\title{\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 6 |
-matrix factorization returning the two matrices that reconstruct |
|
| 7 |
-the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 8 |
-activity with nPerm set to 500} |
|
| 5 |
+\title{CoGAPS Matrix Factorization Algorithm}
|
|
| 9 | 6 |
\usage{
|
| 10 |
-CoGAPS(data, unc, ABins = data.frame(), PBins = data.frame(), GStoGenes, |
|
| 11 |
- nFactor = 7, simulation_id = "simulation", nEquil = 1000, |
|
| 12 |
- nSample = 1000, nOutR = 1000, output_atomic = FALSE, |
|
| 13 |
- fixedBinProbs = FALSE, fixedDomain = "N", sampleSnapshots = TRUE, |
|
| 14 |
- numSnapshots = 100, plot = TRUE, nPerm = 500, alphaA = 0.01, |
|
| 15 |
- nMaxA = 1e+05, max_gibbmass_paraA = 100, alphaP = 0.01, nMaxP = 1e+05, |
|
| 16 |
- max_gibbmass_paraP = 100) |
|
| 7 |
+CoGAPS(D, S, ...) |
|
| 17 | 8 |
} |
| 18 | 9 |
\arguments{
|
| 19 |
-\item{data}{data matrix}
|
|
| 20 |
- |
|
| 21 |
-\item{unc}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 22 |
- |
|
| 23 |
-\item{ABins}{a matrix of same size as A which gives relative
|
|
| 24 |
-probability of that element being non-zero} |
|
| 25 |
- |
|
| 26 |
-\item{PBins}{a matrix of same size as P which gives relative
|
|
| 27 |
-probability of that element being non-zero} |
|
| 28 |
- |
|
| 29 |
-\item{GStoGenes}{data.frame or list with gene sets}
|
|
| 30 |
- |
|
| 31 |
-\item{nFactor}{number of patterns (basis vectors, metagenes)}
|
|
| 32 |
- |
|
| 33 |
-\item{simulation_id}{name to attach to atoms files if created}
|
|
| 34 |
- |
|
| 35 |
-\item{nEquil}{number of iterations for burn-in}
|
|
| 36 |
- |
|
| 37 |
-\item{nSample}{number of iterations for sampling}
|
|
| 38 |
- |
|
| 39 |
-\item{nOutR}{how often to print status into R by iterations}
|
|
| 40 |
- |
|
| 41 |
-\item{output_atomic}{whether to write atom files (large)}
|
|
| 42 |
- |
|
| 43 |
-\item{fixedBinProbs}{Boolean for using relative probabilities
|
|
| 44 |
-given in Abins and Pbins} |
|
| 10 |
+\item{D}{data matrix}
|
|
| 45 | 11 |
|
| 46 |
-\item{fixedDomain}{character to indicate whether A or P is
|
|
| 47 |
-domain for relative probabilities} |
|
| 48 |
- |
|
| 49 |
-\item{sampleSnapshots}{Boolean to indicate whether to capture
|
|
| 50 |
-individual samples from Markov chain during sampling} |
|
| 51 |
- |
|
| 52 |
-\item{numSnapshots}{the number of individual samples to capture}
|
|
| 53 |
- |
|
| 54 |
-\item{plot}{Boolean to indicate whether to produce output graphics}
|
|
| 55 |
- |
|
| 56 |
-\item{nPerm}{number of permutations in gene set test}
|
|
| 57 |
- |
|
| 58 |
-\item{alphaA}{sparsity parameter for A domain}
|
|
| 59 |
- |
|
| 60 |
-\item{nMaxA}{PRESENTLY UNUSED, future = limit number of atoms}
|
|
| 61 |
- |
|
| 62 |
-\item{max_gibbmass_paraA}{limit truncated normal to max size}
|
|
| 63 |
- |
|
| 64 |
-\item{alphaP}{sparsity parameter for P domain}
|
|
| 65 |
- |
|
| 66 |
-\item{nMaxP}{PRESENTLY UNUSED, future = limit number of atoms}
|
|
| 67 |
- |
|
| 68 |
-\item{max_gibbmass_paraP}{limit truncated normal to max size}
|
|
| 12 |
+\item{S}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 13 |
+} |
|
| 14 |
+\value{
|
|
| 15 |
+list with A and P matrix estimates |
|
| 69 | 16 |
} |
| 70 | 17 |
\description{
|
| 71 |
-\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 18 |
+CoGAPS Matrix Factorization Algorithm |
|
| 19 |
+} |
|
| 20 |
+\details{
|
|
| 21 |
+calls the C++ MCMC code and performs Bayesian |
|
| 72 | 22 |
matrix factorization returning the two matrices that reconstruct |
| 73 |
-the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 74 |
-activity with nPerm set to 500 |
|
| 23 |
+the data matrix |
|
| 75 | 24 |
} |
| 76 | 25 |
|
sherman5 authored on 06/09/2017 19:58:09 
Tom Sherman authored on 18/04/2017 18:38:46 Now package.R populates CoGAPS-package.Rd, instead of CoGAPS.Rd, which
was causing an error due to the same alias (CoGAPS-package) being in
two different Rd files. The old CoGAPS.Rd file (mostly generated by
Rcpp?) is overwritten by Roygen but the resulting file is mostly
unchanged (a few formatting differences).
Jacob Carey authored on 28/12/2015 17:01:59 | ... | ... |
@@ -1,9 +1,7 @@ |
| 1 | 1 |
% Generated by roxygen2: do not edit by hand |
| 2 |
-% Please edit documentation in R/CoGAPS.R, R/package.R |
|
| 3 |
-\docType{package}
|
|
| 2 |
+% Please edit documentation in R/CoGAPS.R |
|
| 4 | 3 |
\name{CoGAPS}
|
| 5 | 4 |
\alias{CoGAPS}
|
| 6 |
-\alias{CoGAPS-package}
|
|
| 7 | 5 |
\title{\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
| 8 | 6 |
matrix factorization returning the two matrices that reconstruct |
| 9 | 7 |
the data matrix and then calls calcCoGAPSStat to estimate gene set |
| ... | ... |
@@ -74,7 +72,5 @@ individual samples from Markov chain during sampling} |
| 74 | 72 |
matrix factorization returning the two matrices that reconstruct |
| 75 | 73 |
the data matrix and then calls calcCoGAPSStat to estimate gene set |
| 76 | 74 |
activity with nPerm set to 500 |
| 77 |
- |
|
| 78 |
-Coordinated Gene Activity in Pattern Sets |
|
| 79 | 75 |
} |
| 80 | 76 |
|
Jacob Carey authored on 18/12/2015 21:07:44 | ... | ... |
@@ -1,119 +1,80 @@ |
| 1 |
-% Generated by roxygen2 (4.1.1): do not edit by hand |
|
| 2 |
-% Please edit documentation in R/CoGAPS.R |
|
| 3 |
-\name{CoGAPS}
|
|
| 4 |
-\alias{CoGAPS}
|
|
| 5 |
-\title{\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 6 |
-matrix factorization returning the two matrices that reconstruct |
|
| 7 |
-the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 8 |
-activity with nPerm set to 500} |
|
| 9 |
-\usage{
|
|
| 10 |
-CoGAPS(data, unc, ABins = data.frame(), PBins = data.frame(), GStoGenes, |
|
| 11 |
- nFactor = 7, simulation_id = "simulation", nEquil = 1000, |
|
| 12 |
- nSample = 1000, nOutR = 1000, output_atomic = FALSE, |
|
| 13 |
- fixedBinProbs = FALSE, fixedDomain = "N", sampleSnapshots = TRUE, |
|
| 14 |
- numSnapshots = 100, plot = TRUE, nPerm = 500, alphaA = 0.01, |
|
| 15 |
- nMaxA = 1e+05, max_gibbmass_paraA = 100, alphaP = 0.01, nMaxP = 1e+05, |
|
| 16 |
- max_gibbmass_paraP = 100) |
|
| 17 |
-} |
|
| 18 |
-\arguments{
|
|
| 19 |
-\item{data}{data matrix}
|
|
| 20 |
- |
|
| 21 |
-\item{unc}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 22 |
- |
|
| 23 |
-\item{ABins}{a matrix of same size as A which gives relative
|
|
| 24 |
-probability of that element being non-zero} |
|
| 25 |
- |
|
| 26 |
-\item{PBins}{a matrix of same size as P which gives relative
|
|
| 27 |
-probability of that element being non-zero} |
|
| 28 |
- |
|
| 29 |
-\item{GStoGenes}{data.frame or list with gene sets}
|
|
| 30 |
- |
|
| 31 |
-\item{nFactor}{number of patterns (basis vectors, metagenes)}
|
|
| 32 |
- |
|
| 33 |
-\item{simulation_id}{name to attach to atoms files if created}
|
|
| 34 |
- |
|
| 35 |
-\item{nEquil}{number of iterations for burn-in}
|
|
| 36 |
- |
|
| 37 |
-\item{nSample}{number of iterations for sampling}
|
|
| 38 |
- |
|
| 39 |
-\item{nOutR}{how often to print status into R by iterations}
|
|
| 40 |
- |
|
| 41 |
-\item{output_atomic}{whether to write atom files (large)}
|
|
| 42 |
- |
|
| 43 |
-\item{fixedBinProbs}{Boolean for using relative probabilities
|
|
| 44 |
-given in Abins and Pbins} |
|
| 45 |
- |
|
| 46 |
-\item{fixedDomain}{character to indicate whether A or P is
|
|
| 47 |
-domain for relative probabilities} |
|
| 48 |
- |
|
| 49 |
-\item{sampleSnapshots}{Boolean to indicate whether to capture
|
|
| 50 |
-individual samples from Markov chain during sampling} |
|
| 51 |
- |
|
| 52 |
-\item{numSnapshots}{the number of individual samples to capture}
|
|
| 53 |
- |
|
| 54 |
-\item{plot}{Boolean to indicate whether to produce output graphics}
|
|
| 55 |
- |
|
| 56 |
-\item{nPerm}{number of permutations in gene set test}
|
|
| 57 |
- |
|
| 58 |
-\item{alphaA}{sparsity parameter for A domain}
|
|
| 59 |
- |
|
| 60 |
-\item{nMaxA}{PRESENTLY UNUSED, future = limit number of atoms}
|
|
| 61 |
- |
|
| 62 |
-\item{max_gibbmass_paraA}{limit truncated normal to max size}
|
|
| 63 |
- |
|
| 64 |
-\item{alphaP}{sparsity parameter for P domain}
|
|
| 65 |
- |
|
| 66 |
-\item{nMaxP}{PRESENTLY UNUSED, future = limit number of atoms}
|
|
| 67 |
- |
|
| 68 |
-\item{max_gibbmass_paraP}{limit truncated normal to max size}
|
|
| 69 |
-} |
|
| 70 |
-\description{
|
|
| 71 |
-\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 72 |
-matrix factorization returning the two matrices that reconstruct |
|
| 73 |
-the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 74 |
-activity with nPerm set to 500 |
|
| 75 |
-} |
|
| 76 |
- |
|
| 77 |
- |
|
| 78 |
-\details{
|
|
| 79 |
- CoGAPS first decomposes the data matrix using GAPS, \eqn{{\bf{D}}}, into a basis of underlying patterns and then determines the gene set activity in each of these patterns.
|
|
| 80 |
- |
|
| 81 |
- The GAPS decomposition is achieved by finding amplitude and pattern matrices (\eqn{{\bf{A}}} and \eqn{{\bf{P}}}, respectively) for which \deqn{{\bf{D}} = {\bf{A}}{\bf{P}} + \Sigma,} where \eqn{\Sigma} is the matrix of uncertainties given by unc. The matrices \eqn{\bf{A}} and \eqn{\bf{P}} are assumed to have the atomic prior described in Sibisi and Skilling (1997) and are found with MCMC sampling.
|
|
| 82 |
- |
|
| 83 |
-Then, the patterns identified in the columns of \eqn{\bf{P}} are linked to activity in each of the gene sets specified in GStoGenes using a novel z-score based statistic developed in Ochs et al. (2009). Specifically, the z-score for pattern \eqn{p} and gene set \eqn{G_{i}} containing $G$ total genes is given by \deqn{Z_{i,p} = \frac{1}{G} \sum_{g in \mathcal{G_{i}}} {\frac{{\bf{A}_{gp}}}{Asd_{gp}}},}
|
|
| 84 |
-where \eqn{g} indexes the genes in the set and \eqn{Asd_{gp}} is the standard deviation of \eqn{{\bf{A}}_{gp}} obtained from MCMC sampling. CoGAPS then uses the specified \code{nPerm} random sample tests to compute a consistent p value estimate from that z score. Note that the data from Ochs et al. (2009) are provided with this package in GIST_TS_20084.RData and TFGSList.RData are also provided with this package for further validation.}
|
|
| 85 |
- |
|
| 86 |
-\value{
|
|
| 87 |
- A list containing: |
|
| 88 |
- \item{meanChi2}{Value of \eqn{chi^2} for Amean and Pmean.}
|
|
| 89 |
- \item{D}{Data matrix \eqn{{\bf{D}}} input to factorization.}
|
|
| 90 |
- \item{Sigma}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 91 |
- \item{Amean}{Sampled mean value of the amplitude matrix \eqn{{\bf{A}}}.}
|
|
| 92 |
- \item{Asd}{Sampled standard deviation of the amplitude matrix \eqn{{\bf{A}}}.}
|
|
| 93 |
- \item{Pmean}{Sampled mean value of the amplitude matrix \eqn{{\bf{P}}}.}
|
|
| 94 |
- \item{Psd}{Sampled standard deviation of the amplitude matrix \eqn{{\bf{P}}}.}
|
|
| 95 |
- \item{GSUpreg}{p-values for upregulation of each gene set in each pattern.}
|
|
| 96 |
- \item{GSDownreg}{p-values for downregulation of each gene set in each pattern.}
|
|
| 97 |
- \item{GSActEst}{p-values for activity of each gene set in each pattern.}
|
|
| 98 |
-} |
|
| 99 |
- |
|
| 100 |
-\examples{
|
|
| 101 |
-\dontrun{
|
|
| 102 |
-## Load data |
|
| 103 |
-nIter <- 5000 |
|
| 104 |
- |
|
| 105 |
-## Run GAPS matrix decomposition with gene set statistic |
|
| 106 |
-results <- CoGAPS(data=SimpSim.D, unc=SimpSim.S, |
|
| 107 |
- GStoGenes=GSets, |
|
| 108 |
- nFactor=3, |
|
| 109 |
- nEquil=nIter, nSample=nIter, |
|
| 110 |
- plot=FALSE) |
|
| 111 |
- |
|
| 112 |
- |
|
| 113 |
-## Plot the results |
|
| 114 |
-plotGAPS(results$Amean, results$Pmean, 'GSFigs') |
|
| 115 |
-} |
|
| 116 |
-} |
|
| 117 |
- |
|
| 118 |
-\seealso{\code{\link{gapsRun}},\code{\link{calcCoGAPSStat}}}
|
|
| 119 |
-\keyword{misc}
|
|
| 1 |
+% Generated by roxygen2: do not edit by hand |
|
| 2 |
+% Please edit documentation in R/CoGAPS.R, R/package.R |
|
| 3 |
+\docType{package}
|
|
| 4 |
+\name{CoGAPS}
|
|
| 5 |
+\alias{CoGAPS}
|
|
| 6 |
+\alias{CoGAPS-package}
|
|
| 7 |
+\title{\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 8 |
+matrix factorization returning the two matrices that reconstruct |
|
| 9 |
+the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 10 |
+activity with nPerm set to 500} |
|
| 11 |
+\usage{
|
|
| 12 |
+CoGAPS(data, unc, ABins = data.frame(), PBins = data.frame(), GStoGenes, |
|
| 13 |
+ nFactor = 7, simulation_id = "simulation", nEquil = 1000, |
|
| 14 |
+ nSample = 1000, nOutR = 1000, output_atomic = FALSE, |
|
| 15 |
+ fixedBinProbs = FALSE, fixedDomain = "N", sampleSnapshots = TRUE, |
|
| 16 |
+ numSnapshots = 100, plot = TRUE, nPerm = 500, alphaA = 0.01, |
|
| 17 |
+ nMaxA = 1e+05, max_gibbmass_paraA = 100, alphaP = 0.01, nMaxP = 1e+05, |
|
| 18 |
+ max_gibbmass_paraP = 100) |
|
| 19 |
+} |
|
| 20 |
+\arguments{
|
|
| 21 |
+\item{data}{data matrix}
|
|
| 22 |
+ |
|
| 23 |
+\item{unc}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 24 |
+ |
|
| 25 |
+\item{ABins}{a matrix of same size as A which gives relative
|
|
| 26 |
+probability of that element being non-zero} |
|
| 27 |
+ |
|
| 28 |
+\item{PBins}{a matrix of same size as P which gives relative
|
|
| 29 |
+probability of that element being non-zero} |
|
| 30 |
+ |
|
| 31 |
+\item{GStoGenes}{data.frame or list with gene sets}
|
|
| 32 |
+ |
|
| 33 |
+\item{nFactor}{number of patterns (basis vectors, metagenes)}
|
|
| 34 |
+ |
|
| 35 |
+\item{simulation_id}{name to attach to atoms files if created}
|
|
| 36 |
+ |
|
| 37 |
+\item{nEquil}{number of iterations for burn-in}
|
|
| 38 |
+ |
|
| 39 |
+\item{nSample}{number of iterations for sampling}
|
|
| 40 |
+ |
|
| 41 |
+\item{nOutR}{how often to print status into R by iterations}
|
|
| 42 |
+ |
|
| 43 |
+\item{output_atomic}{whether to write atom files (large)}
|
|
| 44 |
+ |
|
| 45 |
+\item{fixedBinProbs}{Boolean for using relative probabilities
|
|
| 46 |
+given in Abins and Pbins} |
|
| 47 |
+ |
|
| 48 |
+\item{fixedDomain}{character to indicate whether A or P is
|
|
| 49 |
+domain for relative probabilities} |
|
| 50 |
+ |
|
| 51 |
+\item{sampleSnapshots}{Boolean to indicate whether to capture
|
|
| 52 |
+individual samples from Markov chain during sampling} |
|
| 53 |
+ |
|
| 54 |
+\item{numSnapshots}{the number of individual samples to capture}
|
|
| 55 |
+ |
|
| 56 |
+\item{plot}{Boolean to indicate whether to produce output graphics}
|
|
| 57 |
+ |
|
| 58 |
+\item{nPerm}{number of permutations in gene set test}
|
|
| 59 |
+ |
|
| 60 |
+\item{alphaA}{sparsity parameter for A domain}
|
|
| 61 |
+ |
|
| 62 |
+\item{nMaxA}{PRESENTLY UNUSED, future = limit number of atoms}
|
|
| 63 |
+ |
|
| 64 |
+\item{max_gibbmass_paraA}{limit truncated normal to max size}
|
|
| 65 |
+ |
|
| 66 |
+\item{alphaP}{sparsity parameter for P domain}
|
|
| 67 |
+ |
|
| 68 |
+\item{nMaxP}{PRESENTLY UNUSED, future = limit number of atoms}
|
|
| 69 |
+ |
|
| 70 |
+\item{max_gibbmass_paraP}{limit truncated normal to max size}
|
|
| 71 |
+} |
|
| 72 |
+\description{
|
|
| 73 |
+\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 74 |
+matrix factorization returning the two matrices that reconstruct |
|
| 75 |
+the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 76 |
+activity with nPerm set to 500 |
|
| 77 |
+ |
|
| 78 |
+Coordinated Gene Activity in Pattern Sets |
|
| 79 |
+} |
|
| 80 |
+ |
git-svn-id: https://hedgehog.fhcrc.org/bioconductor/trunk/madman/Rpacks/CoGAPS@107512 bc3139a8-67e5-0310-9ffc-ced21a209358
e.fertig authored on 17/08/2015 18:28:25 | ... | ... |
@@ -1,108 +1,119 @@ |
| 1 |
-% Generated by roxygen2 (4.0.1): do not edit by hand |
|
| 2 |
-\name{CoGAPS}
|
|
| 3 |
-\alias{CoGAPS}
|
|
| 4 |
-\title{\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 5 |
-matrix factorization returning the two matrices that reconstruct |
|
| 6 |
-the data matrix and then calls calcCoGAPSStat to estimate gene set |
|
| 7 |
-activity with nPerm set to 500} |
|
| 8 |
-\usage{
|
|
| 9 |
-CoGAPS(data, unc, GStoGenes, nFactor = "7", nEquil = 1000, nSample = 1000, |
|
| 10 |
- nOutR = 1000, output_atomic = "false", simulation_id = "simulation", |
|
| 11 |
- plot = TRUE, nPerm = 500, alphaA = "0.01", nMaxA = "100000", |
|
| 12 |
- max_gibbmass_paraA = "100.0", lambdaA_scale_factor = "1.0", |
|
| 13 |
- alphaP = "0.01", nMaxP = "100000", max_gibbmass_paraP = "100.0", |
|
| 14 |
- lambdaP_scale_factor = "1.0") |
|
| 15 |
-} |
|
| 16 |
-\arguments{
|
|
| 17 |
-\item{data}{data matrix}
|
|
| 18 |
- |
|
| 19 |
-\item{unc}{uncertainty matrix (std devs for chi-squared of Log Likelihood)}
|
|
| 20 |
- |
|
| 21 |
-\item{GStoGenes}{data.frame or list with gene sets}
|
|
| 22 |
- |
|
| 23 |
-\item{nFactor}{number of patterns (basis vectors, metagenes)}
|
|
| 24 |
- |
|
| 25 |
-\item{simulation_id}{name to attach to atoms files if created}
|
|
| 26 |
- |
|
| 27 |
-\item{plot}{logical to determine if plots produced}
|
|
| 28 |
- |
|
| 29 |
-\item{nPerm}{number of permutations for gene set test}
|
|
| 30 |
- |
|
| 31 |
-\item{nEquil}{number of iterations for burn-in}
|
|
| 32 |
- |
|
| 33 |
-\item{nSample}{number of iterations for sampling}
|
|
| 34 |
- |
|
| 35 |
-\item{nOutR}{how often to print status into R by iterations}
|
|
| 36 |
- |
|
| 37 |
-\item{output_atomic}{whether to write atom files (large)}
|
|
| 38 |
- |
|
| 39 |
-\item{alphaA}{sparsity parameter for A domain}
|
|
| 40 |
- |
|
| 41 |
-\item{alphaP}{sparsity parameter for P domain}
|
|
| 42 |
- |
|
| 43 |
-\item{max_gibbmass_paraA}{limit truncated normal to max size for A}
|
|
| 44 |
- |
|
| 45 |
-\item{max_gibbmass_paraP}{limit truncated normal to max size for P}
|
|
| 46 |
- |
|
| 47 |
-\item{nMaxA}{PRESENTLY UNUSED, future = limit number of atoms for A}
|
|
| 48 |
- |
|
| 49 |
-\item{nMaxP}{PRESENTLY UNUSED, future = limit number of atoms for P}
|
|
| 50 |
- |
|
| 51 |
-\item{lambdaA_scale_factor}{lambda factor in penalized likelihood for A}
|
|
| 52 |
- |
|
| 53 |
-\item{lambdaP_scale_factor}{lambda factor in penalized likelihood for P}
|
|
| 54 |
-} |
|
| 55 |
-\description{
|
|
| 56 |
-\code{CoGAPS} calls the C++ MCMC code through gapsRun and performs Bayesian
|
|
| 57 |
-matrix factorization returning the two matrices that reconstruct |
|
| 58 |
-the data matrix and then calls calcCoGAPSStat to estimate gene set |
|