Browse code

Fix some bugs in the documentation

sherman5 authored on 30/06/2020 05:53:52
Showing1 changed files
... ...
@@ -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)}
Browse code

Fix up remaining errors during package check

Tom Sherman authored on 09/09/2019 23:42:33
Showing1 changed files
... ...
@@ -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{
Browse code

fix number of iterations used in vignette

Tom Sherman authored on 05/09/2019 20:47:09
Showing1 changed files
... ...
@@ -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,
Browse code

tidied up notation around pattern markers

Tom Sherman authored on 05/09/2019 18:14:53
Showing1 changed files
... ...
@@ -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)}
Browse code

update documentation

Tom Sherman authored on 24/06/2019 15:32:30
Showing1 changed files
... ...
@@ -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{
Browse code

still issues with file parsing

sherman5 authored on 19/06/2019 22:02:20
Showing1 changed files
... ...
@@ -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)}
Browse code

version bump

Tom Sherman authored on 22/02/2019 16:39:40
Showing1 changed files
... ...
@@ -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
 }
Browse code

allow rds for parameters; move all critical parametrers in CogapsParams

Tom Sherman authored on 18/02/2019 19:33:02
Showing1 changed files
... ...
@@ -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
Browse code

fixed bug when reading csv/tsv files with sparseOptimization

Tom Sherman authored on 03/01/2019 16:44:24
Showing1 changed files
... ...
@@ -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
 }
Browse code

added PUMP back to COGAPS

Tom Sherman authored on 21/12/2018 00:52:37
Showing1 changed files
... ...
@@ -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
 
Browse code

added fixedPatterns option for normal CoGAPS

Tom Sherman authored on 19/12/2018 14:59:41
Showing1 changed files
... ...
@@ -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{
Browse code

use boost pooled allocator instead of custom one

Tom Sherman authored on 12/11/2018 19:57:23
Showing1 changed files
... ...
@@ -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
 }
Browse code

cleaned up version and regenerated vignette

Tom Sherman authored on 02/11/2018 20:05:05
Showing1 changed files
... ...
@@ -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
 }
Browse code

updated config to commit file permissions

Tom Sherman authored on 29/10/2018 19:56:14
Showing1 changed files
1 1
old mode 100644
2 2
new mode 100755
Browse code

added more features to GWCoGAPS and scCoGAPS

Tom Sherman authored on 08/08/2018 22:34:56
Showing1 changed files
... ...
@@ -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{
Browse code

save all diagnostic data from GWCoGAPS/scCoGAPS

Tom Sherman authored on 06/08/2018 20:43:24
Showing1 changed files
... ...
@@ -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
Showing1 changed files
... ...
@@ -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{
Browse code

clean up output

Tom Sherman authored on 02/08/2018 16:52:31
Showing1 changed files
... ...
@@ -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
 }
Browse code

tests passing

Tom Sherman authored on 31/07/2018 18:45:03
Showing1 changed files
... ...
@@ -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, ...)
Browse code

moved file writers to FileParser

Tom Sherman authored on 30/07/2018 02:34:58
Showing1 changed files
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
+}
Browse code

updated documentation

Tom Sherman authored on 25/06/2018 20:35:29
Showing1 changed files
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
-
Browse code

cleaned up config files

Tom Sherman authored on 19/06/2018 19:15:53
Showing1 changed files
... ...
@@ -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
Browse code

compiling and running after merging with develop

Tom Sherman authored on 18/06/2018 22:20:06
Showing1 changed files
... ...
@@ -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
Showing0 changed files
Browse code

make sure nCores argument sent to c++ code

Tom Sherman authored on 18/06/2018 14:20:08
Showing1 changed files
... ...
@@ -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, ...)
Browse code

framework with dispatcher compiling and running

Tom Sherman authored on 18/06/2018 13:49:06
Showing1 changed files
... ...
@@ -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, ...)
Browse code

added back PUMP and FixedPatterns option

Tom Sherman authored on 13/06/2018 21:35:14
Showing1 changed files
... ...
@@ -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}
Browse code

passing checks

Tom Sherman authored on 05/06/2018 22:04:02
Showing1 changed files
... ...
@@ -51,6 +51,8 @@ the fixed patterns}
51 51
 
52 52
 \item{checkpointFile}{name of the checkpoint file}
53 53
 
54
+\item{nCores}{number of cpu cores to run in parallel over}
55
+
54 56
 \item{...}{keeps backwards compatibility with arguments from older versions}
55 57
 }
56 58
 \value{
Browse code

added back docs; fixed RNG syntax

Tom Sherman authored on 31/05/2018 21:34:33
Showing1 changed files
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
+
Browse code

make lintr happy

Tom Sherman authored on 31/05/2018 21:28:52
Showing1 changed files
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
-
Browse code

parallelization appears to be working

Tom Sherman authored on 09/05/2018 21:27:17
Showing1 changed files
... ...
@@ -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}
Browse code

making progress towards passing Bioc Checks

sherman5 authored on 19/04/2018 20:02:35
Showing1 changed files
... ...
@@ -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
+
Browse code

export patternMatch4parallel and ReOrderBySet

Genevieve Stein-O'Brien authored on 26/02/2018 17:11:49
Showing1 changed files
... ...
@@ -73,4 +73,3 @@ the data matrix
73 73
 data(SimpSim)
74 74
 result <- CoGAPS(SimpSim.D, SimpSim.S, nFactor=3, nOutputs=250)
75 75
 }
76
-
Browse code

passing checks

sherman5 authored on 07/02/2018 18:08:11
Showing1 changed files
... ...
@@ -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{
Browse code

nPumpSamples does something now

sherman5 authored on 06/02/2018 22:36:35
Showing1 changed files
... ...
@@ -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}
Browse code

pump working on C++ side, no fixed patterns

sherman5 authored on 06/02/2018 19:49:56
Showing1 changed files
... ...
@@ -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}
Browse code

parallel checkpoints mostly working - some bugs with processing output

sherman5 authored on 30/01/2018 20:50:37
Showing1 changed files
... ...
@@ -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{
Browse code

more warning clean up

sherman5 authored on 26/01/2018 20:18:09
Showing1 changed files
... ...
@@ -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
 
Browse code

get rid of most warnings

sherman5 authored on 26/01/2018 17:42:54
Showing1 changed files
... ...
@@ -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
Browse code

fixed up some documentation

sherman5 authored on 26/01/2018 00:56:10
Showing1 changed files
... ...
@@ -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
 
Browse code

installing properly again

sherman5 authored on 25/01/2018 21:13:26
Showing1 changed files
... ...
@@ -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}
Browse code

cleaning up notes for v3

sherman5 authored on 24/01/2018 23:38:02
Showing1 changed files
... ...
@@ -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
 
Browse code

added C++ unit tests

sherman5 authored on 06/09/2017 19:58:09
Showing1 changed files
... ...
@@ -73,3 +73,4 @@ matrix factorization returning the two matrices that reconstruct
73 73
 the data matrix and then calls calcCoGAPSStat to estimate gene set
74 74
 activity with nPerm set to 500
75 75
 }
76
+
Tom Sherman authored on 18/04/2017 18:38:46
Showing1 changed files
... ...
@@ -73,4 +73,3 @@ matrix factorization returning the two matrices that reconstruct
73 73
 the data matrix and then calls calcCoGAPSStat to estimate gene set
74 74
 activity with nPerm set to 500
75 75
 }
76
-
Browse code

Correct roxygen for package information.

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
Showing1 changed files
... ...
@@ -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
 
Browse code

Run `document`

Jacob Carey authored on 18/12/2015 21:07:44
Showing1 changed files
... ...
@@ -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
+
Browse code

Added new core for mapping version

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
Showing1 changed files
... ...
@@ -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