Title: | ClustAll: Data driven strategy to find groups of patients within complex diseases |
---|---|
Description: | Data driven strategy to find hidden groups of patients with complex diseases using clinical data. ClustAll facilitates the unsupervised identification of multiple robust stratifications. ClustAll, is able to overcome the most common limitations found when dealing with clinical data (missing values, correlated data, mixed data types). |
Authors: | Asier Ortega-Legarreta [aut, cre]
|
Maintainer: | Asier Ortega-Legarreta <[email protected]> |
License: | GPL-2 |
Version: | 1.1.0 |
Built: | 2024-07-17 11:36:03 UTC |
Source: | https://github.com/bioc/ClustAll |
Generic function to add validation data to the
ClustAllObject-class
object
addValidationData(Object, dataValidation)
addValidationData(Object, dataValidation)
Object |
|
dataValidation |
numericOrCharacter |
addValidationData
ClustAllObject-class
object
data("BreastCancerWisconsin", package = "ClustAll") label <- as.numeric(as.factor(wdbc$Diagnosis)) wdbc <- wdbc[,-c(1, 2)] # delete patients IDs & label obj_noNA <- createClustAll(data = wdbc) obj_noNA <- addValidationData(Object = obj_noNA, dataValidation = label)
data("BreastCancerWisconsin", package = "ClustAll") label <- as.numeric(as.factor(wdbc$Diagnosis)) wdbc <- wdbc[,-c(1, 2)] # delete patients IDs & label obj_noNA <- createClustAll(data = wdbc) obj_noNA <- addValidationData(Object = obj_noNA, dataValidation = label)
Contains either character, NULL or missing object
characterOrNA class object
Stores the original data used, the imputed datasets and the results of the ClustAll pipeline.
ClustAllObject class object
data
Data Frame of the data used. Maybe modified from the input data.
dataOriginal
Data Frame of the original data introduced.
dataImputed
Mids object derived from the mice package that stores the imputed data, in case imputation was applied. Otherwise NULL.
dataValidation
labelling numericOrNA. Original data labelling.
nImputation
Number of multiple imputations to be applied.
processed
Logical if the ClustAll pipeline has been executed previously
summary_clusters
listOrNULL. List with the resulting stratifications for each combination of clustering methods (distance + clustering algorithm) and depth, in case ClustAll pipeline has been executed previously. Otherwise NULL.
JACCARD_DISTANCE_F
matrixOrNULL. Matrix containing the Jaccard distances derived from the robust populations stratifications if ClustAll pipeline has been executed previously. Otherwise NULL.
Returns the original data in a dataframe, including the selected robust
stratification(s) as varaibles. The representative stratification names can
be obtained using the method. resStratification
cluster2data(Object, stratificationName)
cluster2data(Object, stratificationName)
Object |
|
stratificationName |
Character vector with one or more stratification names |
data.frame
resStratification
,plotJACCARD
,
ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE) df <- cluster2data(Object = obj_noNA1, stratificationName = c("cuts_a_1","cuts_b_5","cuts_a_5"))
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE) df <- cluster2data(Object = obj_noNA1, stratificationName = c("cuts_a_1","cuts_b_5","cuts_a_5"))
This pipeline creates the ClustAllObject and computes the imputations if the
dataset contains missing values. The next step would be
runClustAll
createClustAll(data=data, nImputation=NULL, dataImputed=NULL, colValidation=NULL)
createClustAll(data=data, nImputation=NULL, dataImputed=NULL, colValidation=NULL)
data |
Data Frame of the using data. It may contain missing (NA) values. |
nImputation |
Numeric value with the number of imputations to be computed in case the data contains NAs. |
dataImputed |
mids object created with mice package. The introduced data for the imputation and the data using must be the same. |
colValidation |
Character value with the original labelling of the input data. |
An object of class ClustAllObject-class
runClustAll
, ClustAllObject-class
# Scenario 1: data does not contain missing values data("BreastCancerWisconsin", package = "ClustAll") wdbc <- wdbc[,-c(1,2)] obj_noNA <- createClustAll(data = wdbc) # Scenario 2: data contains NAs and imputed data is provided automatically data("BreastCancerWisconsinMISSING", package = "ClustAll") # load example data obj_NA <- createClustAll(wdbcNA, nImputation = 5) # Scenario 3: data contains NAs and imputed data is provided manually data("BreastCancerWisconsinMISSING", package = "ClustAll") # load the example data ini <- mice::mice(wdbcNA, maxit = 0, print = FALSE) pred <- ini$pred # predictor matrix pred["radius1", c("perimeter1", "area1", "smoothness1")] <- 0 # example of how to remove predictors imp <- mice::mice(wdbcNA, m=5, pred=pred, maxit=5, seed=1234, print=FALSE) obj_imp <- createClustAll(data=wdbcNA, dataImputed = imp)
# Scenario 1: data does not contain missing values data("BreastCancerWisconsin", package = "ClustAll") wdbc <- wdbc[,-c(1,2)] obj_noNA <- createClustAll(data = wdbc) # Scenario 2: data contains NAs and imputed data is provided automatically data("BreastCancerWisconsinMISSING", package = "ClustAll") # load example data obj_NA <- createClustAll(wdbcNA, nImputation = 5) # Scenario 3: data contains NAs and imputed data is provided manually data("BreastCancerWisconsinMISSING", package = "ClustAll") # load the example data ini <- mice::mice(wdbcNA, maxit = 0, print = FALSE) pred <- ini$pred # predictor matrix pred["radius1", c("perimeter1", "area1", "smoothness1")] <- 0 # example of how to remove predictors imp <- mice::mice(wdbcNA, m=5, pred=pred, maxit=5, seed=1234, print=FALSE) obj_imp <- createClustAll(data=wdbcNA, dataImputed = imp)
Generic function to retrieve the imputed data obtained in
createClustAll
from a ClustAllObject-class
object
dataImputed(Object)
dataImputed(Object)
Object |
|
Mids class object with the imputed data or NULL if imputation was not required
createClustAll
, ClustAllObject-class
,
runClustAll
data("BreastCancerWisconsinMISSING", package = "ClustAll") data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_NA <- createClustAll(data = wdbcNA, colValidation = "Diagnosis", dataImputed = wdbcMIDS) dataImputed(obj_NA)
data("BreastCancerWisconsinMISSING", package = "ClustAll") data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_NA <- createClustAll(data = wdbcNA, colValidation = "Diagnosis", dataImputed = wdbcMIDS) dataImputed(obj_NA)
Generic function to retrieve the initial data used for
createClustAll
from a ClustAllObject-class
object
dataOriginal(Object)
dataOriginal(Object)
Object |
|
The Data Frame with the initial data
createClustAll
, ClustAllObject-class
,
runClustAll
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_noNA <- createClustAll(data = wdbc, colValidation = "Diagnosis") dataOriginal(obj_noNA)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_noNA <- createClustAll(data = wdbc, colValidation = "Diagnosis") dataOriginal(obj_noNA)
Generic function to retrieve numeric vector if it has been added with the
true labels from a ClustAllObject-class
object
dataValidation(Object)
dataValidation(Object)
Object |
|
numeric vector if true labels have been added. Otherwise NULL
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_noNA <- createClustAll(data = wdbc, colValidation="Diagnosis") dataValidation(obj_noNA)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_noNA <- createClustAll(data = wdbc, colValidation="Diagnosis") dataValidation(obj_noNA)
constuctor for ClustAllObject-class
## S4 method for signature 'ClustAllObject' initialize( .Object, data, dataOriginal, dataImputed, dataValidation, nImputation, processed, summary_clusters, JACCARD_DISTANCE_F )
## S4 method for signature 'ClustAllObject' initialize( .Object, data, dataOriginal, dataImputed, dataValidation, nImputation, processed, summary_clusters, JACCARD_DISTANCE_F )
.Object |
initializing object |
data |
Data Frame of the data used. Maybe modified from the input data. |
dataOriginal |
Data Frame of the original data introduced. |
dataImputed |
Mids object derived from the mice package that stores the imputed data, in case imputation was applied. Otherwise NULL. |
dataValidation |
labelling numericOrNA. Original data labelling. |
nImputation |
Number of multiple imputations to be applied. |
processed |
Logical if the ClustAll pipeline has been executed previously |
summary_clusters |
listOrNULL. List with the resulting stratifications for each combination of clustering methods (distance + clustering algorithm) and depth, in case ClustAll pipeline has been executed previously. Otherwise NULL. |
JACCARD_DISTANCE_F |
matrixOrNULL. Matrix containing the Jaccard distances derived from the robust populations stratifications if ClustAll pipeline has been executed previously. Otherwise NULL. |
An object of class ClustAllObject-class
Generic function to retrieve the matrix with the Jaccard distances derived
from the robust populations stratifications inrunClustAll
from
a ClustAllObject-class
object
JACCARD_DISTANCE_F(Object)
JACCARD_DISTANCE_F(Object)
Object |
|
Matrix containing the Jaccard distances derived from the robust populations stratifications or NULL if runClustAll method has not been executed yet
runClustAll
, ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = FALSE) JACCARD_DISTANCE_F(obj_noNA1)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = FALSE) JACCARD_DISTANCE_F(obj_noNA1)
Contains either list, NULL or missing object
Class union of list, null or missing
listOrNULL class object
Contains either logical, NULL or missing object
Class union of logical, null or missing
logicalOrNA class object
Contains either matrix or NULL object
Class union of matrix, null or missing
matrixOrNULL class object
Generic function to retrieve the number of imputations in
createClustAll
from a ClustAllObject-class
object
nImputation(Object)
nImputation(Object)
Object |
|
Numeric vector that contains the number of imputations. 0 in the case of no imputations were required
createClustAll
, ClustAllObject-class
,
runClustAll
data("BreastCancerWisconsinMISSING", package = "ClustAll") data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_NA <- createClustAll(data = wdbcNA, colValidation = "Diagnosis", dataImputed = wdbcMIDS) nImputation(obj_NA)
data("BreastCancerWisconsinMISSING", package = "ClustAll") data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_NA <- createClustAll(data = wdbcNA, colValidation = "Diagnosis", dataImputed = wdbcMIDS) nImputation(obj_NA)
Contains either numeric or character object
Class union of numericor character
numericOrCharacter class object
Contains either numeric, NULL or missing object
Class union of numeric, null or missing
numericOrNA class object
Processed wdbc as appear in vignette
data("testData", package = "ClustAll")
data("testData", package = "ClustAll")
A processed ClustAllObject
ClustAllObject Object
Processed wdbc as appear in vignette, with simplify TRUE parameter
data("testData", package = "ClustAll")
data("testData", package = "ClustAll")
A processed ClustAllObject
ClustAllObject Object
Processed wdbc as appear in vignette, with no validation data
data("testData", package = "ClustAll")
data("testData", package = "ClustAll")
A processed ClustAllObject
ClustAllObject Object
This function plots the correlation matrix heatmap showing the Jaccard Distance between robust stratifications
plotJACCARD(Object, paint=TRUE, stratification_similarity=0.7)
plotJACCARD(Object, paint=TRUE, stratification_similarity=0.7)
Object |
|
paint |
Logical vector with the annotation for the different stratifications |
stratification_similarity |
The minimum Jaccard Distance value to consider two stratifications similar. Default is 0.7. |
plot
resStratification
,cluster2data
,
ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) plotJACCARD(obj_noNA1, paint = TRUE, stratification_similarity = 0.9)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) plotJACCARD(obj_noNA1, paint = TRUE, stratification_similarity = 0.9)
This function plots the Sankey Diagram with the cluster distribution and shifts between a pair of stratifications
plotSANKEY(Object, clusters, validationData=FALSE)
plotSANKEY(Object, clusters, validationData=FALSE)
Object |
|
clusters |
Character vector with the names of a pair of stratifications. Check resStratification to obtain the stratification names. |
validationData |
Logical value to use original labelling data to compare with the ClustALL selected stratification. |
plot
resStratification
,cluster2data
,
ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") label <- as.numeric(as.factor(wdbc$Diagnosis)) wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] label <- label[16:30] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE) plotSANKEY(Object = obj_noNA1, clusters = c("cuts_a_1","cuts_b_5")) obj_noNA1 <- addValidationData(obj_noNA1, label) plotSANKEY(Object = obj_noNA1, clusters = "cuts_a_1", validationData=TRUE)
data("BreastCancerWisconsin", package = "ClustAll") label <- as.numeric(as.factor(wdbc$Diagnosis)) wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] label <- label[16:30] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE) plotSANKEY(Object = obj_noNA1, clusters = c("cuts_a_1","cuts_b_5")) obj_noNA1 <- addValidationData(obj_noNA1, label) plotSANKEY(Object = obj_noNA1, clusters = "cuts_a_1", validationData=TRUE)
Generic function to retrieve the logical if runClustAll
have
been runned from a ClustAllObject-class
object
processed(Object)
processed(Object)
Object |
|
TRUE if runClustAll has been already executed. Otherwise FALSE
runClustAll
, ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) processed(obj_noNA)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) processed(obj_noNA)
This function returns the stratifications representatives by keeping those clusters with a minimum percentage of the population. Default is 0.05. It returns all the robust stratification (TRUE) or the representative for each group of stratifications (FALSE). Default is FALSE
resStratification(Object, population=0.05, all=FALSE, stratification_similarity=0.7)
resStratification(Object, population=0.05, all=FALSE, stratification_similarity=0.7)
Object |
|
population |
Numeric vector with the minimum percentage of the total population that a stratification must have to be considered as representative |
all |
Logical vector to return all the representative stratifications per group of clusters. If it is FALSE, only the centroid stratification of each group of clusters is returned |
stratification_similarity |
The minimum Jaccard distance value to consider two groups similar. Default is 0.7 |
list
plotJACCARD
,cluster2data
,
ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE)
This method runs the ClustAll pipeline
runClustAll(Object, threads=1, simplify=FALSE)
runClustAll(Object, threads=1, simplify=FALSE)
Object |
|
threads |
Numeric vector that indicates the number of cores to use |
simplify |
if TRUE computes one out of four depths of the dendrogram |
An object of class ClustAllObject-class
resStratification
,plotJACCARD
,
cluster2data
,ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE)
Show method for a ClustAllObject-class
object
## S4 method for signature 'ClustAllObject' show(object)
## S4 method for signature 'ClustAllObject' show(object)
object |
|
summarize information about the object
Generic function to retrieve the initial data used for
createClustAll
from a ClustAllObject-class
object
showData(Object)
showData(Object)
Object |
|
The Data Frame with the initial data
createClustAll
, ClustAllObject-class
,
runClustAll
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_noNA <- createClustAll(data = wdbc, colValidation = "Diagnosis") showData(obj_noNA)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=-ID) obj_noNA <- createClustAll(data = wdbc, colValidation = "Diagnosis") showData(obj_noNA)
Generic function to retrieve the resulting stratifications for each
combination of clusterings method (distance + clustering algorithm) and
depth of runClustAll
from a ClustAllObject-class
object
summary_clusters(Object)
summary_clusters(Object)
Object |
|
List with the resulting stratifications for each combination of clusterings method (distance + clustering algorithm) and depth methods or NULL if runClustAll method has not been executed yet.
runClustAll
, ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = FALSE) summary_clusters(obj_noNA1)
data("BreastCancerWisconsin", package = "ClustAll") wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = FALSE) summary_clusters(obj_noNA1)
Returns the sensitivity and specifity of the selected stratification the
original lebelling. The representative stratification names can be obtained
using the method resStratification
validateStratification(Object, stratificationName)
validateStratification(Object, stratificationName)
Object |
|
stratificationName |
Character vector with the name a stratification. Check resStratification to obtain stratification names. |
numeric
resStratification
,plotJACCARD
,
ClustAllObject-class
data("BreastCancerWisconsin", package = "ClustAll") label <- as.numeric(as.factor(wdbc$Diagnosis)) wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] label <- label[16:30] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE) obj_noNA1 <- addValidationData(Object = obj_noNA1, dataValidation = label) validateStratification(obj_noNA1, "cuts_a_1")
data("BreastCancerWisconsin", package = "ClustAll") label <- as.numeric(as.factor(wdbc$Diagnosis)) wdbc <- subset(wdbc,select=c(-ID, -Diagnosis)) wdbc <- wdbc[1:15,1:8] label <- label[16:30] obj_noNA <- createClustAll(data = wdbc) obj_noNA1 <- runClustAll(Object = obj_noNA, threads = 1, simplify = TRUE) resStratification(Object = obj_noNA1, population = 0.05, stratification_similarity = 0.88, all = FALSE) obj_noNA1 <- addValidationData(Object = obj_noNA1, dataValidation = label) validateStratification(obj_noNA1, "cuts_a_1")
A dataset containing Features are computed from a digitized image of a fine needle aspirate (FNA) of a breast mass. They describe characteristics of the cell nuclei present in the image.
data("BreastCancerWisconsin", package = "ClustAll")
data("BreastCancerWisconsin", package = "ClustAll")
A data frame with 660 rows and 31 variables
The dataset comprises two types of features —categorical and numerical— derived from a digitized image of a fine needle aspirate (FNA) of a breast mass from 659 patients. Each patient is characterized by 31 features (10x3) and belongs to one of two target classes: ‘malignant’ or ‘benign’.
wdbc dataset
<https://archive.ics.uci.edu/dataset/17/breast+cancer+wisconsin+diagnostic>
Diagnosis Label says tumor is malingnant or benignant
radius. Mean of distances from the center to points on the perimeter
perimeter
area
smoothness. Local variation in radius lengths
compactness. (Perimeter^2 / Area) - 1.0
concavity. Severity of concave portions of the contour
concave points. Number of concave portions of the contour
symmetry.
fractal dimension. “Coastline approximation” - 1.
We introduced imputed random values to the wdbcNA dataset.
Using Mice. It is a mids object. wdbc
data("BreastCancerWisconsinMISSING", package = "ClustAll")
data("BreastCancerWisconsinMISSING", package = "ClustAll")
A data frame with 660 rows and 31 variables
wdbcMIDS dataset
We introduced random missing values to the wdbc dataset. wdbc
data("BreastCancerWisconsinMISSING", package = "ClustAll")
data("BreastCancerWisconsinMISSING", package = "ClustAll")
A data frame with 660 rows and 31 variables
wdbcNA dataset