improper Bagging Subsample Survival Tree

Description

Fit a bagging survival tree on a mixture of population (susceptible and nonsusceptible)using either a pseudo R2 criterion or an adjusted Logrank criterion. The predictor is evaluated using the Out Of Bag Integrated Brier Score (IBS) and several scores of importanceare computed for variable selection. The thresholds values for variable selection are computed using a nonparametric permutation test. See Cyprien Mbogning and Philippe Broet (2016)<doi:10.1186/s12859-016-1090-x> for an overview about the methods implemented in this package.

Details

Author(s)

Cyprien Mbogning and Philippe Broet

Maintainer: Cyprien Mbogning <cyprien.mbogning@gmail.com>

References

Mbogning, C. and Broet, P. (2016). Bagging survival tree procedure for variable selection and prediction in the presence of nonsusceptible patients. BMC bioinformatics, 17(1), 1.

Duhaze Julianne et al. (2020). A Machine Learning Approach for High-Dimensional Time-to-Event Prediction With Application to Immunogenicity of Biotherapies in the ABIRISK Cohort. Frontiers in Immunology, 11.

See Also

Bagg_Surv Bagg_pred_Surv improper_tree

Examples

## Not run: 
 data(burn)
 myarg = list(cp = 0, maxcompete = 0, maxsurrogate = 0, maxdepth = 2)
 Y.names = c("T3" ,"D3")
 P.names = 'Z2'
 T.names = c("Z1", paste("Z", 3:11, sep = ''))
 mybag = 40
 feat_samp = length(T.names)
 set.seed(5000)
 
 ## fit an improper survival tree
 burn.tree <- suppressWarnings(improper_tree(burn, 
    Y.names, 
    P.names, 
    T.names, 
    method = "R2", 
    args.rpart = myarg))
    
 plot(burn.tree)
 text(burn.tree, cex = .7, xpd = TRUE)
 
 ## fit an improper Bagging survival tree with the adjusted Logrank criterion
 burn.BagEssai0 <- suppressWarnings(Bagg_Surv(burn, 
    Y.names, 
    P.names, 
    T.names, 
    method = "LR", 
    args.rpart = myarg, 
    args.parallel = list(numWorkers = 1), 
    Bag = mybag, feat = feat_samp))
 
 ## fit an improper Bagging survival tree with the pseudo R2 criterion
 burn.BagEssai1 <- suppressWarnings(Bagg_Surv(burn, 
    Y.names, 
    P.names, 
    T.names, 
    method = "R2", 
    args.rpart = myarg, 
    args.parallel = list(numWorkers = 1), 
    Bag = mybag, feat = feat_samp))

 ## Plot the variable importance scores
 par(mfrow=c(1,3))
barplot(burn.BagEssai1$IIS, 
   main = 'IIS', 
   horiz = TRUE, 
   las = 1,
   cex.names = .8, 
   col = 'lightblue')
   
barplot(burn.BagEssai1$DIIS, 
   main = 'DIIS', 
   horiz = TRUE, 
   las = 1,
   cex.names = .8, 
   col = 'grey') 
   
barplot(burn.BagEssai1$DEPTH, 
   main = 'MinDepth', 
   horiz = TRUE, 
   las = 1,
   cex.names = .8, 
   col = 'purple')


 ## evaluation of the Bagging predictors 
pred0 <- suppressWarnings(Bagg_pred_Surv(burn, 
   Y.names, 
   P.names, 
   burn.BagEssai0, 
   args.parallel = list(numWorkers = 1), 
   OOB = TRUE)) 
 
 
 pred1 <- suppressWarnings(Bagg_pred_Surv(burn, 
   Y.names, 
   P.names, 
   burn.BagEssai1, 
   args.parallel = list(numWorkers = 1), 
   OOB = TRUE)) 
 
## End(Not run)

Bagging improper survival trees

Description

Bagging sunbsampling procedure to aggregate several improper trees using either the pseudo-R2 procedure or the adjusted Logrank procedure. Several scores for variables importance are computed.

Usage

Bagg_Surv(xdata, 
     Y.names, 
     P.names, 
     T.names, 
     method = "R2", 
     args.rpart, 
     args.parallel = list(numWorkers = 1), 
     Bag = 100, feat = 5)

Arguments

Details

For the Bagging procedure, it is mendatory to set maxcompete = 0 and maxsurrogate = 0 within the args.rpart arguments. This will ensured the correct calculation of the importance of variables and also a better computation time.

Value

A list of ten elements

Note

This version of the code allows for the moment only one variable to have an impact on the cured population.The next version will allow more than one variable.

Author(s)

Cyprien Mbogning and Philippe Broet

References

Mbogning, C. and Broet, P. (2016). Bagging survival tree procedure for variable selection and prediction in the presence of nonsusceptible patients. BMC bioinformatics, 17(1), 1.

Duhaze Julianne et al. (2020). A Machine Learning Approach for High-Dimensional Time-to-Event Prediction With Application to Immunogenicity of Biotherapies in the ABIRISK Cohort. Frontiers in Immunology, 11.

See Also

Bagg_pred_Surv

Examples

## Not run: 
 data(burn)
 myarg = list(cp = 0, maxcompete = 0, maxsurrogate = 0, maxdepth = 2)
 Y.names = c("T3" ,"D3")
 P.names = 'Z2'
 T.names = c("Z1", paste("Z", 3:11, sep = ''))
 mybag = 40
 feat_samp = length(T.names)
 set.seed(5000)
 
 burn.BagEssai0 <- suppressWarnings(Bagg_Surv(burn, 
    Y.names, 
    P.names, 
    T.names, 
    method = "LR", 
    args.rpart = myarg, 
    args.parallel = list(numWorkers = 1), 
    Bag = mybag, feat = feat_samp))
 
 burn.BagEssai1 <- suppressWarnings(Bagg_Surv(burn, 
    Y.names, 
    P.names, 
    T.names, 
    method = "R2", 
    args.rpart = myarg, 
    args.parallel = list(numWorkers = 1), 
    Bag = mybag, feat = feat_samp))


## End(Not run)

Bagging survival tree prediction

Description

Use the Bagging improper survival tree to predict on new features and to evaluate the predictor using Out Of Bag Integrated Brier Scores with either the Nelson Aalen estimator or the Breslow estimator. A permutation importance score is also computed using OOB observations.

Usage

Bagg_pred_Surv(xdata, Y.names, P.names, resBag, args.parallel = list(numWorkers = 1), 
               new_data = data.frame(), OOB = FALSE)

Arguments

Value

Author(s)

Cyprien Mbogning and Philippe Broet

References

Mbogning, C. and Broet, P. (2016). Bagging survival tree procedure for variable selection and prediction in the presence of nonsusceptible patients. BMC bioinformatics, 17(1), 1.

Duhaze Julianne et al. (2020). A Machine Learning Approach for High-Dimensional Time-to-Event Prediction With Application to Immunogenicity of Biotherapies in the ABIRISK Cohort. Frontiers in Immunology, (11).

See Also

Bagg_Surv

Examples

## Not run: 
 data(burn)
 myarg = list(cp = 0, maxcompete = 0, maxsurrogate = 0, maxdepth = 2)
 Y.names = c("T3" ,"D3")
 P.names = 'Z2'
 T.names = c("Z1", paste("Z", 3:11, sep = ''))
 mybag = 40
 feat_samp = length(T.names)
 set.seed(5000)
 
 burn.BagEssai0 <- suppressWarnings(Bagg_Surv(burn, 
        Y.names, 
        P.names, 
        T.names, 
        method = "LR", 
        args.rpart = myarg, 
        args.parallel = list(numWorkers = 1), 
        Bag = mybag, feat = feat_samp))
 
 burn.BagEssai1 <- suppressWarnings(Bagg_Surv(burn, 
        Y.names, 
        P.names, 
        T.names, 
        method = "R2", 
        args.rpart = myarg, 
        args.parallel = list(numWorkers = 1), 
        Bag = mybag, feat = feat_samp))

pred0 <- Bagg_pred_Surv(burn, 
    Y.names, 
    P.names, 
    burn.BagEssai0, 
    args.parallel = list(numWorkers = 1), 
    OOB = TRUE) 
 
 
 pred1 <- Bagg_pred_Surv(burn, 
    Y.names, 
    P.names, 
    burn.BagEssai1, 
    args.parallel = list(numWorkers = 1), 
    OOB = TRUE) 
 
## End(Not run)

Pseudo R2 criterion

Description

Pseudo R2 criterion for a mixture of population (susceptible and nonsusceptible populations)

Usage

PseudoR2.Cure(ygene, ydelai, yetat, strate, ordered = FALSE)

Arguments

Value

A pseudo R2 value lying between 0 and 1.

Author(s)

Cyprien Mbogning and Philippe Broet

References

Mbogning, C. and Broet, P. (2016). Bagging survival tree procedure for variable selection and prediction in the presence of nonsusceptible patients. BMC bioinformatics, 17(1), 1.

See Also

Bagg_Surv Bagg_pred_Surv improper_tree

Examples

data(burn)
PseudoR2.Cure(ygene = burn$Z3, 
   ydelai = burn$T3, 
   yetat = burn$D3, 
   strate = burn$Z2)
   
PseudoR2.Cure(ygene = burn$Z2, 
   ydelai = burn$T3, 
   yetat = burn$D3, 
   strate = burn$Z2)

burn dataset

Description

The burn data frame has 154 rows and 17 columns.

Usage

data(burn)

Format

A data frame with 154 observations on the following 17 variables.

Obs

Observation number

Z1

Treatment: 0-routine bathing 1-Body cleansing

Z2

Gender (0=male 1=female)

Z3

Race: 0=nonwhite 1=white

Z4

Percentage of total surface area burned

Z5

Burn site indicator: head 1=yes, 0=no

Z6

Burn site indicator: buttock 1=yes, 0=no

Z7

Burn site indicator: trunk 1=yes, 0=no

Z8

Burn site indicator: upper leg 1=yes, 0=no

Z9

Burn site indicator: lower leg 1=yes, 0=no

Z10

Burn site indicator: respiratory tract 1=yes, 0=no

Z11

Type of burn: 1=chemical, 2=scald, 3=electric, 4=flame

T1

Time to excision or on study time

D1

Excision indicator: 1=yes 0=no

T2

Time to prophylactic antibiotic treatment or on study time

D2

Prophylactic antibiotic treatment: 1=yes 0=no

T3

Time to straphylocous aureaus infection or on study time

D3

Straphylocous aureaus infection: 1=yes 0=no

Source

Klein and Moeschberger (1997) Survival Analysis Techniques for Censored and truncated data, Springer.

Ichida et al. Stat. Med. 12 (1993): 301-310.

Examples

data(burn)
## maybe str(burn) ;

imprper survival tree

Description

Fit an improper survival tree for the mixed population (susceptible and nonsusceptible) using either the proposed pseudo R2 criterion or an adjusted Logrank criterion

Usage

improper_tree(xdata, 
     Y.names, 
     P.names, 
     T.names, 
     method = "R2", 
     args.rpart)

Arguments

Value

An unprunned improper survival tree

Author(s)

Cyprien Mbogning and Philippe Broet

References

Mbogning, C. and Broet, P. (2016). Bagging survival tree procedure for variable selection and prediction in the presence of nonsusceptible patients. BMC bioinformatics, 17(1), 1.

See Also

Bagg_Surv Bagg_pred_Surv

Examples

## Not run: 
 data(burn)
 myarg = list(cp = 0, maxcompete = 0, maxsurrogate = 0, maxdepth = 3)
 Y.names = c("T3" ,"D3")
 P.names = 'Z2'
 T.names = c("Z1", paste("Z", 3:11, sep = ''))
 burn.tree <- suppressWarnings(improper_tree(burn, 
    Y.names, 
    P.names, 
    T.names, 
    method = "R2", 
    args.rpart = myarg))
    
 plot(burn.tree)
 text(burn.tree, cex = .7, xpd = TRUE)
 
## End(Not run)

permutation variable selection

Description

Variable selection using the permutation test on several scores of importance: IIS, DIIS and DEPTH.

Usage

permute_select_surv(xdata, 
    Y.names, 
    P.names, 
    T.names, 
    importance = "IIS", 
    method = "R2",
    Bag, 
    args.rpart, 
    args.parallel = list(numWorkers = 1), 
    nperm = 50)

Arguments

Details

Testing weither the importance score is null or not.

Value

A list of five elements:

Author(s)

Cyprien Mbogning and Philippe Broet

References

Mbogning, C. and Broet, P. (2016). Bagging survival tree procedure for variable selection and prediction in the presence of nonsusceptible patients. BMC bioinformatics, 17(1), 1.

See Also

Bagg_Surv Bagg_pred_Surv

Examples

## Not run: 
 myarg = list(cp = 0, maxcompete = 0, maxsurrogate = 0, maxdepth = 2)
 Y.names = c("T3" ,"D3")
 P.names = 'Z2'
 T.names = c("Z1", paste("Z", 3:11, sep = ''))
 mybag = 40
 set.seed(5000)
 
 data(burn)
 resperm0 <- suppressWarnings(permute_select_surv(xdata = burn, 
       Y.names, 
       P.names, 
       T.names, 
       method = "LR", 
       Bag = mybag, 
       args.rpart = myarg, 
       args.parallel = list(numWorkers = 1), 
       nperm = 150))
 
## End(Not run)

Simple function using Rcpp

Description

Simple function using Rcpp

Usage

rcpp_hello_world()	

Examples

## Not run: 
rcpp_hello_world()

## End(Not run)

From a tree to indicators (or dummy variables)

Description

Coerces a given tree structure inheriting from rpart to binary covariates.

Usage

tree2indicators(fit)

Arguments

Value

a list of indicators defining the leaf nodes of the fitted tree from left to right

Author(s)

Cyprien Mbogning

Examples

fit <- rpart(Kyphosis ~ Age + Number + Start, data = kyphosis)
tree2indicators(fit)