Titanic data with passenger names and other details removed.

Format

A data frame with 1046 observations on 6 variables.

`pclass`	passenger class, unordered factor: 1st 2nd 3rd
`survived`	factor: died or survived
`sex`	unordered factor: male female
`age`	age in years, min 0.167 max 80.0
`sibsp`	number of siblings or spouses aboard, integer: 0...8
`parch`	number of parents or children aboard, integer: 0...6

Source

The dataset was compiled by Frank Harrell and Robert Dawson:
http://biostat.mc.vanderbilt.edu/wiki/pub/Main/DataSets/titanic.html.

For this version of the Titanic data, passenger details were deleted, survived was cast as a factor, and the name changed to ptitanic to minimize confusion with other versions.

In this data the crew are conspicuous by their absence.

Contents of ptitanic:

         pclass survived    sex    age sibsp parch
    1       1st survived female 29.000     0     0
    2       1st survived   male  0.917     1     2
    3       1st     died female  2.000     1     2
    4       1st     died   male 30.000     1     2
    5       1st     died female 25.000     1     2
    ...
    1309    3rd     died   male 29.000     0     0

How ptitanic was built:

    load("titanic3.sav") # from Dr. Harrell's web site
    # discard name, ticket, fare, cabin, embarked, body, home.dest
    ptitanic <- titanic3[,c(1,2,4,5,6,7)]
    # change survived from integer to factor
    ptitanic$survived <- factor(ptitanic$survived, labels = c("died", "survived"))
    save(ptitanic, file = "ptitanic.rda")

This version of the data differs from etitanic in the earth package in that here survived is a factor (not an integer) and age has some NAs.

Examples

data(ptitanic)
summary(ptitanic)
#>  pclass        survived       sex           age              sibsp       
#>  1st:323   died    :809   female:466   Min.   : 0.1667   Min.   :0.0000  
#>  2nd:277   survived:500   male  :843   1st Qu.:21.0000   1st Qu.:0.0000  
#>  3rd:709                               Median :28.0000   Median :0.0000  
#>                                        Mean   :29.8811   Mean   :0.4989  
#>                                        3rd Qu.:39.0000   3rd Qu.:1.0000  
#>                                        Max.   :80.0000   Max.   :8.0000  
#>                                        NA's   :263                       
#>      parch      
#>  Min.   :0.000  
#>  1st Qu.:0.000  
#>  Median :0.000  
#>  Mean   :0.385  
#>  3rd Qu.:0.000  
#>  Max.   :9.000  
#>                 

# survival rate was greater for females
rpart.rules(rpart(survived ~ sex, data = ptitanic))
#>  survived                   
#>      0.19 when sex is   male
#>      0.73 when sex is female

# survival rate was greater for higher classes
rpart.rules(rpart(survived ~ pclass, data = ptitanic))
#>  survived                   
#>      0.26 when pclass is 3rd
#>      0.43 when pclass is 2nd
#>      0.62 when pclass is 1st

# survival rate was greater for children
rpart.rules(rpart(survived ~ age, data = ptitanic))
#>  survived                
#>      0.39 when age >= 8.5
#>      0.64 when age <  8.5

# main indicator of missing data is 3rd class esp. with many children
obs.with.nas <- rowSums(is.na(ptitanic)) > 0
rpart.rules(rpart(obs.with.nas ~ ., data = ptitanic, method = "class"))
#>  obs.with.nas                                       
#>          0.09 when pclass is 1st or 2nd             
#>          0.29 when pclass is        3rd & sibsp <  7
#>          0.89 when pclass is        3rd & sibsp >= 7