Search for Corresponding Node in a Regression Tree Using Rpart

Find the data elements in a data frame that pass the rule for a node in a tree model?

I modified the code in path.rpart to return the subset of the data that falls within a particular node rather than returning information about that node. It works by either clicking on the plot or by passing nodes just as the path.rpart function does. Here is the code

subset.rpart <- function (tree, df, nodes) {
    if (!inherits(tree, "rpart")) 
        stop("Not a legitimate \"rpart\" object")
    stopifnot(nrow(df)==length(tree$where))
    frame <- tree$frame
    n <- row.names(frame)
    node <- as.numeric(n)

    if (missing(nodes)) {
        xy <- rpart:::rpartco(tree)
        i <- identify(xy, n = 1L, plot = FALSE)
        if(i> 0L) {
             return( df[tree$where==i, ] )
        } else {
            return(df[0,])
        }
    }
    else {
        if (length(nodes <- rpart:::node.match(nodes, node)) == 0L) 
            return(df[0,])
        return ( df[tree$where %in% as.numeric(nodes), ] )
    }
}

I will use it on some sample data from the package

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

And then to find the observations at a particular node, run

subset.rpart(fit, kyphosis)

and click on a node on the plot. After you do, all the observations at that node will be returned. You must use the same data.frame that was used for modeling for this to work properly. Rather than clicking on a point, you can also pass in a node name that you you discover with path.rpart

# path.rpart(fit)  
#  node number: 10  ---> looks interesting
#    root
#    Start>=8.5
#    Start< 14.5
#    Age< 55

subset.rpart(fit, kyphosis, 10)
#    Kyphosis Age Number Start
# 14   absent   1      4    12
# 20   absent  27      4     9
# 26   absent   9      5    13
# 37   absent   1      3     9
# 39   absent  20      6     9
# 42   absent  35      3    13
# 57   absent   2      3    13
# 59   absent  51      7     9
# 66   absent  17      4    10
# 69   absent  18      4    11
# 78   absent  26      7    13
# 81   absent  36      4    13

Extracting regression tree models built with R rpart

I believe what you are trying to do can be done using a little hack of the Rattle package/GUI. If you install Rattle then from the command line use the asRules() function on your rpart fit object and you will get back a human readable rules set. I have used these rules to quickly convert to SQL or other languages in seconds. I hope this helps.

rpart node assignment

Try using the partykit package:

library(rpart)
z.auto <- rpart(Mileage ~ Weight, car.test.frame)
library(partykit)
z.auto2 <- as.party(z.auto)
predict(z.auto2, car.test.frame[1:3,], type = "node")

# Eagle Summit 4 Ford Escort   4  Ford Festiva 4 
#              7               7               7 

How can I get the data in one terminal node of a decision tree in RPART

Try this:

fit$where[fit$where==3] # for Node 3
#  Hornet 4 Drive   Hornet Sportabout             Valiant          Duster 360           Merc 240D            Merc 280           Merc 280C 
#              3                   3                   3                   3                   3                   3                   3 
#     Merc 450SE          Merc 450SL         Merc 450SLC  Cadillac Fleetwood Lincoln Continental   Chrysler Imperial    Dodge Challenger 
#              3                   3                   3                   3                   3                   3                   3 
#    AMC Javelin          Camaro Z28    Pontiac Firebird 
              3                   3                   3 
length(fit$where[fit$where==3])
#[1] 17

mtcars[rownames(mtcars) %in% names(fit$where[fit$where==3]),]
#                     mpg cyl  disp  hp drat    wt  qsec vs am gear carb  x
#Hornet 4 Drive      21.4   6 258.0 110 3.08 3.215 19.44  1  0    3    1  4
#Hornet Sportabout   18.7   8 360.0 175 3.15 3.440 17.02  0  0    3    2  5
#Valiant             18.1   6 225.0 105 2.76 3.460 20.22  1  0    3    1  6
#Duster 360          14.3   8 360.0 245 3.21 3.570 15.84  0  0    3    4  7
#Merc 240D           24.4   4 146.7  62 3.69 3.190 20.00  1  0    4    2  8
#Merc 280            19.2   6 167.6 123 3.92 3.440 18.30  1  0    4    4 10
#Merc 280C           17.8   6 167.6 123 3.92 3.440 18.90  1  0    4    4 11
#Merc 450SE          16.4   8 275.8 180 3.07 4.070 17.40  0  0    3    3 12
#Merc 450SL          17.3   8 275.8 180 3.07 3.730 17.60  0  0    3    3 13
#Merc 450SLC         15.2   8 275.8 180 3.07 3.780 18.00  0  0    3    3 14
#Cadillac Fleetwood  10.4   8 472.0 205 2.93 5.250 17.98  0  0    3    4 15
#Lincoln Continental 10.4   8 460.0 215 3.00 5.424 17.82  0  0    3    4 16
#Chrysler Imperial   14.7   8 440.0 230 3.23 5.345 17.42  0  0    3    4 17
#Dodge Challenger    15.5   8 318.0 150 2.76 3.520 16.87  0  0    3    2 22
#AMC Javelin         15.2   8 304.0 150 3.15 3.435 17.30  0  0    3    2 23
#Camaro Z28          13.3   8 350.0 245 3.73 3.840 15.41  0  0    3    4 24
#Pontiac Firebird    19.2   8 400.0 175 3.08 3.845 17.05  0  0    3    2 25

compute numbers of leaf in rpart

The fit object has all the information that you need. You can examine it using str(fit).

Two ways to find the number of leaves are:

sum(fit$frame$ncompete == 0)
[1] 11

AND

sum(fit$frame$var == "<leaf>")
[1] 11

---

Related Topics