Rpart Classification
result<-data.frame(rtree_classification$cptable)
result$nsplit<-factor(result$nsplit+1)
minimun_size<-as.numeric(as.character(result[which.min(result[,"xerror"]),"nsplit"]))
initial_model<-rtree_classification
model<-rpart::prune(rtree_classification,cp=rtree_classification$cptable[which.min(rtree_classification$cptable[,"xerror"]),"CP"])
importance<-model$variable.importance
importance<-data.frame(names=names(importance),importance=importance)
importance$names<-factor(importance$names,levels=rev(as.character(importance$names)))
plot_importance<-ggplot(importance,aes(x=names,y=importance))+
geom_bar(stat='identity')+
labs(title="Importance Plot",y="Relative Influence",x="Predictor")+
theme_bw(base_size=10)+
scale_x_discrete(limits=rev(levels(names)))+
coord_flip()
plot_importance
## n= 223
##
## node), split, n, loss, yval, (yprob)
## * denotes terminal node
##
## 1) root 223 74 0 (0.6681614 0.3318386)
## 2) spontaneous< 0.5 124 24 0 (0.8064516 0.1935484) *
## 3) spontaneous>=0.5 99 49 1 (0.4949495 0.5050505)
## 6) parity>=3.5 20 5 0 (0.7500000 0.2500000) *
## 7) parity< 3.5 79 34 1 (0.4303797 0.5696203)
## 14) age< 30.5 46 21 0 (0.5434783 0.4565217)
## 28) age>=28.5 9 2 0 (0.7777778 0.2222222) *
## 29) age< 28.5 37 18 1 (0.4864865 0.5135135)
## 58) spontaneous< 1.5 27 12 0 (0.5555556 0.4444444)
## 116) parity>=1.5 12 3 0 (0.7500000 0.2500000) *
## 117) parity< 1.5 15 6 1 (0.4000000 0.6000000) *
## 59) spontaneous>=1.5 10 3 1 (0.3000000 0.7000000) *
## 15) age>=30.5 33 9 1 (0.2727273 0.7272727) *
##
## Classification tree:
## rpart::rpart(formula = infert_formula, data = train_test_classification$f$train$f1,
## model = TRUE, x = TRUE, y = TRUE)
##
## Variables actually used in tree construction:
## [1] age parity spontaneous
##
## Root node error: 74/223 = 0.33184
##
## n= 223
##
## CP nsplit rel error xerror xstd
## 1 0.074324 0 1.00000 1.00000 0.095022
## 2 0.054054 2 0.85135 0.91892 0.092904
## 3 0.031532 3 0.79730 0.89189 0.092117
## 4 0.010000 6 0.70270 0.87838 0.091707
## Warning in rpart::rsq.rpart(model): may not be applicable for this method
error<-data.frame(model$cptable)
error$nsplit<-factor(error$nsplit+1)
tree_size<-error[which.min(error[,"xerror"]),"nsplit"]
error<-reshape2::melt(error,id.vars="nsplit")
names(error)<-c("Split","Metric","value")
plot_prune<-ggplot(error,aes(x=Split,y=value,color=Metric))+
geom_line(aes(group=Metric))+
geom_point()+
labs(title=paste("Error Plot","Suggested Size:",tree_size),y="Metric value",x="Size of Tree")+
theme_bw(base_size=10)
plot_prune
rpart.plot::rpart.plot(model,type=1)
frame<-data.frame(model$frame)
cp<-data.frame(model$cptable)
parameters<-data.frame(parameters=unlist(model$control))
splits<-data.frame(name=row.names(model$splits),model$splits,row.names=NULL)
importance<-data.frame(importance=model$variable.importance)
ordered<-data.frame(ordered=model$ordered)
data<-data.frame(y=model$y,x=model$x,model=model$model)
call<-data.frame(call=call_to_string(model))
result<-list(frame=frame,cp=cp,parameters=parameters,splits=splits,importance=importance,ordered=ordered,call=call)
print(result)
## $frame
## var n wt dev yval complexity ncompete nsurrogate yval2.V1 yval2.V2 yval2.V3 yval2.V4 yval2.V5 yval2.nodeprob
## 1 spontaneous 223 223 74 1 0.074324324 4 4 1.00000000 149.00000000 74.00000000 0.66816143 0.33183857 1.00000000
## 2 124 124 24 1 0.000000000 0 0 1.00000000 100.00000000 24.00000000 0.80645161 0.19354839 0.55605381
## 3 parity 99 99 49 2 0.074324324 4 2 2.00000000 49.00000000 50.00000000 0.49494949 0.50505051 0.44394619
## 6 20 20 5 1 0.000000000 0 0 1.00000000 15.00000000 5.00000000 0.75000000 0.25000000 0.08968610
## 7 age 79 79 34 2 0.054054054 4 0 2.00000000 34.00000000 45.00000000 0.43037975 0.56962025 0.35426009
## 14 age 46 46 21 1 0.031531532 4 0 1.00000000 25.00000000 21.00000000 0.54347826 0.45652174 0.20627803
## 28 9 9 2 1 0.010000000 0 0 1.00000000 7.00000000 2.00000000 0.77777778 0.22222222 0.04035874
## 29 spontaneous 37 37 18 2 0.031531532 4 0 2.00000000 18.00000000 19.00000000 0.48648649 0.51351351 0.16591928
## 58 parity 27 27 12 1 0.031531532 3 3 1.00000000 15.00000000 12.00000000 0.55555556 0.44444444 0.12107623
## 116 12 12 3 1 0.010000000 0 0 1.00000000 9.00000000 3.00000000 0.75000000 0.25000000 0.05381166
## 117 15 15 6 2 0.010000000 0 0 2.00000000 6.00000000 9.00000000 0.40000000 0.60000000 0.06726457
## 59 10 10 3 2 0.010000000 0 0 2.00000000 3.00000000 7.00000000 0.30000000 0.70000000 0.04484305
## 15 33 33 9 2 0.006756757 0 0 2.00000000 9.00000000 24.00000000 0.27272727 0.72727273 0.14798206
##
## $cp
## CP nsplit rel.error xerror xstd
## 1 0.07432432 0 1.0000000 1.0000000 0.09502215
## 2 0.05405405 2 0.8513514 0.9189189 0.09290437
## 3 0.03153153 3 0.7972973 0.8918919 0.09211655
## 4 0.01000000 6 0.7027027 0.8783784 0.09170670
##
## $parameters
## parameters
## minsplit 20.00
## minbucket 7.00
## cp 0.01
## maxcompete 4.00
## maxsurrogate 5.00
## usesurrogate 2.00
## surrogatestyle 0.00
## maxdepth 30.00
## xval 10.00
##
## $splits
## name count ncat improve index adj
## 1 spontaneous 223 -1 1.068327e+01 0.5 0.00000000
## 2 age 223 1 7.907546e-02 35.5 0.00000000
## 3 induced 223 -1 3.572929e-02 1.5 0.00000000
## 4 parity 223 -1 2.339838e-02 4.5 0.00000000
## 5 education 223 3 2.122571e-02 1.0 0.00000000
## 6 parity 0 -1 6.367713e-01 2.5 0.18181818
## 7 induced 0 1 6.233184e-01 0.5 0.15151515
## 8 age 0 1 5.695067e-01 24.5 0.03030303
## 9 education 0 3 5.605381e-01 2.0 0.01010101
## 10 parity 99 1 3.260772e+00 3.5 0.00000000
## 11 spontaneous 99 -1 2.585859e+00 1.5 0.00000000
## 12 age 99 -1 1.487542e+00 30.5 0.00000000
## 13 induced 99 1 2.727273e-01 0.5 0.00000000
## 14 education 99 3 2.430976e-01 3.0 0.00000000
## 15 induced 0 1 8.282828e-01 1.5 0.15000000
## 16 education 0 3 8.181818e-01 4.0 0.10000000
## 17 age 79 -1 2.817181e+00 30.5 0.00000000
## 18 spontaneous 79 -1 2.651214e+00 1.5 0.00000000
## 19 parity 79 1 5.822955e-01 1.5 0.00000000
## 20 induced 79 1 4.876055e-02 0.5 0.00000000
## 21 education 79 3 1.240302e-02 5.0 0.00000000
## 22 age 46 1 1.228489e+00 28.5 0.00000000
## 23 parity 46 1 9.356204e-01 1.5 0.00000000
## 24 spontaneous 46 -1 9.146650e-01 1.5 0.00000000
## 25 education 46 3 5.786225e-01 6.0 0.00000000
## 26 induced 46 1 1.129305e-04 0.5 0.00000000
## 27 spontaneous 37 -1 9.531532e-01 1.5 0.00000000
## 28 age 37 -1 7.007722e-01 26.5 0.00000000
## 29 parity 37 1 3.773956e-01 1.5 0.00000000
## 30 education 37 3 6.486486e-03 7.0 0.00000000
## 31 induced 37 1 5.005005e-03 0.5 0.00000000
## 32 parity 27 1 1.633333e+00 1.5 0.00000000
## 33 age 27 -1 1.096491e-01 24.5 0.00000000
## 34 induced 27 1 1.096491e-01 0.5 0.00000000
## 35 education 27 3 9.803922e-02 8.0 0.00000000
## 36 induced 0 1 8.518519e-01 0.5 0.66666667
## 37 age 0 1 7.037037e-01 23.5 0.33333333
## 38 education 0 3 6.296296e-01 9.0 0.16666667
##
## $importance
## importance
## spontaneous 11.6364186
## parity 6.8365175
## age 4.9138503
## induced 3.1966813
## education 0.7062112
##
## $ordered
## ordered
## age FALSE
## parity FALSE
## education FALSE
## spontaneous FALSE
## induced FALSE
##
## $call
## call
## 1 rpart::rpart(formula=infert_formula,data=train_test_classification$f$train$f1,,model=TRUE,x=TRUE,y=TRUE)
Rpart Regression
result<-data.frame(rtree_regression$cptable)
result$nsplit<-factor(result$nsplit+1)
minimun_size<-as.numeric(as.character(result[which.min(result[,"xerror"]),"nsplit"]))
initial_model<-rtree_regression
model<-rpart::prune(rtree_regression,cp=rtree_regression$cptable[which.min(rtree_regression$cptable[,"xerror"]),"CP"])
importance<-model$variable.importance
importance<-data.frame(names=names(importance),importance=importance)
importance$names<-factor(importance$names,levels=rev(as.character(importance$names)))
plot_importance<-ggplot(importance,aes(x=names,y=importance))+
geom_bar(stat='identity')+
labs(title="Importance Plot",y="Relative Influence",x="Predictor")+
theme_bw(base_size=10)+
scale_x_discrete(limits=rev(levels(names)))+
coord_flip()
plot_importance
## n= 455
##
## node), split, n, deviance, yval
## * denotes terminal node
##
## 1) root 455 36882.4700 22.27934
## 2) rm< 6.941 392 15593.1200 19.86454
## 4) lstat>=14.4 160 3093.8000 14.85125
## 8) nox>=0.607 95 1158.4370 12.65263
## 16) lstat>=19.645 53 469.1419 10.56415 *
## 17) lstat< 19.645 42 166.4040 15.28810 *
## 9) nox< 0.607 65 804.9686 18.06462 *
## 5) lstat< 14.4 232 5704.7180 23.32198
## 10) rm< 6.543 179 3125.3570 21.96313
## 20) crim< 7.24712 172 1449.2880 21.64651 *
## 21) crim>=7.24712 7 1235.1570 29.74286 *
## 11) rm>=6.543 53 1132.5530 27.91132 *
## 3) rm>=6.941 63 4780.4090 37.30476
## 6) rm< 7.4545 39 1376.0590 32.40513
## 12) lstat>=5.495 18 603.9711 28.97778 *
## 13) lstat< 5.495 21 379.4114 35.34286 *
## 7) rm>=7.4545 24 946.6933 45.26667 *
##
## Regression tree:
## rpart::rpart(formula = boston_formula, data = train_test_regression$f$train$f1,
## model = TRUE, x = TRUE, y = TRUE)
##
## Variables actually used in tree construction:
## [1] crim lstat nox rm
##
## Root node error: 36882/455 = 81.06
##
## n= 455
##
## CP nsplit rel error xerror xstd
## 1 0.447609 0 1.00000 1.00304 0.089852
## 2 0.184223 1 0.55239 0.63906 0.060331
## 3 0.066635 2 0.36817 0.39400 0.045195
## 4 0.039228 3 0.30153 0.35010 0.043677
## 5 0.030649 4 0.26231 0.31264 0.044314
## 6 0.014177 5 0.23166 0.26498 0.042395
## 7 0.011955 6 0.21748 0.25069 0.039729
## 8 0.010647 7 0.20552 0.24530 0.039714
## 9 0.010000 8 0.19488 0.24468 0.039712
error<-data.frame(model$cptable)
error$nsplit<-factor(error$nsplit+1)
tree_size<-error[which.min(error[,"xerror"]),"nsplit"]
error<-reshape2::melt(error,id.vars="nsplit")
names(error)<-c("Split","Metric","value")
plot_prune<-ggplot(error,aes(x=Split,y=value,color=Metric))+
geom_line(aes(group=Metric))+
geom_point()+
labs(title=paste("Error Plot","Suggested Size:",tree_size),y="Metric value",x="Size of Tree")+
theme_bw(base_size=10)
plot_prune
rpart.plot::rpart.plot(model,type=1)
frame<-data.frame(model$frame)
cp<-data.frame(model$cptable)
parameters<-data.frame(parameters=unlist(model$control))
splits<-data.frame(name=row.names(model$splits),model$splits,row.names=NULL)
importance<-data.frame(importance=model$variable.importance)
ordered<-data.frame(ordered=model$ordered)
data<-data.frame(y=model$y,x=model$x,model=model$model)
call<-data.frame(call=call_to_string(model))
result<-list(frame=frame,cp=cp,parameters=parameters,splits=splits,importance=importance,ordered=ordered,call=call)
print(result)
## $frame
## var n wt dev yval complexity ncompete nsurrogate
## 1 rm 455 455 36882.4658 22.27934 0.447609447 4 5
## 2 lstat 392 392 15593.1171 19.86454 0.184223027 4 5
## 4 nox 160 160 3093.7998 14.85125 0.030648555 4 5
## 8 lstat 95 95 1158.4368 12.65263 0.014177222 4 5
## 16 53 53 469.1419 10.56415 0.004942787 0 0
## 17 42 42 166.4040 15.28810 0.010000000 0 0
## 9 65 65 804.9686 18.06462 0.006818168 0 0
## 5 rm 232 232 5704.7179 23.32198 0.039227530 4 5
## 10 crim 179 179 3125.3566 21.96313 0.011954504 4 2
## 20 172 172 1449.2879 21.64651 0.007874345 0 0
## 21 7 7 1235.1571 29.74286 0.010000000 0 0
## 11 53 53 1132.5532 27.91132 0.008671637 0 0
## 3 rm 63 63 4780.4086 37.30476 0.066634814 4 5
## 6 lstat 39 39 1376.0590 32.40513 0.010646697 4 5
## 12 18 18 603.9711 28.97778 0.010000000 0 0
## 13 21 21 379.4114 35.34286 0.002712095 0 0
## 7 24 24 946.6933 45.26667 0.005777875 0 0
##
## $cp
## CP nsplit rel.error xerror xstd
## 1 0.44760945 0 1.0000000 1.0030382 0.08985226
## 2 0.18422303 1 0.5523906 0.6390649 0.06033078
## 3 0.06663481 2 0.3681675 0.3940014 0.04519524
## 4 0.03922753 3 0.3015327 0.3501001 0.04367740
## 5 0.03064856 4 0.2623052 0.3126368 0.04431441
## 6 0.01417722 5 0.2316566 0.2649762 0.04239499
## 7 0.01195450 6 0.2174794 0.2506874 0.03972945
## 8 0.01064670 7 0.2055249 0.2453018 0.03971366
## 9 0.01000000 8 0.1948782 0.2446765 0.03971197
##
## $parameters
## parameters
## minsplit 20.00
## minbucket 7.00
## cp 0.01
## maxcompete 4.00
## maxsurrogate 5.00
## usesurrogate 2.00
## surrogatestyle 0.00
## maxdepth 30.00
## xval 10.00
##
## $splits
## name count ncat improve index adj
## 1 rm 455 -1 0.44760945 6.941000 0.00000000
## 2 lstat 455 1 0.43832216 9.950000 0.00000000
## 3 indus 455 1 0.27457781 6.660000 0.00000000
## 4 ptratio 455 1 0.24752787 19.650000 0.00000000
## 5 nox 455 1 0.22601220 0.669500 0.00000000
## 6 lstat 0 1 0.89890110 4.830000 0.26984127
## 7 ptratio 0 1 0.88131868 14.150000 0.14285714
## 8 zn 0 -1 0.87252747 87.500000 0.07936508
## 9 indus 0 1 0.87252747 1.605000 0.07936508
## 10 crim 0 1 0.86373626 0.013355 0.01587302
## 11 lstat 392 1 0.43574350 14.400000 0.00000000
## 12 nox 392 1 0.28729760 0.669500 0.00000000
## 13 crim 392 1 0.25216185 5.848030 0.00000000
## 14 ptratio 392 1 0.23720257 19.900000 0.00000000
## 15 age 392 1 0.22884396 75.750000 0.00000000
## 16 age 0 1 0.82142857 84.600000 0.56250000
## 17 dis 0 -1 0.77806122 2.239350 0.45625000
## 18 nox 0 1 0.77551020 0.576500 0.45000000
## 19 indus 0 1 0.77295918 16.570000 0.44375000
## 20 tax 0 1 0.76275510 434.500000 0.41875000
## 21 nox 160 1 0.36537410 0.607000 0.00000000
## 22 crim 160 1 0.35899922 5.781900 0.00000000
## 23 tax 160 1 0.31905818 567.500000 0.00000000
## 24 dis 160 -1 0.29272223 1.990400 0.00000000
## 25 ptratio 160 1 0.29088455 19.900000 0.00000000
## 26 indus 0 1 0.87500000 16.010000 0.69230769
## 27 tax 0 1 0.86875000 397.000000 0.67692308
## 28 dis 0 -1 0.83750000 2.384050 0.60000000
## 29 crim 0 1 0.82500000 1.400920 0.56923077
## 30 rad 0 1 0.75000000 16.000000 0.38461538
## 31 lstat 95 1 0.45137628 19.645000 0.00000000
## 32 crim 95 1 0.42470492 10.452400 0.00000000
## 33 tax 95 1 0.21801780 551.500000 0.00000000
## 34 dis 95 -1 0.19770744 1.951250 0.00000000
## 35 rad 95 1 0.14739896 14.500000 0.00000000
## 36 dis 0 -1 0.78947368 1.842250 0.52380952
## 37 crim 0 1 0.74736842 8.951885 0.42857143
## 38 rm 0 -1 0.69473684 5.627500 0.30952381
## 39 nox 0 -1 0.67368421 0.706500 0.26190476
## 40 age 0 1 0.63157895 98.850000 0.16666667
## 41 rm 232 -1 0.25361605 6.543000 0.00000000
## 42 lstat 232 1 0.20737698 7.685000 0.00000000
## 43 dis 232 1 0.13514617 1.600900 0.00000000
## 44 chas 232 -1 0.07354552 0.500000 0.00000000
## 45 indus 232 1 0.06337127 4.100000 0.00000000
## 46 lstat 0 1 0.84913793 5.055000 0.33962264
## 47 zn 0 -1 0.79741379 31.500000 0.11320755
## 48 crim 0 1 0.78879310 0.017895 0.07547170
## 49 indus 0 1 0.78017241 4.010000 0.03773585
## 50 dis 0 -1 0.77586207 10.648000 0.01886792
## 51 crim 179 -1 0.14107561 7.247120 0.00000000
## 52 lstat 179 1 0.11894361 9.660000 0.00000000
## 53 dis 179 1 0.11625341 1.615600 0.00000000
## 54 rm 179 -1 0.07798460 6.142000 0.00000000
## 55 ptratio 179 1 0.05518417 20.550000 0.00000000
## 56 dis 0 1 0.98882682 1.551100 0.71428571
## 57 age 0 -1 0.96648045 99.450000 0.14285714
## 58 rm 63 -1 0.51411009 7.454500 0.00000000
## 59 lstat 63 1 0.29110784 5.185000 0.00000000
## 60 ptratio 63 1 0.20314402 19.700000 0.00000000
## 61 black 63 1 0.09352138 392.805000 0.00000000
## 62 crim 63 1 0.08962926 1.676395 0.00000000
## 63 lstat 0 1 0.74603175 3.990000 0.33333333
## 64 crim 0 -1 0.69841270 0.112760 0.20833333
## 65 ptratio 0 1 0.69841270 14.750000 0.20833333
## 66 black 0 1 0.68253968 390.065000 0.16666667
## 67 indus 0 -1 0.65079365 18.840000 0.08333333
## 68 lstat 39 1 0.28536309 5.495000 0.00000000
## 69 nox 39 1 0.14197199 0.488500 0.00000000
## 70 tax 39 1 0.11804967 267.000000 0.00000000
## 71 rad 39 1 0.11441342 7.500000 0.00000000
## 72 ptratio 39 1 0.10870099 18.900000 0.00000000
## 73 nox 0 1 0.74358974 0.470500 0.44444444
## 74 age 0 1 0.74358974 58.950000 0.44444444
## 75 dis 0 -1 0.74358974 3.546600 0.44444444
## 76 crim 0 1 0.69230769 0.102345 0.33333333
## 77 zn 0 -1 0.64102564 33.500000 0.22222222
##
## $importance
## importance
## rm 20575.2516
## lstat 13475.5477
## indus 5367.3187
## dis 4568.9258
## nox 4499.4345
## age 4146.6209
## tax 3610.4285
## ptratio 2870.4317
## crim 2322.6065
## zn 1561.2844
## rad 434.7670
## black 409.6094
##
## $ordered
## ordered
## crim FALSE
## zn FALSE
## indus FALSE
## chas FALSE
## nox FALSE
## rm FALSE
## age FALSE
## dis FALSE
## rad FALSE
## tax FALSE
## ptratio FALSE
## black FALSE
## lstat FALSE
##
## $call
## call
## 1 rpart::rpart(formula=boston_formula,data=train_test_regression$f$train$f1,,model=TRUE,x=TRUE,y=TRUE)