20.17 Summary of the Model
## Call:
## rpart(formula=form, data=ds[tr, vars], model=TRUE)
## n= 145946
##
## CP nsplit rel error xerror xstd
## 1 0.15141265 0 1.0000000 1.0000000 0.005036230
## 2 0.03653233 1 0.8485874 0.8470732 0.004729950
## 3 0.03472826 2 0.8120550 0.8164363 0.004662043
## 4 0.01000000 3 0.7773268 0.7795496 0.004577084
##
## Variable importance
## humidity_3pm temp_3pm sunshine rainfall rain_today max_temp
## 77 5 4 4 3 3
## cloud_3pm humidity_9am
## 2 2
##
## Node number 1: 145946 observations, complexity param=0.1514126
## predicted class=No expected loss=0.2126883 P(node) =1
## class counts: 114905 31041
## probabilities: 0.787 0.213
## left son=2 (122288 obs) right son=3 (23658 obs)
## Primary splits:
## humidity_3pm < 71.5 to the left, improve=8441.871, (0 missing)
## rainfall < 0.75 to the left, improve=4936.526, (0 missing)
## rain_today splits as LR, improve=4890.129, (0 missing)
## cloud_3pm < 6.5 to the left, improve=3401.813, (0 missing)
## sunshine < 5.85 to the right, improve=2786.708, (0 missing)
## Surrogate splits:
## temp_3pm < 10.55 to the right, agree=0.847, adj=0.055, (0 split)
## sunshine < 0.45 to the right, agree=0.846, adj=0.051, (0 split)
## max_temp < 11.35 to the right, agree=0.843, adj=0.031, (0 split)
## rainfall < 30.5 to the left, agree=0.840, adj=0.015, (0 split)
## cloud_3pm < 7.5 to the left, agree=0.840, adj=0.011, (0 split)
##
## Node number 2: 122288 observations
## predicted class=No expected loss=0.1378876 P(node) =0.8378989
## class counts: 105426 16862
## probabilities: 0.862 0.138
##
## Node number 3: 23658 observations, complexity param=0.03653233
....In the following pages we dissect the various components of this summary.
Your donation will support ongoing availability and give you access to the PDF version of this book. Desktop Survival Guides include Data Science, GNU/Linux, and MLHub. Books available on Amazon include Data Mining with Rattle and Essentials of Data Science. Popular open source software includes rattle, wajig, and mlhub. Hosted by Togaware, a pioneer of free and open source software since 1984. Copyright © 1995-2022 Graham.Williams@togaware.com Creative Commons Attribution-ShareAlike 4.0
