santhosh project
santhosh
4/20/2022
R Markdown
This is an R Markdown document. Markdown is a simple formatting syntax for authoring HTML, PDF, and MS Word documents. For more details on using R Markdown see http://rmarkdown.rstudio.com.
When you click the Knit button a document will be generated that includes both content as well as the output of any embedded R code chunks within the document. You can embed an R code chunk like this:
library(C50)
library(rpart)
library(rpart.plot)
library(gmodels)
NewResin_Train <- read.csv("C:/Users/gujjes/Downloads/MLTraining(1).csv", header=TRUE)
set.seed(12345)
resin2_rand <- NewResin_Train[order(runif(340)),]
resin2_train <-resin2_rand[1:270,]
resin2_test<-resin2_rand[271:340,]
table(resin2_train$Final.Resin.Product)##
## 34.9928 44.6572 44.8958 45.2194 48.1125 48.2656 48.5985 48.9419 49.6108 49.8733
## 1 1 1 1 1 1 1 1 1 1
## 50.0562 50.539 50.7018 50.8264 51.0388 51.0645 51.4173 51.6389 51.7499 51.7542
## 1 1 3 1 1 1 1 1 1 1
## 51.9275 52.0255 52.2982 52.4 52.6632 52.6775 52.8568 53.603 53.6901 53.7431
## 1 1 1 1 1 1 1 1 1 1
## 53.7647 53.818 54.2838 54.7161 54.7379 54.7551 54.9143 54.9633 55.1662 55.3368
## 1 1 1 1 1 1 1 1 1 1
## 55.384 55.6709 55.7156 55.9854 55.9974 56.1077 56.1411 56.2317 56.2341 56.2578
## 1 1 1 1 1 1 1 1 1 1
## 56.2895 56.3344 56.4587 56.4832 56.8166 56.8912 56.9286 57.002 57.0878 57.2828
## 1 1 1 1 1 1 1 1 5 1
## 57.3665 57.3837 57.4046 57.4496 57.4905 57.5374 57.5674 57.5757 57.8292 57.8816
## 1 1 1 1 1 1 1 1 1 1
## 57.9132 57.9862 58.0684 58.2107 58.4465 58.5547 58.5715 58.6142 58.6497 58.7121
## 1 1 1 2 1 1 1 1 1 1
## 58.7922 58.8721 58.9861 59.0097 59.0224 59.0736 59.0983 59.1129 59.1402 59.1799
## 1 1 1 1 1 1 1 1 2 1
## 59.2057 59.2104 59.228 59.2459 59.2728 59.2894 59.4259 59.447 59.4888 59.57
## 1 1 1 1 1 1 1 1 1 1
## 59.5771 59.6041 59.6207 59.6351 59.7572 59.8172 60.0028 60.0155 60.0915 60.1198
## 1 1 1 1 1 2 1 1 2 1
## 60.3389 60.348 60.3841 60.4143 60.5704 60.5905 60.5985 60.6612 60.7256 60.8884
## 2 1 1 1 2 1 1 1 1 1
## 60.8994 60.934 61.034 61.046 61.1001 61.2245 61.2518 61.2966 61.3639 61.5648
## 2 2 1 1 1 1 1 1 1 2
## 61.5701 61.6041 61.6135 61.6949 61.7537 61.8115 61.8191 61.9076 61.9088 61.9613
## 1 1 1 1 1 2 1 1 1 1
## 61.9674 62.0589 62.1702 62.2274 62.3344 62.4202 62.4287 62.5098 62.52 62.6006
## 1 1 1 1 1 1 1 1 1 1
## 62.6577 62.6643 62.6892 62.7188 62.7826 62.8249 62.8315 62.8651 62.8687 62.8756
## 1 1 1 1 1 1 1 1 1 2
## 62.99 63.1088 63.1921 63.2418 63.3183 63.4132 63.5061 63.5566 63.5663 63.5726
## 1 1 1 1 1 1 1 1 1 1
## 63.5974 63.7812 63.8119 63.828 63.9541 64.0708 64.0758 64.1042 64.1197 64.2108
## 1 1 1 1 1 1 2 1 1 1
## 64.2601 64.2815 64.36 64.3693 64.5671 64.6052 64.814 64.8494 64.8729 64.9631
## 1 1 1 1 1 1 1 1 1 1
## 65.0551 65.0753 65.0813 65.095 65.2009 65.3427 65.3517 65.3891 65.4238 65.4736
## 1 1 1 1 1 1 1 1 2 1
## 65.5486 65.668 65.7484 65.756 65.8051 65.8464 66.2678 66.2943 66.6801 66.7297
## 1 1 1 1 1 1 1 1 1 1
## 66.8832 66.9857 67.0707 67.3765 67.4257 67.4869 67.7875 67.9357 67.9925 68.0716
## 1 1 1 1 1 1 1 1 1 1
## 68.0742 68.2574 68.2739 68.4055 68.5176 68.7697 68.8784 68.9225 68.9638 68.9814
## 1 1 1 1 1 1 1 1 1 1
## 69.279 69.2873 69.7735 69.8379 70.1889 70.9129 70.9775 70.9784 71.2017 71.3125
## 1 1 1 1 1 1 1 1 1 1
## 71.3756 71.7253 72.3614 72.4389 73.0963 73.326 74.1765 74.3419 75.1404 75.5366
## 1 1 1 1 1 1 1 1 1 1
## 76.5254
## 1table(resin2_test$Final.Resin.Product)##
## 47.5572 48.3608 48.6229 49.5325 52.1271 53.8438 54.0355 54.5898 54.9711 55.0156
## 1 1 1 1 1 1 1 1 1 1
## 55.0341 55.6709 55.7981 55.8494 56.3987 56.8221 57.008 57.0878 57.1346 57.3655
## 1 1 1 1 1 1 1 1 1 1
## 57.3984 57.5674 58.0194 58.1832 58.19 58.3068 58.5035 58.5396 58.6643 58.8232
## 1 1 1 1 1 1 1 1 1 1
## 58.9058 58.9658 60.466 60.8805 60.8994 60.9887 61.0097 61.0295 61.0485 61.3695
## 1 1 1 1 1 1 1 1 1 1
## 61.4811 61.5819 61.6108 61.6857 61.7275 61.8042 61.9567 62.6759 62.7188 62.8934
## 1 1 1 2 1 1 1 1 1 1
## 63.006 63.3174 63.3907 63.5974 63.6159 63.8075 64.0664 64.2346 64.769 65.3747
## 1 1 1 1 1 1 1 1 1 1
## 66.4547 67.2839 67.3574 67.8877 68.2994 68.6541 69.0085 72.9945 78.5917
## 1 1 1 1 1 1 1 1 1str(resin2_test)## 'data.frame': 70 obs. of 35 variables:
## $ TreeSite : chr "Mix Harvest" "City Harvest" "City Harvest" "Village Harvest" ...
## $ Pressure : num 3 0 0 2 3 2 0 1 2 4 ...
## $ Step1.Temp : int 18 18 18 18 18 18 18 18 18 18 ...
## $ Step1.conductivity : num 13.5 13.2 13.5 13.2 13.5 13.5 13.5 11.4 13.5 12.8 ...
## $ Step1.Resin : num 14049 14064 13496 13272 14266 ...
## $ Step2.pH : num 5.6 5.66 5.6 5.64 5.67 5.64 5.64 5.67 5.62 5.6 ...
## $ Step2.Conductivity : int 49 50 46 49 50 50 50 48 48 49 ...
## $ Step2.Yield : int 8 9 10 8 8 9 8 8 9 7 ...
## $ Step3.Temp : num 19.7 19.7 19 19.7 19.7 ...
## $ Step3.pH : num 5.64 5.68 5.66 5.65 5.68 5.68 5.67 5.68 5.68 5.63 ...
## $ Step3.Conductivity : int 40 40 41 40 40 40 40 40 40 41 ...
## $ Step3.yield : int 18 15 14 17 15 17 14 16 15 17 ...
## $ Step3.compression : num 1.18 1.18 1.18 1.16 1.18 ...
## $ Step4.pH : num 5.47 7.69 7.2 7.55 5.9 6.61 5.33 6.81 5.6 7.27 ...
## $ Step5.yield : num 77 81 83 82 78 80 79 79 81 78 ...
## $ Step5.temp : num 18 19 18 19 19 19 18 19 19 20 ...
## $ Step6.pH.start : num 4.61 4.69 4.62 4.69 4.61 4.63 4.67 4.62 4.64 4.63 ...
## $ Step6.Conductivity.start: num 106 107 105 106 120 ...
## $ Step6.temperature : int 17 18 18 18 17 17 18 17 18 18 ...
## $ Step7.symmetry : num 1.8 1.75 1.69 9.67 1.12 1.52 1.61 1.12 1.8 1.46 ...
## $ Step8.temperature : num 18.5 18.5 19 18.5 18 ...
## $ step.8.yield : int 91 92 92 92 92 93 94 92 93 90 ...
## $ Step7.UV : num 3.01 2.97 2.98 3.02 3 ...
## $ Step8.UV : num 0.692 0.694 0.72 0.69 0.712 ...
## $ Step.1.pH : num 7.18 7.2 7.24 7.1 7.2 ...
## $ step.1.particles : num 2647813 2780000 3255000 33000000 3059091 ...
## $ step.1.quality. : num 93.6 91.8 86 95 90 ...
## $ step0.total.particles : num 998582 920513 1060000 850000 1100000 ...
## $ step0.warmup.hrs : num 16.9 16.9 16.9 16 16.9 ...
## $ step0.input.pH : num 7.2 7.1 7.2 7.1 7.2 ...
## $ step0.output.pH : num 7.32 7.27 7.25 7.3 7.27 ...
## $ Step0.tmpressure : num 15.3 19.1 16.2 14.7 14.3 ...
## $ Step0.return.flow : num 81.9 81.9 81.9 80.3 81.9 ...
## $ Step0.temperature : num 10 10.12 10.12 10 9.88 ...
## $ Final.Resin.Product : num 56.4 61.7 55 62.7 61 ...str(resin2_train)## 'data.frame': 270 obs. of 35 variables:
## $ TreeSite : chr "Village Harvest" "Village Harvest" "Village Harvest" "City Harvest" ...
## $ Pressure : num 2 0 0 0 1 1 2 3 3 2 ...
## $ Step1.Temp : int 18 18 18 18 18 18 18 18 18 18 ...
## $ Step1.conductivity : num 13.2 13.5 13.3 13.5 13.4 13 13.7 13.4 13.4 13.2 ...
## $ Step1.Resin : num 13159 12483 12785 14100 12563 ...
## $ Step2.pH : num 5.64 5.64 5.65 5.66 5.61 5.64 5.63 5.61 5.67 5.64 ...
## $ Step2.Conductivity : int 49 49 49 49 50 49 49 50 46 49 ...
## $ Step2.Yield : int 8 7 9 9 8 7 7 8 9 8 ...
## $ Step3.Temp : num 19 19.7 19.7 19.7 19.7 ...
## $ Step3.pH : num 5.63 5.65 5.67 5.68 5.66 5.64 5.62 5.7 5.68 5.65 ...
## $ Step3.Conductivity : int 41 40 40 40 40 40 40 40 40 40 ...
## $ Step3.yield : int 17 21 16 14 20 18 17 15 17 17 ...
## $ Step3.compression : num 1.26 1.16 1.18 1.18 1.16 ...
## $ Step4.pH : num 6.43 6.91 7.3 6.7 7.02 6.99 7.49 7.21 7.33 7.55 ...
## $ Step5.yield : num 80 82 82 79 84 81 82 79 87 82 ...
## $ Step5.temp : num 18 19 18 18 20 20 19 19 20 19 ...
## $ Step6.pH.start : num 4.68 4.69 4.7 4.63 4.65 4.7 4.59 4.66 4.64 4.69 ...
## $ Step6.Conductivity.start: num 102 103 105 104 104 ...
## $ Step6.temperature : int 17 18 18 18 18 18 18 17 18 18 ...
## $ Step7.symmetry : num 1.64 1.61 1.41 1.41 1.87 1.65 1.87 1.06 1.8 1.75 ...
## $ Step8.temperature : num 18.5 18.5 18.5 18.5 18.5 ...
## $ step.8.yield : int 90 93 93 93 90 92 91 91 92 92 ...
## $ Step7.UV : num 3.02 2.98 2.99 2.89 3.01 ...
## $ Step8.UV : num 0.701 0.703 0.688 0.713 0.727 ...
## $ Step.1.pH : num 7.2 7.2 7.2 7.1 7.2 ...
## $ step.1.particles : num 2427500 2640000 2770000 3670000 2780000 ...
## $ step.1.quality. : num 94.2 98.2 95.1 94.6 95.1 ...
## $ step0.total.particles : num 868889 840000 950000 1010256 960000 ...
## $ step0.warmup.hrs : num 19.7 16 16.9 16.9 17 ...
## $ step0.input.pH : num 7.2 7.2 7.1 7.1 7.2 ...
## $ step0.output.pH : num 7.34 7.31 7.3 7.3 7.32 ...
## $ Step0.tmpressure : num 13.8 15.6 14.4 17.4 14.8 ...
## $ Step0.return.flow : num 80.4 80.3 81.9 81.9 81.2 ...
## $ Step0.temperature : num 9.56 10.1 10.34 10.12 10.06 ...
## $ Final.Resin.Product : num 59.3 71 61 56.3 67.9 ...resin2_train$TreeSite = as.factor(resin2_train$TreeSite)
resin2_test$TreeSite = as.factor(resin2_test$TreeSite)
##Re-running the decision tree and random forest models with modified data
#revised decision tree model with rpart
rpartmodel2 <- rpart(Final.Resin.Product~., data=resin2_train, method = 'anova')
summary(rpartmodel2)## Call:
## rpart(formula = Final.Resin.Product ~ ., data = resin2_train,
## method = "anova")
## n= 270
##
## CP nsplit rel error xerror xstd
## 1 0.27791485 0 1.0000000 1.0055874 0.10489951
## 2 0.12178424 1 0.7220851 0.7358733 0.08309345
## 3 0.05422432 2 0.6003009 0.6458785 0.06759378
## 4 0.04176818 3 0.5460766 0.6351812 0.06604956
## 5 0.03741543 4 0.5043084 0.6213511 0.06285652
## 6 0.02942100 5 0.4668930 0.6096217 0.06118237
## 7 0.02619517 6 0.4374720 0.5943916 0.05938842
## 8 0.01947945 7 0.4112768 0.6027184 0.06229733
## 9 0.01704187 8 0.3917974 0.6047151 0.06499826
## 10 0.01527325 10 0.3577136 0.6097708 0.06550507
## 11 0.01509449 11 0.3424404 0.6141552 0.06497369
## 12 0.01316619 12 0.3273459 0.6106708 0.06610380
## 13 0.01155538 13 0.3141797 0.6164629 0.06093879
## 14 0.01000000 14 0.3026243 0.6159412 0.06163970
##
## Variable importance
## TreeSite step0.output.pH Step1.Resin
## 17 12 11
## Step0.temperature step.1.quality. Step.1.pH
## 10 9 8
## Step0.return.flow Step0.tmpressure Step5.yield
## 6 4 3
## Step7.symmetry Step3.compression Step6.Conductivity.start
## 3 3 2
## Step2.Conductivity Step4.pH Step1.conductivity
## 1 1 1
## Step3.yield step0.input.pH step0.warmup.hrs
## 1 1 1
## Step3.pH Step2.pH Step8.temperature
## 1 1 1
##
## Node number 1: 270 observations, complexity param=0.2779149
## mean=60.86954, MSE=36.5422
## left son=2 (154 obs) right son=3 (116 obs)
## Primary splits:
## TreeSite splits as LLR, improve=0.2779149, (0 missing)
## Step0.temperature < 10.11453 to the right, improve=0.1623813, (0 missing)
## Step1.Resin < 12480.5 to the left, improve=0.1602819, (0 missing)
## step0.output.pH < 7.289643 to the left, improve=0.1521203, (0 missing)
## Step5.yield < 80.39818 to the left, improve=0.1402625, (0 missing)
## Surrogate splits:
## step0.output.pH < 7.298438 to the left, agree=0.881, adj=0.724, (0 split)
## Step0.temperature < 10.10913 to the right, agree=0.752, adj=0.422, (0 split)
## Step.1.pH < 7.197619 to the left, agree=0.748, adj=0.414, (0 split)
## Step0.return.flow < 81.90516 to the right, agree=0.719, adj=0.345, (0 split)
## step.1.quality. < 93.36066 to the left, agree=0.711, adj=0.328, (0 split)
##
## Node number 2: 154 observations, complexity param=0.1217842
## mean=58.10373, MSE=30.15883
## left son=4 (35 obs) right son=5 (119 obs)
## Primary splits:
## Step1.Resin < 12808.5 to the left, improve=0.2587107, (0 missing)
## step.1.quality. < 87.88226 to the left, improve=0.1397313, (0 missing)
## Step3.compression < 1.188815 to the right, improve=0.1126030, (0 missing)
## Step0.temperature < 10.11453 to the right, improve=0.1038099, (0 missing)
## Step0.tmpressure < 11.30958 to the left, improve=0.1000074, (0 missing)
## Surrogate splits:
## Step3.compression < 1.255 to the right, agree=0.812, adj=0.171, (0 split)
## Step0.tmpressure < 11.22096 to the left, agree=0.812, adj=0.171, (0 split)
## Step3.pH < 5.575 to the left, agree=0.792, adj=0.086, (0 split)
## step.8.yield < 88.5 to the left, agree=0.786, adj=0.057, (0 split)
## Step.1.pH < 7.282807 to the right, agree=0.786, adj=0.057, (0 split)
##
## Node number 3: 116 observations, complexity param=0.05422432
## mean=64.54139, MSE=21.3786
## left son=6 (108 obs) right son=7 (8 obs)
## Primary splits:
## Step7.symmetry < 1.09 to the right, improve=0.2157324, (0 missing)
## Step2.Conductivity < 46.5 to the right, improve=0.2137481, (0 missing)
## Step1.Resin < 12016.95 to the left, improve=0.1925142, (0 missing)
## step0.input.pH < 7.109916 to the left, improve=0.1856576, (0 missing)
## Step0.tmpressure < 15.28579 to the left, improve=0.1340051, (0 missing)
## Surrogate splits:
## Step2.Conductivity < 45.5 to the right, agree=0.957, adj=0.375, (0 split)
##
## Node number 4: 35 observations, complexity param=0.02619517
## mean=52.95318, MSE=28.8939
## left son=8 (13 obs) right son=9 (22 obs)
## Primary splits:
## step.1.quality. < 90.06 to the left, improve=0.2555674, (0 missing)
## Step5.yield < 80.5 to the left, improve=0.2384527, (0 missing)
## Step4.pH < 6.825 to the left, improve=0.2024653, (0 missing)
## Step6.Conductivity.start < 104.0245 to the left, improve=0.1733643, (0 missing)
## Step0.tmpressure < 11.69608 to the left, improve=0.1517861, (0 missing)
## Surrogate splits:
## Step4.pH < 6.825 to the left, agree=0.857, adj=0.615, (0 split)
## Step6.Conductivity.start < 104.0245 to the left, agree=0.829, adj=0.538, (0 split)
## Step3.compression < 1.255 to the right, agree=0.771, adj=0.385, (0 split)
## Step5.yield < 78.5 to the left, agree=0.743, adj=0.308, (0 split)
## step.1.particles < 2965385 to the left, agree=0.743, adj=0.308, (0 split)
##
## Node number 5: 119 observations, complexity param=0.03741543
## mean=59.6186, MSE=20.43363
## left son=10 (111 obs) right son=11 (8 obs)
## Primary splits:
## Step5.yield < 82.5 to the left, improve=0.15181570, (0 missing)
## step.1.quality. < 88.11333 to the left, improve=0.13551610, (0 missing)
## Step3.yield < 15.5 to the left, improve=0.13437600, (0 missing)
## Step8.UV < 0.69485 to the left, improve=0.11975150, (0 missing)
## step0.input.pH < 7.148436 to the left, improve=0.09762502, (0 missing)
## Surrogate splits:
## Step3.yield < 19.5 to the left, agree=0.958, adj=0.375, (0 split)
##
## Node number 6: 108 observations, complexity param=0.04176818
## mean=63.95689, MSE=17.1909
## left son=12 (7 obs) right son=13 (101 obs)
## Primary splits:
## Step1.Resin < 11697.75 to the left, improve=0.22196350, (0 missing)
## step0.input.pH < 7.109916 to the left, improve=0.19551480, (0 missing)
## Step5.yield < 83.5 to the left, improve=0.09623668, (0 missing)
## Step0.tmpressure < 15.28579 to the left, improve=0.08981585, (0 missing)
## step.1.quality. < 87.775 to the left, improve=0.07204267, (0 missing)
##
## Node number 7: 8 observations
## mean=72.43206, MSE=11.03769
##
## Node number 8: 13 observations
## mean=49.41813, MSE=31.68906
##
## Node number 9: 22 observations, complexity param=0.01509449
## mean=55.04207, MSE=15.4944
## left son=18 (7 obs) right son=19 (15 obs)
## Primary splits:
## Step0.tmpressure < 12.13498 to the left, improve=0.4368973, (0 missing)
## Step1.Resin < 12309.65 to the left, improve=0.4230592, (0 missing)
## Step8.UV < 0.70935 to the left, improve=0.3818082, (0 missing)
## Step5.yield < 80.5 to the left, improve=0.3781286, (0 missing)
## step.1.quality. < 93.00179 to the left, improve=0.3153516, (0 missing)
## Surrogate splits:
## Step2.pH < 5.625 to the left, agree=0.864, adj=0.571, (0 split)
## Step3.compression < 1.235 to the right, agree=0.864, adj=0.571, (0 split)
## step0.input.pH < 7.203334 to the right, agree=0.864, adj=0.571, (0 split)
## Step1.conductivity < 12.4 to the left, agree=0.818, adj=0.429, (0 split)
## Step2.Yield < 7.5 to the left, agree=0.818, adj=0.429, (0 split)
##
## Node number 10: 111 observations, complexity param=0.029421
## mean=59.14576, MSE=16.74951
## left son=20 (19 obs) right son=21 (92 obs)
## Primary splits:
## step.1.quality. < 88.045 to the left, improve=0.15613160, (0 missing)
## Step0.temperature < 10.03571 to the right, improve=0.14071740, (0 missing)
## Step3.yield < 15.5 to the left, improve=0.12832540, (0 missing)
## Step8.UV < 0.69485 to the left, improve=0.12466030, (0 missing)
## Step0.tmpressure < 15.9099 to the right, improve=0.08834975, (0 missing)
## Surrogate splits:
## Step8.UV < 0.6879 to the left, agree=0.856, adj=0.158, (0 split)
## step0.warmup.hrs < 18.3 to the right, agree=0.838, adj=0.053, (0 split)
##
## Node number 11: 8 observations
## mean=66.17926, MSE=25.40629
##
## Node number 12: 7 observations
## mean=56.53693, MSE=12.97275
##
## Node number 13: 101 observations, complexity param=0.01947945
## mean=64.47115, MSE=13.40304
## left son=26 (90 obs) right son=27 (11 obs)
## Primary splits:
## Step0.tmpressure < 15.28579 to the left, improve=0.14197460, (0 missing)
## step0.input.pH < 7.108064 to the left, improve=0.11890590, (0 missing)
## Step5.yield < 83.5 to the left, improve=0.09538495, (0 missing)
## Step8.UV < 0.7202 to the left, improve=0.07473181, (0 missing)
## Step0.temperature < 9.9214 to the right, improve=0.07340118, (0 missing)
##
## Node number 18: 7 observations
## mean=51.2334, MSE=4.023933
##
## Node number 19: 15 observations
## mean=56.81945, MSE=10.91874
##
## Node number 20: 19 observations
## mean=55.58729, MSE=17.89809
##
## Node number 21: 92 observations, complexity param=0.01527325
## mean=59.88066, MSE=13.3571
## left son=42 (65 obs) right son=43 (27 obs)
## Primary splits:
## Step0.temperature < 10.03571 to the right, improve=0.12262810, (0 missing)
## Step7.symmetry < 1.565 to the left, improve=0.11603610, (0 missing)
## TreeSite splits as LR-, improve=0.08158899, (0 missing)
## step0.input.pH < 7.117524 to the left, improve=0.07876424, (0 missing)
## Step3.compression < 1.188815 to the right, improve=0.07780008, (0 missing)
## Surrogate splits:
## TreeSite splits as LR-, agree=0.859, adj=0.519, (0 split)
## Step0.return.flow < 82.0625 to the left, agree=0.815, adj=0.370, (0 split)
## step0.warmup.hrs < 16.87402 to the right, agree=0.772, adj=0.222, (0 split)
## step0.output.pH < 7.296482 to the left, agree=0.772, adj=0.222, (0 split)
## Step6.Conductivity.start < 103.487 to the right, agree=0.761, adj=0.185, (0 split)
##
## Node number 26: 90 observations, complexity param=0.01704187
## mean=63.98889, MSE=11.98437
## left son=52 (57 obs) right son=53 (33 obs)
## Primary splits:
## Step0.temperature < 9.9214 to the right, improve=0.15553330, (0 missing)
## step0.input.pH < 7.108064 to the left, improve=0.11215220, (0 missing)
## Step8.UV < 0.7225 to the left, improve=0.10525300, (0 missing)
## Step0.return.flow < 79.38854 to the right, improve=0.07686055, (0 missing)
## step0.warmup.hrs < 17.55584 to the right, improve=0.07659861, (0 missing)
## Surrogate splits:
## Step3.Conductivity < 40.5 to the left, agree=0.756, adj=0.333, (0 split)
## step0.warmup.hrs < 16.55238 to the right, agree=0.722, adj=0.242, (0 split)
## Step3.compression < 1.188815 to the left, agree=0.700, adj=0.182, (0 split)
## Step8.UV < 0.71375 to the left, agree=0.700, adj=0.182, (0 split)
## Step0.return.flow < 82.49074 to the left, agree=0.700, adj=0.182, (0 split)
##
## Node number 27: 11 observations
## mean=68.41692, MSE=7.538337
##
## Node number 42: 65 observations, complexity param=0.01316619
## mean=59.05581, MSE=10.25407
## left son=84 (30 obs) right son=85 (35 obs)
## Primary splits:
## Step6.Conductivity.start < 105.956 to the left, improve=0.1948986, (0 missing)
## Step8.UV < 0.699 to the left, improve=0.1553981, (0 missing)
## Step7.symmetry < 1.565 to the left, improve=0.1352056, (0 missing)
## Step5.temp < 18.5 to the left, improve=0.1326888, (0 missing)
## Step3.compression < 1.165 to the right, improve=0.1251886, (0 missing)
## Surrogate splits:
## Step0.tmpressure < 16.17419 to the right, agree=0.754, adj=0.467, (0 split)
## Step1.Resin < 13856.55 to the right, agree=0.738, adj=0.433, (0 split)
## step0.input.pH < 7.157158 to the left, agree=0.723, adj=0.400, (0 split)
## Step3.Temp < 19.85664 to the left, agree=0.677, adj=0.300, (0 split)
## Step.1.pH < 7.146774 to the right, agree=0.662, adj=0.267, (0 split)
##
## Node number 43: 27 observations, complexity param=0.01155538
## mean=61.86642, MSE=15.24616
## left son=86 (13 obs) right son=87 (14 obs)
## Primary splits:
## Step5.yield < 80.5 to the left, improve=0.2769610, (0 missing)
## Step0.temperature < 9.8375 to the left, improve=0.2264512, (0 missing)
## step0.input.pH < 7.190474 to the left, improve=0.2167303, (0 missing)
## Step0.tmpressure < 13.105 to the left, improve=0.1908182, (0 missing)
## Pressure < 0.5 to the left, improve=0.1458057, (0 missing)
## Surrogate splits:
## Step0.tmpressure < 14.23011 to the left, agree=0.778, adj=0.538, (0 split)
## Step1.conductivity < 13.15 to the left, agree=0.741, adj=0.462, (0 split)
## step.1.quality. < 90.67079 to the left, agree=0.741, adj=0.462, (0 split)
## step0.warmup.hrs < 16.87402 to the left, agree=0.704, adj=0.385, (0 split)
## Step4.pH < 5.565 to the left, agree=0.667, adj=0.308, (0 split)
##
## Node number 52: 57 observations, complexity param=0.01704187
## mean=62.95007, MSE=9.521646
## left son=104 (47 obs) right son=105 (10 obs)
## Primary splits:
## Step0.temperature < 10.58523 to the left, improve=0.3105141, (0 missing)
## Step1.Resin < 14444.7 to the left, improve=0.2048429, (0 missing)
## step0.total.particles < 1003125 to the left, improve=0.1635155, (0 missing)
## step.1.particles < 2953000 to the left, improve=0.1304884, (0 missing)
## step0.output.pH < 7.312019 to the left, improve=0.1152205, (0 missing)
## Surrogate splits:
## Step1.Resin < 14516.1 to the left, agree=0.930, adj=0.6, (0 split)
## Step8.temperature < 18.77311 to the left, agree=0.912, adj=0.5, (0 split)
## step0.total.particles < 1043963 to the left, agree=0.877, adj=0.3, (0 split)
## Step1.conductivity < 12.35 to the right, agree=0.860, adj=0.2, (0 split)
## step.1.quality. < 95.3 to the left, agree=0.842, adj=0.1, (0 split)
##
## Node number 53: 33 observations
## mean=65.78321, MSE=11.15461
##
## Node number 84: 30 observations
## mean=57.52886, MSE=7.226515
##
## Node number 85: 35 observations
## mean=60.36463, MSE=9.137614
##
## Node number 86: 13 observations
## mean=59.73395, MSE=8.770849
##
## Node number 87: 14 observations
## mean=63.84656, MSE=13.11537
##
## Node number 104: 47 observations
## mean=62.15693, MSE=6.5464
##
## Node number 105: 10 observations
## mean=66.67781, MSE=6.652656rpart.plot(rpartmodel2,box.palette="RdBu", shadow.col="gray", nn=TRUE)
rpart_pred = predict(rpartmodel2,resin2_test)
mean((rpart_pred-resin2_test$Final.Resin.Product)^2) #MeanSquareError## [1] 20.49385print(rpart_pred-resin2_test$Final.Resin.Product)## 164 297 87 262 151 52
## 3.33525385 -1.32107143 11.16366250 -0.56186596 -1.29554615 -0.47107143
## 180 292 169 256 159 332
## -1.54731053 -1.25474615 -4.08194333 3.54252857 -6.14246596 -2.10449394
## 163 280 108 37 320 81
## 2.88980606 -6.15963750 -1.13544333 5.14816250 1.23045385 3.68505667
## 136 2 255 340 316 192
## 7.99606250 -0.53477143 3.96693404 -3.23591053 -3.23277143 -7.07897143
## 4 202 225 106 38 116
## -2.60813571 6.96438947 7.60685385 -15.46564333 -0.77794333 1.14723404
## 227 243 264 238 149 178
## -1.32107143 -0.86147143 8.03008947 -0.59158182 1.67945667 -3.27204615
## 43 91 247 188 241 215
## -6.49716596 4.94076429 0.35273404 1.05951818 -2.61206596 4.72111000
## 127 135 242 70 67 261
## -1.01074333 3.27682857 7.28695333 3.19113404 2.78395333 4.80976250
## 20 313 146 252 337 77
## 1.78535333 3.50336250 4.35051818 10.39751818 0.40850606 -3.64654667
## 326 51 23 171 185 103
## 1.02135333 -3.58344615 -6.57124667 0.67583404 1.94466250 0.16335667
## 65 117 195 237 63 10
## -0.08361053 -5.17166923 0.42943404 0.57503404 8.45865333 -1.65056596
## 147 167 274 269
## -0.03854333 -3.35164333 2.79806429 1.84835333#CrossTable(resin2_test$Final.Resin.Product,rpart_pred,prop.chisq = FALSE, prop.c = FALSE, prop.r = FALSE, dnn = #c('ActualResin','PredictedResin'))
#(pdtr2 <- table(rpart_pred,resin2_test$Final.Resin.Product))
#(accuracy <- sum(diag(pdtr2))/sum(pdtr2)*100)
##Re-running random forest model with modified data
library(randomForest)## randomForest 4.7-1## Type rfNews() to see new features/changes/bug fixes.random_resin2 <- randomForest(Final.Resin.Product~.,data=resin2_train,mtry = 3,importance = TRUE, na.action = na.omit)
summary(random_resin2)## Length Class Mode
## call 6 -none- call
## type 1 -none- character
## predicted 270 -none- numeric
## mse 500 -none- numeric
## rsq 500 -none- numeric
## oob.times 270 -none- numeric
## importance 68 -none- numeric
## importanceSD 34 -none- numeric
## localImportance 0 -none- NULL
## proximity 0 -none- NULL
## ntree 1 -none- numeric
## mtry 1 -none- numeric
## forest 11 -none- list
## coefs 0 -none- NULL
## y 270 -none- numeric
## test 0 -none- NULL
## inbag 0 -none- NULL
## terms 3 terms callprint(random_resin2)##
## Call:
## randomForest(formula = Final.Resin.Product ~ ., data = resin2_train, mtry = 3, importance = TRUE, na.action = na.omit)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 3
##
## Mean of squared residuals: 15.95867
## % Var explained: 56.33random_pred2 <- predict(random_resin2,resin2_test)
#(prn2<-table(random_pred2,resin2_test$Final.Resin.Product))
#(randaccuracy2 <-sum(diag(prn2))/sum(prn2)*100)
random_resin2##
## Call:
## randomForest(formula = Final.Resin.Product ~ ., data = resin2_train, mtry = 3, importance = TRUE, na.action = na.omit)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 3
##
## Mean of squared residuals: 15.95867
## % Var explained: 56.33importance(random_resin2)## %IncMSE IncNodePurity
## TreeSite 15.839251 729.04997
## Pressure 2.941823 132.71770
## Step1.Temp 0.000000 0.00000
## Step1.conductivity 8.802062 324.66025
## Step1.Resin 15.848069 659.11561
## Step2.pH 1.805252 204.46077
## Step2.Conductivity 7.727103 327.48764
## Step2.Yield 5.821304 102.53391
## Step3.Temp 4.488733 114.45172
## Step3.pH 3.291567 172.90717
## Step3.Conductivity 2.742881 35.85039
## Step3.yield 8.866270 310.53510
## Step3.compression 9.071007 239.35637
## Step4.pH 6.525684 260.50672
## Step5.yield 12.218167 538.91068
## Step5.temp 2.827892 59.58037
## Step6.pH.start 1.732550 166.30152
## Step6.Conductivity.start 3.076291 237.67058
## Step6.temperature 2.950301 47.95199
## Step7.symmetry 11.855867 403.42573
## Step8.temperature 4.289990 101.74009
## step.8.yield 3.912101 163.60147
## Step7.UV 1.111764 213.33184
## Step8.UV 1.574684 241.29533
## Step.1.pH 9.077051 324.22097
## step.1.particles 4.844773 235.04789
## step.1.quality. 10.197762 481.53981
## step0.total.particles 4.998940 228.43611
## step0.warmup.hrs 4.767805 125.40080
## step0.input.pH 14.116110 438.50199
## step0.output.pH 8.968009 469.60789
## Step0.tmpressure 10.850799 554.11512
## Step0.return.flow 5.527307 172.19363
## Step0.temperature 9.432181 420.23620order(importance(random_resin2))## [1] 3 37 23 24 17 6 11 16 2 19 18 10 22 21 9 29 26 28 33 8 14 7 4 12 31
## [26] 13 25 34 27 32 20 15 30 1 5 45 53 50 55 42 43 63 36 56 51 67 44 40 57 62
## [51] 60 52 47 58 48 46 59 38 41 54 68 64 65 61 49 66 39 35##Mean Square Error of RandomForesModel
mean((random_pred2-resin2_test$Final.Resin.Product)^2) #MeanSquareError## [1] 14.35697#importance(random_resin2)[order(importance(random_resin2)),]##Selected features from Random forest model results: Model3 step.1.particles step0.total.particles Step.1.pH Step8.UV
Step7.symmetry,Step6.Conductivity.start,Step0.tmpressure,Step0.temperature,step0.input.pH,step.1.quality., Step5.yield,step0.output.pH,TreeSite,Final.Resin.Product
Creating new dataset according to features identified from Random Forest
## step0.input.pH Step0.tmpressure Step7.symmetry Step0.temperature
## 14 7.2 13.83333 1.64 9.555556
## 32 7.2 15.58065 1.61 10.096774
## 196 7.1 14.41379 1.41 10.344828
## 83 7.1 17.41250 1.41 10.117944
## 119 7.2 14.78788 1.87 10.060606
## 300 7.1 15.86667 1.65 10.117944
## step0.output.pH step.1.quality. Step5.yield Step1.Resin TreeSite
## 14 7.340000 94.24722 80 13158.6 Village Harvest
## 32 7.309677 98.20000 82 12482.8 Village Harvest
## 196 7.296154 95.10000 82 12785.4 Village Harvest
## 83 7.296875 94.60000 79 14099.5 City Harvest
## 119 7.318182 95.10000 84 12563.1 Village Harvest
## 300 7.248148 88.50000 81 13577.2 City Harvest
## Final.Resin.Product
## 14 59.2894
## 32 70.9784
## 196 61.0340
## 83 56.3344
## 119 67.9357
## 300 59.6207## 'data.frame': 270 obs. of 10 variables:
## $ step0.input.pH : num 7.2 7.2 7.1 7.1 7.2 ...
## $ Step0.tmpressure : num 13.8 15.6 14.4 17.4 14.8 ...
## $ Step7.symmetry : num 1.64 1.61 1.41 1.41 1.87 1.65 1.87 1.06 1.8 1.75 ...
## $ Step0.temperature : num 9.56 10.1 10.34 10.12 10.06 ...
## $ step0.output.pH : num 7.34 7.31 7.3 7.3 7.32 ...
## $ step.1.quality. : num 94.2 98.2 95.1 94.6 95.1 ...
## $ Step5.yield : num 80 82 82 79 84 81 82 79 87 82 ...
## $ Step1.Resin : num 13159 12483 12785 14100 12563 ...
## $ TreeSite : Factor w/ 3 levels "City Harvest",..: 3 3 3 1 3 1 1 1 3 3 ...
## $ Final.Resin.Product: num 59.3 71 61 56.3 67.9 ...## step0.input.pH Step0.tmpressure Step7.symmetry Step0.temperature
## 14 7.2 13.83333 1.64 9.555556
## 32 7.2 15.58065 1.61 10.096774
## 196 7.1 14.41379 1.41 10.344828
## 83 7.1 17.41250 1.41 10.117944
## 119 7.2 14.78788 1.87 10.060606
## 300 7.1 15.86667 1.65 10.117944
## step0.output.pH step.1.quality. Step5.yield Step1.Resin TreeSite
## 14 7.340000 94.24722 80 13158.6 Village Harvest
## 32 7.309677 98.20000 82 12482.8 Village Harvest
## 196 7.296154 95.10000 82 12785.4 Village Harvest
## 83 7.296875 94.60000 79 14099.5 City Harvest
## 119 7.318182 95.10000 84 12563.1 Village Harvest
## 300 7.248148 88.50000 81 13577.2 City Harvest
## Final.Resin.Product
## 14 59.2894
## 32 70.9784
## 196 61.0340
## 83 56.3344
## 119 67.9357
## 300 59.6207## 'data.frame': 270 obs. of 10 variables:
## $ step0.input.pH : num 7.2 7.2 7.1 7.1 7.2 ...
## $ Step0.tmpressure : num 13.8 15.6 14.4 17.4 14.8 ...
## $ Step7.symmetry : num 1.64 1.61 1.41 1.41 1.87 1.65 1.87 1.06 1.8 1.75 ...
## $ Step0.temperature : num 9.56 10.1 10.34 10.12 10.06 ...
## $ step0.output.pH : num 7.34 7.31 7.3 7.3 7.32 ...
## $ step.1.quality. : num 94.2 98.2 95.1 94.6 95.1 ...
## $ Step5.yield : num 80 82 82 79 84 81 82 79 87 82 ...
## $ Step1.Resin : num 13159 12483 12785 14100 12563 ...
## $ TreeSite : Factor w/ 3 levels "City Harvest",..: 3 3 3 1 3 1 1 1 3 3 ...
## $ Final.Resin.Product: num 59.3 71 61 56.3 67.9 ...
## n= 270
##
## node), split, n, deviance, yval
## * denotes terminal node
##
## 1) root 270 9866.39400 60.86954
## 2) TreeSite=City Harvest,Mix Harvest 154 4644.45900 58.10373
## 4) Step1.Resin< 12808.5 35 1011.28600 52.95318
## 8) step.1.quality.< 90.06 13 411.95780 49.41813 *
## 9) step.1.quality.>=90.06 22 340.87670 55.04207
## 18) Step0.tmpressure< 12.13498 7 28.16753 51.23340 *
## 19) Step0.tmpressure>=12.13498 15 163.78110 56.81945 *
## 5) Step1.Resin>=12808.5 119 2431.60100 59.61860
## 10) Step5.yield< 82.5 111 1859.19600 59.14576
## 20) step.1.quality.< 88.045 19 340.06380 55.58729 *
## 21) step.1.quality.>=88.045 92 1228.85300 59.88066
## 42) Step0.temperature>=10.03571 65 666.51480 59.05581 *
## 43) Step0.temperature< 10.03571 27 411.64620 61.86642
## 86) Step5.yield< 80.5 13 114.02100 59.73395 *
## 87) Step5.yield>=80.5 14 183.61520 63.84656 *
## 11) Step5.yield>=82.5 8 203.25030 66.17926 *
## 3) TreeSite=Village Harvest 116 2479.91700 64.54139
## 6) Step7.symmetry>=1.09 108 1856.61700 63.95689
## 12) Step1.Resin< 11697.75 7 90.80923 56.53693 *
## 13) Step1.Resin>=11697.75 101 1353.70700 64.47115
## 26) Step0.tmpressure< 15.28579 90 1078.59300 63.98889
## 52) Step0.temperature>=9.921399 57 542.73380 62.95007
## 104) Step0.temperature< 10.58523 47 307.68080 62.15693 *
## 105) Step0.temperature>=10.58523 10 66.52656 66.67781 *
## 53) Step0.temperature< 9.921399 33 368.10200 65.78321 *
## 27) Step0.tmpressure>=15.28579 11 82.92171 68.41692 *
## 7) Step7.symmetry< 1.09 8 88.30152 72.43206 *## Call:
## rpart(formula = Final.Resin.Product ~ ., data = feature_train)
## n= 270
##
## CP nsplit rel error xerror xstd
## 1 0.27791485 0 1.0000000 1.0058589 0.10462217
## 2 0.12178424 1 0.7220851 0.7385326 0.08304573
## 3 0.05422432 2 0.6003009 0.6487828 0.06727510
## 4 0.04176818 3 0.5460766 0.6395846 0.06685824
## 5 0.03741543 4 0.5043084 0.6107288 0.06292990
## 6 0.02942100 5 0.4668930 0.6200828 0.06255112
## 7 0.02619517 6 0.4374720 0.6069110 0.05978801
## 8 0.01947945 7 0.4112768 0.5936252 0.05863276
## 9 0.01704187 8 0.3917974 0.5948290 0.06215735
## 10 0.01527325 10 0.3577136 0.5824567 0.06135640
## 11 0.01509449 11 0.3424404 0.5719935 0.06101048
## 12 0.01155538 12 0.3273459 0.5627259 0.06389189
## 13 0.01000000 13 0.3157905 0.5690636 0.06361042
##
## Variable importance
## TreeSite step0.output.pH Step1.Resin Step0.temperature
## 22 16 14 13
## step.1.quality. Step5.yield Step0.tmpressure Step7.symmetry
## 12 10 5 5
## step0.input.pH
## 4
##
## Node number 1: 270 observations, complexity param=0.2779149
## mean=60.86954, MSE=36.5422
## left son=2 (154 obs) right son=3 (116 obs)
## Primary splits:
## TreeSite splits as LLR, improve=0.2779149, (0 missing)
## Step0.temperature < 10.11453 to the right, improve=0.1623813, (0 missing)
## Step1.Resin < 12480.5 to the left, improve=0.1602819, (0 missing)
## step0.output.pH < 7.289643 to the left, improve=0.1521203, (0 missing)
## Step5.yield < 80.39818 to the left, improve=0.1402625, (0 missing)
## Surrogate splits:
## step0.output.pH < 7.298438 to the left, agree=0.881, adj=0.724, (0 split)
## Step0.temperature < 10.10913 to the right, agree=0.752, adj=0.422, (0 split)
## step.1.quality. < 93.36066 to the left, agree=0.711, adj=0.328, (0 split)
## Step5.yield < 81.5 to the left, agree=0.681, adj=0.259, (0 split)
## step0.input.pH < 7.19853 to the left, agree=0.641, adj=0.164, (0 split)
##
## Node number 2: 154 observations, complexity param=0.1217842
## mean=58.10373, MSE=30.15883
## left son=4 (35 obs) right son=5 (119 obs)
## Primary splits:
## Step1.Resin < 12808.5 to the left, improve=0.2587107, (0 missing)
## step.1.quality. < 87.88226 to the left, improve=0.1397313, (0 missing)
## Step0.temperature < 10.11453 to the right, improve=0.1038099, (0 missing)
## Step0.tmpressure < 11.30958 to the left, improve=0.1000074, (0 missing)
## TreeSite splits as LR-, improve=0.0963883, (0 missing)
## Surrogate splits:
## Step0.tmpressure < 11.22096 to the left, agree=0.812, adj=0.171, (0 split)
## Step5.yield < 77.5 to the left, agree=0.779, adj=0.029, (0 split)
##
## Node number 3: 116 observations, complexity param=0.05422432
## mean=64.54139, MSE=21.3786
## left son=6 (108 obs) right son=7 (8 obs)
## Primary splits:
## Step7.symmetry < 1.09 to the right, improve=0.21573240, (0 missing)
## Step1.Resin < 12016.95 to the left, improve=0.19251420, (0 missing)
## step0.input.pH < 7.109916 to the left, improve=0.18565760, (0 missing)
## Step0.tmpressure < 15.28579 to the left, improve=0.13400510, (0 missing)
## Step0.temperature < 9.812317 to the right, improve=0.09473891, (0 missing)
##
## Node number 4: 35 observations, complexity param=0.02619517
## mean=52.95318, MSE=28.8939
## left son=8 (13 obs) right son=9 (22 obs)
## Primary splits:
## step.1.quality. < 90.06 to the left, improve=0.25556740, (0 missing)
## Step5.yield < 80.5 to the left, improve=0.23845270, (0 missing)
## Step0.tmpressure < 11.69608 to the left, improve=0.15178610, (0 missing)
## Step7.symmetry < 1.645 to the left, improve=0.09573133, (0 missing)
## Step1.Resin < 12326.45 to the left, improve=0.06146094, (0 missing)
## Surrogate splits:
## Step5.yield < 78.5 to the left, agree=0.743, adj=0.308, (0 split)
## Step0.tmpressure < 11.69608 to the left, agree=0.714, adj=0.231, (0 split)
## Step7.symmetry < 1.8 to the right, agree=0.686, adj=0.154, (0 split)
## Step1.Resin < 11795.55 to the left, agree=0.686, adj=0.154, (0 split)
## step0.output.pH < 7.257216 to the right, agree=0.657, adj=0.077, (0 split)
##
## Node number 5: 119 observations, complexity param=0.03741543
## mean=59.6186, MSE=20.43363
## left son=10 (111 obs) right son=11 (8 obs)
## Primary splits:
## Step5.yield < 82.5 to the left, improve=0.15181570, (0 missing)
## step.1.quality. < 88.11333 to the left, improve=0.13551610, (0 missing)
## step0.input.pH < 7.148436 to the left, improve=0.09762502, (0 missing)
## Step0.temperature < 10.11453 to the right, improve=0.07675249, (0 missing)
## Step7.symmetry < 1.525 to the left, improve=0.07043341, (0 missing)
##
## Node number 6: 108 observations, complexity param=0.04176818
## mean=63.95689, MSE=17.1909
## left son=12 (7 obs) right son=13 (101 obs)
## Primary splits:
## Step1.Resin < 11697.75 to the left, improve=0.22196350, (0 missing)
## step0.input.pH < 7.109916 to the left, improve=0.19551480, (0 missing)
## Step5.yield < 83.5 to the left, improve=0.09623668, (0 missing)
## Step0.tmpressure < 15.28579 to the left, improve=0.08981585, (0 missing)
## step.1.quality. < 87.775 to the left, improve=0.07204267, (0 missing)
##
## Node number 7: 8 observations
## mean=72.43206, MSE=11.03769
##
## Node number 8: 13 observations
## mean=49.41813, MSE=31.68906
##
## Node number 9: 22 observations, complexity param=0.01509449
## mean=55.04207, MSE=15.4944
## left son=18 (7 obs) right son=19 (15 obs)
## Primary splits:
## Step0.tmpressure < 12.13498 to the left, improve=0.4368973, (0 missing)
## Step1.Resin < 12309.65 to the left, improve=0.4230592, (0 missing)
## Step5.yield < 80.5 to the left, improve=0.3781286, (0 missing)
## step.1.quality. < 93.00179 to the left, improve=0.3153516, (0 missing)
## step0.output.pH < 7.254509 to the left, improve=0.1701276, (0 missing)
## Surrogate splits:
## step0.input.pH < 7.203334 to the right, agree=0.864, adj=0.571, (0 split)
## Step7.symmetry < 1.465 to the left, agree=0.727, adj=0.143, (0 split)
##
## Node number 10: 111 observations, complexity param=0.029421
## mean=59.14576, MSE=16.74951
## left son=20 (19 obs) right son=21 (92 obs)
## Primary splits:
## step.1.quality. < 88.045 to the left, improve=0.15613160, (0 missing)
## Step0.temperature < 10.03571 to the right, improve=0.14071740, (0 missing)
## Step0.tmpressure < 15.9099 to the right, improve=0.08834975, (0 missing)
## step0.input.pH < 7.148436 to the left, improve=0.08538924, (0 missing)
## Step7.symmetry < 1.565 to the left, improve=0.08225724, (0 missing)
##
## Node number 11: 8 observations
## mean=66.17926, MSE=25.40629
##
## Node number 12: 7 observations
## mean=56.53693, MSE=12.97275
##
## Node number 13: 101 observations, complexity param=0.01947945
## mean=64.47115, MSE=13.40304
## left son=26 (90 obs) right son=27 (11 obs)
## Primary splits:
## Step0.tmpressure < 15.28579 to the left, improve=0.14197460, (0 missing)
## step0.input.pH < 7.108064 to the left, improve=0.11890590, (0 missing)
## Step5.yield < 83.5 to the left, improve=0.09538495, (0 missing)
## Step0.temperature < 9.9214 to the right, improve=0.07340118, (0 missing)
## step.1.quality. < 87.66184 to the left, improve=0.06799471, (0 missing)
##
## Node number 18: 7 observations
## mean=51.2334, MSE=4.023933
##
## Node number 19: 15 observations
## mean=56.81945, MSE=10.91874
##
## Node number 20: 19 observations
## mean=55.58729, MSE=17.89809
##
## Node number 21: 92 observations, complexity param=0.01527325
## mean=59.88066, MSE=13.3571
## left son=42 (65 obs) right son=43 (27 obs)
## Primary splits:
## Step0.temperature < 10.03571 to the right, improve=0.12262810, (0 missing)
## Step7.symmetry < 1.565 to the left, improve=0.11603610, (0 missing)
## TreeSite splits as LR-, improve=0.08158899, (0 missing)
## step0.input.pH < 7.117524 to the left, improve=0.07876424, (0 missing)
## Step5.yield < 80.89818 to the left, improve=0.05826906, (0 missing)
## Surrogate splits:
## TreeSite splits as LR-, agree=0.859, adj=0.519, (0 split)
## step0.output.pH < 7.296482 to the left, agree=0.772, adj=0.222, (0 split)
## Step1.Resin < 14608.75 to the left, agree=0.750, adj=0.148, (0 split)
## step.1.quality. < 88.38182 to the right, agree=0.717, adj=0.037, (0 split)
##
## Node number 26: 90 observations, complexity param=0.01704187
## mean=63.98889, MSE=11.98437
## left son=52 (57 obs) right son=53 (33 obs)
## Primary splits:
## Step0.temperature < 9.9214 to the right, improve=0.15553330, (0 missing)
## step0.input.pH < 7.108064 to the left, improve=0.11215220, (0 missing)
## Step1.Resin < 14924.45 to the left, improve=0.06918046, (0 missing)
## step.1.quality. < 87.7208 to the left, improve=0.06258108, (0 missing)
## Step7.symmetry < 5.77 to the right, improve=0.05956377, (0 missing)
## Surrogate splits:
## Step0.tmpressure < 13.14263 to the right, agree=0.678, adj=0.121, (0 split)
## step0.output.pH < 7.326111 to the left, agree=0.667, adj=0.091, (0 split)
## Step7.symmetry < 1.16 to the right, agree=0.656, adj=0.061, (0 split)
## Step5.yield < 77.5 to the right, agree=0.656, adj=0.061, (0 split)
## Step1.Resin < 12782.95 to the right, agree=0.656, adj=0.061, (0 split)
##
## Node number 27: 11 observations
## mean=68.41692, MSE=7.538337
##
## Node number 42: 65 observations
## mean=59.05581, MSE=10.25407
##
## Node number 43: 27 observations, complexity param=0.01155538
## mean=61.86642, MSE=15.24616
## left son=86 (13 obs) right son=87 (14 obs)
## Primary splits:
## Step5.yield < 80.5 to the left, improve=0.27696100, (0 missing)
## Step0.temperature < 9.8375 to the left, improve=0.22645120, (0 missing)
## step0.input.pH < 7.190474 to the left, improve=0.21673030, (0 missing)
## Step0.tmpressure < 13.105 to the left, improve=0.19081820, (0 missing)
## Step7.symmetry < 1.565 to the left, improve=0.08982249, (0 missing)
## Surrogate splits:
## Step0.tmpressure < 14.23011 to the left, agree=0.778, adj=0.538, (0 split)
## step.1.quality. < 90.67079 to the left, agree=0.741, adj=0.462, (0 split)
## Step7.symmetry < 1.63 to the left, agree=0.667, adj=0.308, (0 split)
## step0.input.pH < 7.190474 to the left, agree=0.630, adj=0.231, (0 split)
## Step0.temperature < 9.602564 to the left, agree=0.630, adj=0.231, (0 split)
##
## Node number 52: 57 observations, complexity param=0.01704187
## mean=62.95007, MSE=9.521646
## left son=104 (47 obs) right son=105 (10 obs)
## Primary splits:
## Step0.temperature < 10.58523 to the left, improve=0.31051410, (0 missing)
## Step1.Resin < 14444.7 to the left, improve=0.20484290, (0 missing)
## step0.output.pH < 7.312019 to the left, improve=0.11522050, (0 missing)
## step0.input.pH < 7.108064 to the left, improve=0.10291350, (0 missing)
## Step5.yield < 82.5 to the left, improve=0.08845772, (0 missing)
## Surrogate splits:
## Step1.Resin < 14516.1 to the left, agree=0.930, adj=0.6, (0 split)
## step.1.quality. < 95.3 to the left, agree=0.842, adj=0.1, (0 split)
##
## Node number 53: 33 observations
## mean=65.78321, MSE=11.15461
##
## Node number 86: 13 observations
## mean=59.73395, MSE=8.770849
##
## Node number 87: 14 observations
## mean=63.84656, MSE=13.11537
##
## Node number 104: 47 observations
## mean=62.15693, MSE=6.5464
##
## Node number 105: 10 observations
## mean=66.67781, MSE=6.652656
## [1] 20.07133## Length Class Mode
## call 3 -none- call
## type 1 -none- character
## predicted 270 -none- numeric
## mse 500 -none- numeric
## rsq 500 -none- numeric
## oob.times 270 -none- numeric
## importance 9 -none- numeric
## importanceSD 0 -none- NULL
## localImportance 0 -none- NULL
## proximity 0 -none- NULL
## ntree 1 -none- numeric
## mtry 1 -none- numeric
## forest 11 -none- list
## coefs 0 -none- NULL
## y 270 -none- numeric
## test 0 -none- NULL
## inbag 0 -none- NULL
## terms 3 terms call##
## Call:
## randomForest(formula = Final.Resin.Product ~ ., data = feature_train)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 3
##
## Mean of squared residuals: 12.9876
## % Var explained: 64.46## IncNodePurity
## step0.input.pH 685.5047
## Step0.tmpressure 908.2940
## Step7.symmetry 885.3149
## Step0.temperature 769.2187
## step0.output.pH 801.3864
## step.1.quality. 985.5914
## Step5.yield 1064.8589
## Step1.Resin 1651.1822
## TreeSite 1617.2818## [1] 14.08814
Note that the echo = FALSE parameter was added to the
code chunk to prevent printing of the R code that generated the
plot.