Iris Flowers Multi-class Classification

In [21]:

    

# first install packages devtools and pacman manually

#pull in source functions from github
devtools::source_url('https://raw.githubusercontent.com/jsphyg/Machine_Learning_Notebooks/master/myRfunctions.R')
#source("C:\\Work\\myRfunctions.R")
fnRunDate()
fnInstallPackages()


    

    




SHA-1 hash of file is 6de800cc2b55861becf89ae2e1c302676677008e






    





'Project last run on Wed Sep 20 8:22:03 AM 2017'






    





'Package install completed'

In [22]:

    

#**# attach the iris dataset to the environment
data(iris)
# rename the dataset
dataset <- as_tibble(iris)


    

In [23]:

    

glimpse(dataset)


    

    




Observations: 150
Variables: 5
$ Sepal.Length <dbl> 5.1, 4.9, 4.7, 4.6, 5.0, 5.4, 4.6, 5.0, 4.4, 4.9, 5.4,...
$ Sepal.Width  <dbl> 3.5, 3.0, 3.2, 3.1, 3.6, 3.9, 3.4, 3.4, 2.9, 3.1, 3.7,...
$ Petal.Length <dbl> 1.4, 1.4, 1.3, 1.5, 1.4, 1.7, 1.4, 1.5, 1.4, 1.5, 1.5,...
$ Petal.Width  <dbl> 0.2, 0.2, 0.2, 0.2, 0.2, 0.4, 0.3, 0.2, 0.2, 0.1, 0.2,...
$ Species      <fctr> setosa, setosa, setosa, setosa, setosa, setosa, setos...

In [9]:

    

#**# define the target
target <- dplyr::select(dataset, Species)
head(target)


    

    






Species

	
setosa
	
setosa
	
setosa
	
setosa
	
setosa
	
setosa

In [10]:

    

#**# define the predictors
predictors <- dplyr::select(dataset, -dplyr::one_of(c("Species")))
head(predictors)


    

    






Sepal.Length
Sepal.Width
Petal.Length
Petal.Width

	
5.1
3.5
1.4
0.2
	
4.9
3.0
1.4
0.2
	
4.7
3.2
1.3
0.2
	
4.6
3.1
1.5
0.2
	
5.0
3.6
1.4
0.2
	
5.4
3.9
1.7
0.4

In [11]:

    

fnClassDistribution(Class = dataset$Species)


    

    






freq
percentage

	
setosa
50      
33.33333
	
versicolor
50      
33.33333
	
virginica
50      
33.33333

In [12]:

    

psych::describe(dataset, check = T)


    

    






vars
n
mean
sd
median
trimmed
mad
min
max
range
skew
kurtosis
se

	
Sepal.Length
1         
150       
5.843333  
0.8280661 
5.80      
5.808333  
1.03782   
4.3       
7.9       
3.6       
 0.3086407
-0.6058125
0.06761132
	
Sepal.Width
2         
150       
3.057333  
0.4358663 
3.00      
3.043333  
0.44478   
2.0       
4.4       
2.4       
 0.3126147
 0.1387047
0.03558833
	
Petal.Length
3         
150       
3.758000  
1.7652982 
4.35      
3.760000  
1.85325   
1.0       
6.9       
5.9       
-0.2694109
-1.4168574
0.14413600
	
Petal.Width
4         
150       
1.199333  
0.7622377 
1.30      
1.184167  
1.03782   
0.1       
2.5       
2.4       
-0.1009166
-1.3581792
0.06223645
	
Species*
5         
150       
2.000000  
0.8192319 
2.00      
2.000000  
1.48260   
1.0       
3.0       
2.0       
 0.0000000
-1.5199333
0.06689001

In [18]:

    

# the data types
  visdat::vis_dat(dataset)
  # a best guess at the data types looking at the data
  visdat::vis_guess(dataset)
  # what's missing? view NAs
  visdat::vis_miss(dataset, cluster = T, sort_miss = T)


    

In [19]:

    

tabplot::tableplot(
  dataset, sortCol = Species)


    

In [16]:

    

# scatterplot matrix
trellis.par.set(theme = col.whitebg(), warn = FALSE)
caret::featurePlot(x=dataset[, 1:4], y=dataset$Species, plot="ellipse")


    

In [69]:

    

# density plots for each attribute by class value
scales <- list(x=list(relation="free"), y=list(relation="free"))
featurePlot(x=dataset[, 1:4], y=dataset$Species, plot="density", scales=scales)


    

In [70]:

    

# scatterplot matrix
caret::featurePlot(x=dataset[, 1:4], y=dataset$Species, plot="box")


    

In [71]:

    

caret::featurePlot(x=dataset[, 1:4], y=dataset$Species, plot="strip", jitter = TRUE)


    

In [14]:

    

validation_index <- createDataPartition(dataset$Species, p=0.80, list=FALSE)
validation <- dataset[-validation_index,]
dataset <- dataset[validation_index,]


    

In [74]:

    

formula <- Species ~ .


    

In [75]:

    

# Run algorithms using 10-fold cross validation
control <- trainControl(method="cv", number=10)
metric <- "Accuracy"

# a) linear algorithms
# LDA
set.seed(7)
fit.lda <- train(formula, data=dataset, method="lda", metric=metric, trControl=control)
# b) nonlinear algorithms
# CART
set.seed(7)
fit.cart <- train(formula, data=dataset, method="rpart", metric=metric, trControl=control)
# kNN
set.seed(7)
fit.knn <- train(formula, data=dataset, method="knn", metric=metric, trControl=control)
# c) advanced algorithms
# SVM
set.seed(7)
fit.svm <- train(formula, data=dataset, method="svmRadial", metric=metric, trControl=control)
# Random Forest
set.seed(7)
fit.rf <- train(formula, data=dataset, method="rf", metric=metric, trControl=control)
set.seed(13)
fit.xgb <- caret::train(formula, data = dataset, method = "xgbLinear",  metric = metric, trControl = control)


    

    




Loading required package: xgboost
Warning message:
"package 'xgboost' was built under R version 3.3.3"
Attaching package: 'xgboost'

The following object is masked from 'package:dplyr':

    slice

In [76]:

    

# d) compare algorithms
results <- resamples(list(lda=fit.lda, cart=fit.cart, knn=fit.knn, svm=fit.svm, rf=fit.rf, xgb=fit.xgb))
summary(results)


    

    






Call:
summary.resamples(object = results)

Models: lda, cart, knn, svm, rf, xgb 
Number of resamples: 10 

Accuracy 
       Min. 1st Qu. Median   Mean 3rd Qu. Max. NA's
lda  0.9167  0.9375 1.0000 0.9750  1.0000    1    0
cart 0.8333  0.9167 0.9167 0.9250  0.9792    1    0
knn  0.8333  1.0000 1.0000 0.9750  1.0000    1    0
svm  0.8333  0.9167 0.9583 0.9500  1.0000    1    0
rf   0.8333  0.9167 0.9583 0.9417  1.0000    1    0
xgb  0.8333  0.9167 0.9583 0.9500  1.0000    1    0

Kappa 
      Min. 1st Qu. Median   Mean 3rd Qu. Max. NA's
lda  0.875  0.9062 1.0000 0.9625  1.0000    1    0
cart 0.750  0.8750 0.8750 0.8875  0.9688    1    0
knn  0.750  1.0000 1.0000 0.9625  1.0000    1    0
svm  0.750  0.8750 0.9375 0.9250  1.0000    1    0
rf   0.750  0.8750 0.9375 0.9125  1.0000    1    0
xgb  0.750  0.8750 0.9375 0.9250  1.0000    1    0

In [77]:

    

dotplot(results)


    

In [78]:

    

# summarize Best Model
#  Output estimated accuracy of a model.
print(fit.lda)


    

    




Linear Discriminant Analysis 

120 samples
  4 predictor
  3 classes: 'setosa', 'versicolor', 'virginica' 

No pre-processing
Resampling: Cross-Validated (10 fold) 
Summary of sample sizes: 108, 108, 108, 108, 108, 108, ... 
Resampling results:

  Accuracy  Kappa 
  0.975     0.9625

In [82]:

    

set.seed(13)
predictions <- predict(fit.lda, newdata=validation)
confusionMatrix(predictions, validation$Species)


    

    






Confusion Matrix and Statistics

            Reference
Prediction   setosa versicolor virginica
  setosa         10          0         0
  versicolor      0         10         0
  virginica       0          0        10

Overall Statistics
                                     
               Accuracy : 1          
                 95% CI : (0.8843, 1)
    No Information Rate : 0.3333     
    P-Value [Acc > NIR] : 4.857e-15  
                                     
                  Kappa : 1          
 Mcnemar's Test P-Value : NA         

Statistics by Class:

                     Class: setosa Class: versicolor Class: virginica
Sensitivity                 1.0000            1.0000           1.0000
Specificity                 1.0000            1.0000           1.0000
Pos Pred Value              1.0000            1.0000           1.0000
Neg Pred Value              1.0000            1.0000           1.0000
Prevalence                  0.3333            0.3333           0.3333
Detection Rate              0.3333            0.3333           0.3333
Detection Prevalence        0.3333            0.3333           0.3333
Balanced Accuracy           1.0000            1.0000           1.0000

In [ ]: