Chapter 3 R Lab 2 - 31/03/2022

In this lecture we will learn how to implement the logistic regression model, the linear and the quadratic discriminant analysis (LDA and QDA).

The following packages are required: MASS, pROC and tidyverse.

library(MASS) #for LDA and QDA
## 
## Attaching package: 'MASS'
## The following object is masked from 'package:plotly':
## 
##     select
## The following object is masked from 'package:dplyr':
## 
##     select
library(pROC) #for ROC curve
## Type 'citation("pROC")' for a citation.
## 
## Attaching package: 'pROC'
## The following objects are masked from 'package:stats':
## 
##     cov, smooth, var
library(tidyverse)
library(plotly)

3.1 Satisfaction data

The data we use for this lab are from the Kaggle platform link. You can download the dataset from Moodle, or use the following code for recreating it:

# # Data management -> recreate the original dataset from train / test datasets
# satisf_test <- read.csv("~/Downloads/archive/test.csv")
# satisf_test <- satisf_test %>% dplyr::select(satisfaction, Age, Flight.Distance, Departure.Delay.in.Minutes, Arrival.Delay.in.Minutes)
# satisf_test$satisfaction <- ifelse(satisf_test$satisfaction=="satisfied",1,0)
# table(satisf_test$satisfaction) #check!
# 
# satisf_train <- read.csv("~/Downloads/archive/train.csv")
# satisf_train$satisfaction <- ifelse(satisf_train$satisfaction=="satisfied",1,0)
# satisf_train <- satisf_train %>% dplyr::select(satisfaction, Age, Flight.Distance, Departure.Delay.in.Minutes, Arrival.Delay.in.Minutes)
# table(satisf_train$satisfaction) #check!
# satisf <- rbind(satisf_test, satisf_train)
# satisf %>% filter(is.na(satisfaction)==T |
#                     is.na(Age)==T | 
#                     is.na(Flight.Distance)==T |
#                     is.na(Departure.Delay.in.Minutes)==T |
#                     is.na(Arrival.Delay.in.Minutes)==T) #check!
# 
# summary(satisf) #check!
# satisf <- satisf %>% filter(is.na(Arrival.Delay.in.Minutes)==F)
# colnames(satisf) <- c("outcome","age","dist","dep_delay","arr_delay")
# 
# #export
# write_csv(satisf, "~/Downloads/archive/satisf.csv")

The objective of the analysis is to predict whether or not a customer is overall satisfied with the company service, based on other quality measurements.The datasets consists of several predictor variables:

  • outcome: Customer’s overall satisfaction about the airline (0 = Neutral or dissatisfied, 1 = Satisfied)
  • age: Age of passenger (years)
  • dist: Flight distance of the evaluated journey (not indicated :-( )
  • dep_delay: Delay when departure (minutes)
  • arr_delay: Delay delayed at the arrival (minutes)

The response variable is named outcome and is a binary variable: 0 means that the customer was not satisfied about the overall company service, while 1 means he/she was pleased by the experience.

We import the data

satisf <- read.csv("/Users/marco/Google Drive Personale/Formazione/3_Unibg/2021-22/ML_tutor/RLabs/Lab2/archive/satisf.csv")
glimpse(satisf)
## Rows: 129,487
## Columns: 5
## $ outcome   <int> 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 1, 0, 1, …
## $ age       <int> 52, 36, 20, 44, 49, 16, 77, 43, 47, 46, 47, 33, 46, 60, 52, …
## $ dist      <int> 160, 2863, 192, 3377, 1182, 311, 3987, 2556, 556, 1744, 1235…
## $ dep_delay <int> 50, 0, 0, 0, 0, 0, 0, 77, 1, 28, 29, 18, 0, 117, 10, 5, 2, 0…
## $ arr_delay <int> 44, 0, 0, 6, 20, 0, 0, 65, 0, 14, 19, 7, 0, 113, 0, 2, 0, 0,…

We transform this 0/1 integer variable into a factor with categories “Neutral or dissatisfied” and “Satisfied”:

satisf$outcome= factor(satisf$outcome)
levels(satisf$outcome) = c("Neutral-Dissatisfied","Satisfied")
table(satisf$outcome)
## 
## Neutral-Dissatisfied            Satisfied 
##                73225                56262

Some exploratory analysis can be done to study the distribution of the regressors conditionally on the satisfaction condition. We consider here, as an example, Age or Distance:

satisf %>% 
  ggplot() +
  geom_boxplot(aes(age,fill=outcome))

# Seems that the most satisfied customers are also the elder one 

satisf %>% 
  ggplot() +
  geom_density(aes(dist,fill=outcome),alpha=0.3)

# Satisfaction seems to be related also to travel distance. Passenger travelling on longer flights are the one
# that face less dissatisfaction/neutrality.

3.2 Definition of a function for computing performance indexes

For assessing the performance of a classifier we compare predicted categories with observed categories. This can be done by using the confusion matrix which is a 2x2 matrix reporting the joint distribution (with absolute frequencies) of predicted (by row) and observed (by column) categories.

No (obs.) Yes (obs.)
No (pred.) TN FN
Yes (pred.) FP TP
Total N P

We consider in particular the following performance indexes:

  • sensitivity (true positive rate): TP/P
  • specificity (true negative rate): TN/N
  • accuracy (rate of correctly classified observations): (TN+TP)/(N+P)

We define a function named perf_indexes which computes the above defined indexes. The function has only one argument which is a confusion 2x2 matrix named cm. It returns a vector with the 3 indexes.

perf_indexes = function(cm){
  sensitivity = cm[2,2] / (cm[1,2] + cm[2,2])
  specificity = cm[1,1] / (cm[1,1] + cm[2,1])
  accuracy = sum(diag(cm)) / sum(cm)
  return(c(sens=sensitivity,spec=specificity,acc=accuracy))
}

3.3 A new method for creating the training and testing set

To create the training (80%) and test (20%) dataset we use a new approach different from the one introduced in Section 2.1.1 and Section 2.2.

We first create a vector with the indexes we will use for the training dataset. In this case we must set replace to false, since each index will indicate an observation belonging to just train or test. Recall also to set a seed, in order to be able to replicate your results:

set.seed(11, sample.kind="Rejection")
training_index <- sample(1:nrow(satisf),0.8*nrow(satisf), replace=F) # notice that here resample is not allowed!
head(training_index)
## [1]  65570  66457 128113  19004  73612  28886

We are now ready to sample from the satisf dataset 80% of the observations (rows) by using the training_index vector, and leave the non-sampled indexes for test. Just use a minus (-) before the row selection to exclude them:

# recall: dataset [ row , columns ]
s_train <- satisf[training_index,]
s_test<- satisf[-training_index,]

# check datasets
head(s_train)
##                     outcome age dist dep_delay arr_delay
## 65570  Neutral-Dissatisfied  50  541        35        39
## 66457  Neutral-Dissatisfied  39 1013        17        57
## 128113            Satisfied  42 3330         9         2
## 19004  Neutral-Dissatisfied  41  341         0         0
## 73612  Neutral-Dissatisfied  24 1084        45        42
## 28886  Neutral-Dissatisfied  22  258        18        19
head(s_test)
##                 outcome age dist dep_delay arr_delay
## 5             Satisfied  49 1182         0        20
## 12 Neutral-Dissatisfied  33  325        18         7
## 19 Neutral-Dissatisfied  43 1927         0         0
## 29            Satisfied  44 1543         0         0
## 30 Neutral-Dissatisfied  51  235         0         0
## 34 Neutral-Dissatisfied  22 1846        40        68

Note that this procedure could be the easiest choice when we don’t have a third dataset (validation).

We are now ready to train our models.

3.4 Logistic regression

For implementing the logistic regression the function glm (generalized linear model) is used. It requires the specification of the formula, of the dataframe containing the data and of the distribution (family) of the response variable (in this case binomial distribution given that we are working with a binary response):

logreg = glm(outcome ~ ., 
             data=s_train,
             family="binomial") # binomial use as default "logit" link function
summary(logreg)
## 
## Call:
## glm(formula = outcome ~ ., family = "binomial", data = s_train)
## 
## Deviance Residuals: 
##     Min       1Q   Median       3Q      Max  
## -2.2647  -0.9716  -0.7758   1.1563   2.5771  
## 
## Coefficients:
##               Estimate Std. Error z value Pr(>|z|)    
## (Intercept) -1.556e+00  2.037e-02 -76.361  < 2e-16 ***
## age          1.522e-02  4.403e-04  34.572  < 2e-16 ***
## dist         6.214e-04  7.038e-06  88.294  < 2e-16 ***
## dep_delay    3.118e-03  6.908e-04   4.514 6.37e-06 ***
## arr_delay   -6.788e-03  6.842e-04  -9.921  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## (Dispersion parameter for binomial family taken to be 1)
## 
##     Null deviance: 141820  on 103588  degrees of freedom
## Residual deviance: 130815  on 103584  degrees of freedom
## AIC: 130825
## 
## Number of Fisher Scoring iterations: 4

Note that the notation outcome ~. means that all the covariates but outcome are included as regressors (this avoids to write the formula in the standard way: outcome ~ age+dist+...).

The summary contains the parameter estimates and the corresponding p-values of the test checking \(H_0:\beta =0\) vs \(H_1: \beta\neq 0\). Consider for example the parameter of the Age variable which is positive (i.e. higher possibility of being satisfied) and equal to 0.0152229: this means that for a one-unit increase in age, the log-odds increases by 0.0152229. For a simpler interpretation, in the odds scale, we can take the exponential transformation of the parameter:

# odds ratio. Not of our interest
exp(logreg$coefficients[2])
##      age 
## 1.015339
# odds interpretation
(exp(coef(logreg)[2]) -1)*100
##      age 
## 1.533937

This means that this means that for a one-unit increase in the customer’s age, we expect a 1.53 increase in the satisfaction odds. Remember that the odds is strictly connected to the satisfaction probability as it is defined as \(p/(1-p)\)).

We are now interested in computing predictions for the test observations given the estimated logistic model. They can be obtained by using the predict.glm function. It is important to specify that we are interested in the prediction on the outcome scale (otherwise we will get the predictions on the logit scale);

logreg_pred = predict.glm(logreg, #also predict is fine
                          newdata = s_test,
                          type="response")
head(logreg_pred)
##         5        12        19        29        30        34 
## 0.4474067 0.3009241 0.5735149 0.5181946 0.3467344 0.3987565

The object logreg_pred contains the conditional probability of being satisfied given the covariate vector. For example for the first test observation the satisfaction probability is equal to logreg_pred[1]. In order to have a categorical prediction it is necessary to fix a threshold for the probability (default is 0.5) and transform correspondingly the probabilities:

logreg_pred_class = ifelse(logreg_pred > 0.5, "Satisfied", "Neutral-Dissatisfied")
head(logreg_pred_class)
##                      5                     12                     19 
## "Neutral-Dissatisfied" "Neutral-Dissatisfied"            "Satisfied" 
##                     29                     30                     34 
##            "Satisfied" "Neutral-Dissatisfied" "Neutral-Dissatisfied"

As expected, the predicted category for the first observation is “Neutral - Dissatisfied”.

We are now ready to compute the confusion matrix

table(logreg_pred_class, s_test$outcome)
##                       
## logreg_pred_class      Neutral-Dissatisfied Satisfied
##   Neutral-Dissatisfied                12117      6259
##   Satisfied                            2524      4998

and the performance index by using the function we defined in Section 3.2

perf_indexes(table(logreg_pred_class, s_test$outcome))
##      sens      spec       acc 
## 0.4439904 0.8276074 0.6608618

What happen to our performance indexes if we change prediction threshold?

rm(logreg_pred_class) # to erase a specific object
logreg_pred_class = ifelse(logreg_pred > 0.8, "Yes", "No")
perf_indexes(table(logreg_pred_class, s_test$outcome))
##       sens       spec        acc 
## 0.04557164 0.99112083 0.58012202

As you can see, increasing the satisfaction threshold can improve a lot the algorithm specificity (= our algorithm is more able to detect an unsatisfied customer). At the same time, the decreased in the sensitivity and overall accuracy should be considered. In general, the decision of lowering / increasing the decision boundary depends on the context we are applying our knowledge. Think about COVID. Should be “better” to have a diagnostic test with more false positive (= lower sensitivity) or false negative (= lower specificity)?

3.5 Linear discriminant analysis

To run LDA we use the lda function of the MASS package. Similarly, to logistic regression we have to specify the formula and the data:

ldamod = lda(outcome ~ ., data=s_train)
ldamod
## Call:
## lda(outcome ~ ., data = s_train)
## 
## Prior probabilities of groups:
## Neutral-Dissatisfied            Satisfied 
##            0.5655427            0.4344573 
## 
## Group means:
##                           age      dist dep_delay arr_delay
## Neutral-Dissatisfied 37.66313  928.5023  16.29445  17.02332
## Satisfied            41.73423 1526.8704  12.22791  12.36638
## 
## Coefficients of linear discriminants:
##                     LD1
## age        0.0226054119
## dist       0.0009458021
## dep_delay  0.0045190445
## arr_delay -0.0093103673

The output reports the prior probabilities for the two categories and the conditional means of all the covariates. Recall that the prior probability is the probability that a randomly chosen observation comes from the \(k^{th}\) class.

We go straight to the calculation of the predictions:

lda_pred = predict(ldamod,newdata = s_test)
names(lda_pred)
## [1] "class"     "posterior" "x"

Note that the object lda_pred is a list containing 3 objects including the vector of posterior probabilities (posterior) and the vector of predicted categories (`class):

head(lda_pred$posterior)
##    Neutral-Dissatisfied Satisfied
## 5             0.5565935 0.4434065
## 12            0.7088309 0.2911691
## 19            0.4283927 0.5716073
## 29            0.4860507 0.5139493
## 30            0.6639620 0.3360380
## 34            0.5979763 0.4020237
head(lda_pred$class)
## [1] Neutral-Dissatisfied Neutral-Dissatisfied Satisfied           
## [4] Satisfied            Neutral-Dissatisfied Neutral-Dissatisfied
## Levels: Neutral-Dissatisfied Satisfied

To compute the confusion matrix we will use the vector lda_pred$class:

table(lda_pred$class, s_test$outcome)
##                       
##                        Neutral-Dissatisfied Satisfied
##   Neutral-Dissatisfied                12179      6306
##   Satisfied                            2462      4951
perf_indexes(table(lda_pred$class, s_test$outcome))
##      sens      spec       acc 
## 0.4398152 0.8318421 0.6614410

As before, we can choose a different threshold for our decision. In this case, the used functions do not allow to change directly the threshold. So, we need to use the same methodology applied to logistic regression classification method to generate new classes Recall that LDA assign the the observation to the majority class (= class with highest posterior probability). So, there is an implicit probability threshold with two classes (\(0.5\)). Check if by yourself here above:

lda_pred_class = ifelse(lda_pred$posterior[,"Satisfied"] > 0.5, "Satisfied", "Neutral-Dissatisfied")
# same results than previous LDA
table(lda_pred_class, s_test$outcome) 
##                       
## lda_pred_class         Neutral-Dissatisfied Satisfied
##   Neutral-Dissatisfied                12179      6306
##   Satisfied                            2462      4951
# same results than previous LDA
perf_indexes(table(lda_pred_class, s_test$outcome))
##      sens      spec       acc 
## 0.4398152 0.8318421 0.6614410

We can now set an higher probability threshold, with the aim of being able to better select the “real” satisfied customers. The probability threshold will be set on the “Satisfied” class at \(0.8\) :

lda_pred_class = ifelse(lda_pred$posterior[,"Satisfied"] > 0.8, "Satisfied", "Neutral-Dissatisfied") 

table(lda_pred_class, s_test$outcome) #objects are changed w.r.t. before
##                       
## lda_pred_class         Neutral-Dissatisfied Satisfied
##   Neutral-Dissatisfied                14499     10673
##   Satisfied                             142       584
perf_indexes(table(lda_pred_class, s_test$outcome)) #objects are changed w.r.t. before
##       sens       spec        acc 
## 0.05187883 0.99030121 0.58240019

We reduced a lot the false positive observation (from \(2462\) to \(142\)). We were able to correctly detect more real unsatisfied customers (from \(12179\) to \(14499\)), increasing the specificity of our algorithm. However, we dramatically predict with a wrong class lots of truly satisfied customers (\(6306\) to \(10673\)), being able to predict very few true “positive” (Satisfied) travellers (\(584\)).

3.6 Classifiers comparison in terms of performance indexes

By comparing the 2 classifiers with respect to accuracy, sensitivity and specificity

perf_indexes(table(logreg_pred_class, s_test$outcome))
##       sens       spec        acc 
## 0.04557164 0.99112083 0.58012202
perf_indexes(table(lda_pred$class, s_test$outcome))
##      sens      spec       acc 
## 0.4398152 0.8318421 0.6614410

We see that the logistic regression and LDA perform very similarly. In particular, LDA has the highest specificity but the lowest sensitivity. In terms of accuracy it has the best performance.

3.7 Exercises Lab 2

3.7.1 Exercise 1

The data contained in the files titanic_tr.csv (for training) and titanic_te.csv (and testing) are about the Titanic disaster (the files are available in the e-learning). In particular the following variables are available:

  • pclass: ticket class (1 = 1st, 2 = 2nd, 3 = 3rd)
  • survived: survival (0 = No, 1 = Yes)
  • name of the passenger
  • sex
  • age in years
  • sibsp: number of siblings / spouses aboard the Titanic
  • parch number of parents / children aboard the Titanic
  • ticket: ticket number
  • fare: passenger fare
  • cabin: cabin number
  • embarked: port of Embarkation (C = Cherbourg, Q = Queenstown, S = Southampton)

Import the data in R and explore them. Note that the response variable is survived which is classified as a int 0/1 variable in the dataframe. Transform it in a factor type object using the factor function with categories “No” and “Yes”.

  1. Convert pclass (ticket class) to factor (1 = “1st”, 2 = “2nd”, 3 = “3rd”) for both the training and test data set. Plot survived as a function of pclass. In which class do you observe the highest proportion of survivors?

  2. Check how many missing values we have in the variables age and fare separately for training and test data.

  1. Separately for the training and test data, substitute the missing values with the average of age and fare. Note that for computing the mean when you have missing values you have to run mean(...,na.rm=T). After this, check that you don’t have any other missing values.

  1. Consider the training data and survived as response variable. Compute the percentage of survived people. Moreover, represent graphically the relationship between age and survived. Comment the plot.

  2. Consider the training data set, estimate a logistic model for survived considering age as the only predictor (mod1). Provide the model summary.

  1. Comment the age coefficient.

  2. Compute manually (without using predict) by means of the formula of the logistic function the probability of surviving for a person of 50 years old. Remember that estimated coefficients can be retrieved by mod1$coefficients. Repeat the same for a person of 20 years old. Finally compute the odds and the odds-ratio.

  3. Compute the probability of surviving for all the people in the test data set (use predict).

  4. Plot the estimated probabilities as a function of age (impose that the probability axis is between 0 and 1). Use for each point a color corresponding to the survived outcome. Moreover, include in the plot an horizontal line corresponding to the probability threshold (0.5) and the corresponding age threshold (age boundary between categories). Comment the plot.

  5. Compute the confusion matrix for mod1. Moreover, compute the test classification error rate.

  1. Consider the training data set, estimate a logistic model for survived considering pclass as predictor (mod2). Note that pclass is a categorical variable with 3 categories and will be included in the model as a dummy variable with 3-1 categories (one category is the baseline). Provide the model summary and comment the coefficients. The following function contrasts helps you in understanding which is the baseline category:
contrasts(tr$pclass) 
# on the rows you have the categories
# on the column the 2 dummy variables that will be included in the model (for the 2nd and 3rd category)
#the baseline is the category 1st
#in the model 2 coefficients will be estimated for category 2nd and 3rd

For more information about the management of dummy variables in R please read this short note available here. It refers to a linear regression model but it generalizes to any model.

  1. Compute (manually or by using predict) the probability of surviving for a person with a 1st class ticket. Repeat also for the other 2 classes. Compare the three predicted probabilities with the corresponding surviving proportion computed for each class.

  2. Estimate the probability of surviving (and the corresponding categories) for all the observations in the test data set (using predict). Then compute the test overall classification error rate and compare it with the one obtained for mod1.

  1. For the training observations represent graphically fare (amount paid for the ticket) as a function of pclass. Moreover, compute the average and median fare amount separately for each class.

  2. Consider the training data set, estimate a logistic model for survived considering age and fare as predictors (mod3). Provide the model summary.

  1. Comment the coefficient of fare.

  2. Using mod3 compute the predicted probability of surviving for all the test observations and the corresponding categories.

  3. Plot fare (y-axis) as a function of age using the predicted categories to color the points. Moreover, include in the plot the linear boundary (it’s the solution of \(\beta_0+\beta_1 X_1+\beta_2X_2=0\)) which separates the two groups. HINT: derive it considering \(X_2 = Y\) and \(\beta_1=X\).

  4. Given the previous plot, what’s the prediction for a 40 years old individual who paid 200 dollars ?

  5. Compute the test error rate for mod3 and compare it with the other models.

  1. Consider the training data set, estimate a logistic model for survived considering age, fare and pclass as predictors (mod4). Provide the model summary. Compare mod4 with the others by using the test classification error. Which is the best model?

3.7.2 Exercise 2

We will use some simulated data available from the mlbench library (don’t forget to install it) with \(p=2\) regressors and a binary response variable. Use the following code to generate the data and create the data frame.

library(mlbench)

?mlbench.circle
set.seed(3333,sample.kind="Rejection")
data = mlbench.circle(1000) #simulate 1000 data
datadf = data.frame(data$x,
                    Y=factor(data$classes,labels=c("No","Yes")))
head(datadf)
##           X1          X2   Y
## 1  0.4705415 -0.52006500  No
## 2 -0.1253492  0.20921556  No
## 3 -0.4381452  0.52445868  No
## 4  0.5877131 -0.86523067 Yes
## 5  0.1631693 -0.01084204  No
## 6 -0.6190528 -0.98775598 Yes

  1. Plot the data (use the response variable to color the points).

  2. Split the dataset in two parts: 80% of the observations are used for training and 20% for testing. The split is random (use as seed for random number generation the number 456).

  3. Estimate a logistic regression model using X1 and X2 as regressors. Provide the model summary.

    1. Comment the model coefficients.

    2. Plot the training data and add to the plot the linear decision boundary (see the slides for the formula) defined by the logistic model. Do you think that the model is working well?

    3. Compute the predictions using the test observations. Provide the confusion matrix (use 0.5 as probability threshold) and the overall classification error rate.

  4. Define a function named classification_perf which computes the overall classification error rate, the sensitivity and the specificity for a 2x2 confusion matrix which has in the rows the predicted categories and in the columns the observed categories.

    1. Use the classification_perf function for the logistic regression model output. Comment about the performance of the logistic regression model.

    2. Consider now a very simple classifier (null classifier) which uses as prediction for all the test observations the majority class observed in the training dataset (regardless of the values of the predictors). Compute the overall error rate for the null classifier. Does it perform better than the logistic regression model?

  5. Use linear discriminant analysis to classify your data. Compute the predictions using the test observations and provide the confusion matrix together with performance indexes. Comment the results.

    1. [Advanced and optional] Using a grid of values for X1 and X2 together with the outer function, plot the decision boundary together with the training data. Add to the plot also the boundary estimated with the logistic model.

  6. Compare the three previous classifier (LR, null classifier, LDA) using the confusion matrix (with the classification_perf formula) and the error rate. Are you able to compute the performance indexes with the null classifier? Which of the proposed model seems to perform better?

3.7.3 Exercise 3

We will use some simulated data available from the mlbench library with \(p=2\) regressors and a response variable with 4 categories. Use the following code to generate the data.

library(mlbench)

?mlbench.smiley
set.seed(3333,sample.kind="Rejection")
data = mlbench.smiley(1000) #simulate 1000 data
datadf = data.frame(data$x, factor(data$classes))
head(datadf)
##           x4        V2 factor.data.classes.
## 1 -0.7371167 0.9282289                    1
## 2 -0.8580089 1.0668791                    1
## 3 -0.7794051 1.0191270                    1
## 4 -0.9232296 0.9328375                    1
## 5 -0.7723922 0.9318714                    1
## 6 -0.9142181 0.8255049                    1
colnames(datadf) = c("x1", "x2", "y")
head(datadf)
##           x1        x2 y
## 1 -0.7371167 0.9282289 1
## 2 -0.8580089 1.0668791 1
## 3 -0.7794051 1.0191270 1
## 4 -0.9232296 0.9328375 1
## 5 -0.7723922 0.9318714 1
## 6 -0.9142181 0.8255049 1
table(datadf$y)
## 
##   1   2   3   4 
## 167 167 250 416

  1. Plot the data (use the response variable to color the points).

  2. Split the dataset in two parts: 70% of the observations are used for training and 20% for testing. The split is random (use as seed for random number generation the number 456).

  3. Consider linear discriminant analysis for your data. Analyse your results computing the test error rate.

  4. Use the KNN method to classify your data. Choose the best value of \(k\) among a sequence of values between 1 and 100 (with step equal to 5). To tune the hyperparameter \(k\) split the training dataset into two sets: one for training (70%) and one for validation (30%). Use a seed equal to 4.

  5. Which model (between lda and knn) do you prefer? Explain why.