---

title: “Assignment 4 - LDA” output: html_notebook — library(‘dplyr’) library(‘stargazer’) library(‘caret’) library(‘MASS’) library ( “ISLR” ) library(“psych”) library(“MASS”) library(“ggord”) library(“devtools”) library(“ggplot2”) library(“tidyverse”) library(“survival”)

install.packages(“epitools”) library(epitools)

hotel_data <- read.csv("C:/Users/Issac/OneDrive/Desktop/Meharry/MSDS 565 Predictive Modeling and Analytics/Predicitive Modeling/Hotel.csv")

### Visualize the data
# Install the psych package if you haven't already
install.packages("psych")

WARNING: Rtools is required to build R packages but is not currently installed. Please download and install the appropriate version of Rtools before proceeding:

https://cran.rstudio.com/bin/windows/Rtools/
Installing package into ‘C:/Users/Issac/AppData/Local/R/win-library/4.3’
(as ‘lib’ is unspecified)
trying URL 'https://cran.rstudio.com/bin/windows/contrib/4.3/psych_2.4.6.26.zip'
Content type 'application/zip' length 3720730 bytes (3.5 MB)
downloaded 3.5 MB

package ‘psych’ successfully unpacked and MD5 sums checked

The downloaded binary packages are in
    C:\Users\Issac\AppData\Local\Temp\RtmpsxsHlm\downloaded_packages

# Load the psych package
library(psych)

Warning: package ‘psych’ was built under R version 4.3.3
Attaching package: ‘psych’

The following objects are masked from ‘package:ggplot2’:

    %+%, alpha

# Now you can use the pairs.panels function
pairs.panels(hotel_data[15:16], gap = 0, bg = c("red", "green"))

head(hotel_data)

nrow(hotel$status)

Error: object 'hotel' not found

### Odds Ratio_

### Example Approach
# Sample data a <- 50 # previous cancellations with current cancellations b <- 30 # previous cancellations without current cancellations c <- 20 # no previous cancellations with current cancellations d <- 100 # no previous cancellations without current cancellations


## Create a matrix table <- matrix(c(npc, pc, cc, ncc), ncol=2)##


### People who do not have previous cancellations

npc = length(which(hotel_data$previous_cancellations!=0))
npc

[1] 338

### People who do have previous cancellations

pc = length(which(hotel_data$previous_cancellations==0))

### People who do have current cancellations

cc = length(which(hotel_data$status=="Canceled"))

### People who do not have current cancellations

ncc = length(which(hotel_data$status == "Not_Canceled"))

df1 <- cbind(npc,pc,cc,ncc)
df1

     npc    pc    cc   ncc
[1,] 338 35937 11885 24390

table <- matrix(c(npc, pc, cc, ncc), ncol=2)
table

      [,1]  [,2]
[1,]   338 11885
[2,] 35937 24390

## 1. Among previous cancellations, what is the odds that have current cancellation

Odds_Ratio <-(npc*ncc)/(cc*pc)

Odds_Ratio

[1] 0.01930134

### Which type of room is more likely to have current booking cancelled?

 table(hotel_data$room_type,hotel_data$status)

             
              Canceled Not_Canceled
  Room_Type 1     9072        19058
  Room_Type 2      228          464
  Room_Type 3        2            5
  Room_Type 4     2069         3988
  Room_Type 5       72          193
  Room_Type 6      406          560
  Room_Type 7       36          122

 newc = which(hotel_data$status == "Canceled")/which(hotel_data$status == "Not_Canceled")

Warning: longer object length is not a multiple of shorter object length

## Room type  6 is more likely to have current booking cancelled as it has a 42% cancellation rate. 

##Preprossesing
# Remove rows with missing values
hotel_data <- na.omit(hotel_data)

# Replace infinite values with NA
hotel_data[sapply(hotel_data, is.infinite)] <- NA

# Optionally, you can impute missing values
# For example, using the mean for numeric columns
hotel_data[is.na(hotel_data)] <- lapply(hotel_data, function(x) if(is.numeric(x)) mean(x, na.rm = TRUE) else x)

#Pick 5 to 8 predictors, split your data, 70% training and 30% testing, outcome variable is status

Reduced_set <- hotel_data[,c("previous_bookings_not_canceled","status","previous_cancellations","avg_room_price","room_type","lead_time")]

print(Reduced_set)

### Setting Training and Testing Data

install.packages("caret")

Error in install.packages : Updating loaded packages

library(caret)

Warning: package ‘caret’ was built under R version 4.3.3Loading required package: ggplot2
Warning: package ‘ggplot2’ was built under R version 4.3.3Loading required package: lattice

# Determine the number of rows in the dataset
num_rows <- nrow(Reduced_set)

# Calculate the number of rows for the training set (70% of the data)
train_size <- round(0.7 * num_rows)

# Create a random sample of row indices for the training set
train_indices <- sample(1:num_rows, size = train_size)

# Split the data into training and testing sets
training_data <- Reduced_set[train_indices, ]
testing_data <- Reduced_set[-train_indices, ]

# Print a summary of the training and testing data
summary(training_data)

 previous_bookings_not_canceled     status  previous_cancellations
 Min.   : 0.0000                Min.   :0   Min.   : 0.0000       
 1st Qu.: 0.0000                1st Qu.:0   1st Qu.: 0.0000       
 Median : 0.0000                Median :0   Median : 0.0000       
 Mean   : 0.1563                Mean   :0   Mean   : 0.0239       
 3rd Qu.: 0.0000                3rd Qu.:0   3rd Qu.: 0.0000       
 Max.   :58.0000                Max.   :0   Max.   :13.0000       
 avg_room_price    room_type           lead_time     
 Min.   :  0.00   Length:25392       Min.   :  0.00  
 1st Qu.: 80.75   Class :character   1st Qu.: 17.00  
 Median : 99.95   Mode  :character   Median : 57.00  
 Mean   :103.51                      Mean   : 85.59  
 3rd Qu.:120.60                      3rd Qu.:127.00  
 Max.   :540.00                      Max.   :443.00  

summary(testing_data)

 previous_bookings_not_canceled     status  previous_cancellations
 Min.   : 0.0000                Min.   :0   Min.   : 0.00000      
 1st Qu.: 0.0000                1st Qu.:0   1st Qu.: 0.00000      
 Median : 0.0000                Median :0   Median : 0.00000      
 Mean   : 0.1466                Mean   :0   Mean   : 0.02205      
 3rd Qu.: 0.0000                3rd Qu.:0   3rd Qu.: 0.00000      
 Max.   :53.0000                Max.   :0   Max.   :13.00000      
 avg_room_price    room_type           lead_time    
 Min.   :  0.00   Length:10883       Min.   :  0.0  
 1st Qu.: 80.04   Class :character   1st Qu.: 17.0  
 Median : 99.00   Mode  :character   Median : 57.0  
 Mean   :103.22                      Mean   : 84.4  
 3rd Qu.:120.00                      3rd Qu.:124.0  
 Max.   :349.63                      Max.   :443.0  

# Check the dimensions of the split data
dim(training_data)

[1] 25392     6

dim(testing_data)

[1] 10883     6

## Build a linear discriminant analysis model, report confusion matrix and accuracy
library('MASS')
# Fit the LDA model
lda_model <- lda(status ~., data = Reduced_set)

summary(lda_model)

## Visualizing the lda model
ggplot2::aes(hotel_data$status,hotel_data$previous_bookings_not_canceled)
lda2 <- predict(lda1,type="response")

training_data

### Build a logistic regression model, accuracy



unique(Reduced_set$status)

[1] 0

Reduced_set$status <- ifelse(Reduced_set$status == "yes", 1, 0)
Reduced_set$status <- as.numeric(Reduced_set$status)
logit1 <- glm(status ~ ., data = Reduced_set, family = binomial)

Warning: glm.fit: algorithm did not converge

exp(coef(logit1))

                   (Intercept) previous_bookings_not_canceled 
                  2.900701e-12                   1.000000e+00 
        previous_cancellations                 avg_room_price 
                  1.000000e+00                   1.000000e+00 
          room_typeRoom_Type 2           room_typeRoom_Type 3 
                  1.000000e+00                   1.000000e+00 
          room_typeRoom_Type 4           room_typeRoom_Type 5 
                  1.000000e+00                   1.000000e+00 
          room_typeRoom_Type 6           room_typeRoom_Type 7 
                  1.000000e+00                   1.000000e+00 
                     lead_time 
                  1.000000e+00