Scribd
Upload
48 views2 pages

Assignment 1

The document reads in airline data, imputes missing values with medians, identifies outliers which are dropped, examines correlations between variables, and fits linear and logistic regression models to predict recommendations. Key findings are that value_money_rating and cabin_staff_rating are most important to customers, and country-specific models identify variables for airlines like Lufthansa to focus on for different countries.

Uploaded by

Harshit Patel
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
48 views2 pages

Assignment 1

The document reads in airline data, imputes missing values with medians, identifies outliers which are dropped, examines correlations between variables, and fits linear and logistic regression models to predict recommendations. Key findings are that value_money_rating and cabin_staff_rating are most important to customers, and country-specific models identify variables for airlines like Lufthansa to focus on for different countries.

Uploaded by

Harshit Patel
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd

Data <- [Link]("C:/Users/admin/Downloads/airline.

csv")
names(Data)
str(Data)

#---------------------------Imputing missing values with median of the


var-----------------------------

colSums([Link](Data))
#Data$author_country[[Link](Data$author_country)] <- median(Data$author_country,
[Link] = TRUE)
Data$overall_rating[[Link](Data$overall_rating)] <- median(Data$overall_rating,
[Link] = TRUE)
Data$seat_comfort_rating[[Link](Data$seat_comfort_rating)] <-
median(Data$seat_comfort_rating, [Link] = TRUE)
Data$cabin_staff_rating[[Link](Data$cabin_staff_rating)] <-
median(Data$cabin_staff_rating, [Link] = TRUE)
Data$ food_beverages_rating[[Link](Data$ food_beverages_rating)] <- median(Data$
food_beverages_rating, [Link] = TRUE)
Data$inflight_entertainment_rating[[Link](Data$inflight_entertainment_rating)] <-
median(Data$inflight_entertainment_rating, [Link] = TRUE)
#Data$ground_service_rating[[Link](Data$ground_service_rating)] <-
median(Data$ground_service_rating, [Link] = TRUE)
#Data$wifi_connectivity_rating[[Link](Data$wifi_connectivity_rating)] <-
median(Data$wifi_connectivity_rating, [Link] = TRUE)
Data$value_money_rating[[Link](Data$value_money_rating)] <-
median(Data$value_money_rating, [Link] = TRUE)

#----------------------Finding the outliers---------------------


boxplot(Data$overall_rating)$out
boxplot(Data$seat_comfort_rating)$out
boxplot(Data$cabin_staff_rating)$out
boxplot(Data$food_beverages_rating)$out
boxplot(Data$inflight_entertainment_rating)$out
#boxplot(Data$ground_service_rating)$out
#boxplot(Data$wifi_connectivity_rating)$out
boxplot(Data$value_money_rating)$out

#-------wifi rating and ground service has outliers and dropping the
columns-------------------------

df = subset(Data, select = -c(ground_service_rating,wifi_connectivity_rating) )

#Dropping the NA values from author_country

levels(df$author_country)[1] = NA
df<-[Link](df,cols="author_country")

#----------Finding the correlation among the variable-----------------------


round(cor(df[,11:17]),
digits = 2 # rounded to 2 decimals
)

#------------overall_rating & value_money_rating are correlated.. so dropping


overall_rating from the model--------

pos = df %>% group_by(airline_name) %>% summarise(AVG = mean(overall_rating))


# Using Linear Regression
model1 = lm(recommended ~ type_traveller+ cabin_flown + seat_comfort_rating +
cabin_staff_rating + food_beverages_rating + inflight_entertainment_rating +
value_money_rating, data = df )
summary(model1)

# Using Logistic Regression


model2 = glm(recommended ~ type_traveller+
cabin_flown+seat_comfort_rating+cabin_staff_rating+food_beverages_rating+
inflight_entertainment_rating+
value_money_rating,data=df,binomial(link = "logit"))
summary(model2)

#----------------Ans1- Both the model suggests value_money_rating &


cabin_staff_rating are of utmost importance to the customers---------------

#---------------------------Ans2- counntry specific decesion making


variables------------------------
library(dplyr)
fitted_models = dplyr::group_by(df,author_country) %>% dplyr::do(model =
lm(recommended~ seat_comfort_rating+cabin_staff_rating+food_beverages_rating+

inflight_entertainment_rating+

value_money_rating,data=.))
View(fitted_models$model)

#-----------------------------Ans3- counntry specific decesion making variables


forlufthansa airlines -----------------------------

pos3 = df %>% filter(airline_name == "lufthansa") %>% group_by(author_country) %>%


summarise(AVG = mean(overall_rating,[Link] = TRUE),CSR =
mean(cabin_staff_rating,[Link] = TRUE),

SCR = mean(seat_comfort_rating,[Link] = TRUE), IER =


mean(inflight_entertainment_rating,[Link] = TRUE),

VMR = mean(value_money_rating,[Link] = TRUE),FBR = mean(food_beverages_rating,[Link]


= TRUE))

# For example for Ireland the overall average is 4 but FBR is only one hence,
lufthansa should focus on food_beverages_rating in Ireland

You might also like