$14\mathrm{.}11$ HW$2\_$LAB$4$: Prediction with Logistic Regression Run your program as often as you'd like, before submitting for grading. Below, type any needed

input values in the first box, then click Run program and observe the program's output in the

second box.

This lab will be available until June $21$st$,11$:$59$ PM EDT

The file Invistico$\_$Airline$\_$LR$.$csv contains information from an airline using the alias Invistico Airline on customer satisfaction, as well as

details on each customer. The columns of interest are Gender, Age, Class, Arrival$\_$Delay$\_$in$\_$Minutes, and satisfaction.

Read the file Invistico$\_$Airline$\_$LR$.$csv into a data frame.

Obtain user defined values female, age, economy, and delay.

Re$-$code the categorical variables Gender, Class, and satisfaction into dummy variables.

Create a new data frame X from the predictor variables Gender$\_$female, Age, Class$\_$Eco, and Arrival$\_$Delay$\_$in$\_$Minutes, in that order.

Create a response variable $Y$ from the dummy variable satisfaction$\_$satisfied.

Perform logistic regression on $x$ and $Y.$

Use the user defined values to predict the probability that a customer with those values is satisfied.

Ex: If the input is $134010$ the ouput is:

$[0\mathrm{.}62343979]$

here is the code:

# import the necessary modules

import pandas as pd

female $=$ int$($input$())$

age $=$ float$($input$())$

economy $=$ int$($input$())$

delay $=$ float$($input$())$

# read in the file Invistico$\_$Airline$\_$LR$.$csv

flights $=$ pd$.$read$\_$csv $(\text{'}$Invistico$\_$Airline$\_$LR$.$csv$\text{'})$

# remove missing data

flights.dropna$($axis $=0,$ inplace $=$ True$)$

# recode the categorical variables Gender, Class, and satisfaction as dummy variables

flights $=$ pd$.$get$\_$dummies $($flights$,$ columns$=[\text{'}$Gender$\text{'},$ 'Class', 'satisfaction'$],$ drop$\_$first $=$ True$)$

# create a new data frame from the variables Gender$\_$Female, Age, Class$\_$Eco, and Arrival$\_$Delay$\_$in$\_$Minutes, in that order.

X $=$ flights$[[\text{'}$Gender$\_$female', 'Age', 'Class$\_$Eco', 'Arrival$\_$Delay$\_$in$\_$Minutes'$]]$

X $=$ add$\_$constant$($X$)$

Y $=$ flights$[\text{'}$satisfaction$\_$satisfied'$]$

# set Y as the response variable satisfaction$\_$satisfied

# perform logistic regression on X and Y

model $=$ Logit$($Y$,$ X$).$fit$()$

# create an array with $1$ for the intercept, and the user input values female, age, economy, and delay

ex $=[1,$ female, age, economy, delay$]$

prediction $=$ model.predict$($ex$)$

# find the predicted probablility that a customer with the user input values is satisfied

print$($prediction$)$