---

title: ‘WPA#3: Chapters 5-6’ author: “Rebekka Herz” date: “7. Mai 2015” output: html_document

1. Download the dataframe credit.csv from http://nathanieldphillips.com/wp-content/uploads/2015/05/credit.txt. The data are stored in a comma-separated text file with headers. Load the dataframe into an object called credit.

credit <- read.table(file = "http://nathanieldphillips.com/wp-content/uploads/2015/05/credit.txt", header = T, sep = ",", stringsAsFactors = F)
head(credit) #I imported it via the "Import Dataset function"

##   checking_balance months_loan_duration credit_history
## 1           < 0 DM                    6       critical
## 2       1 - 200 DM                   48           good
## 3          unknown                   12       critical
## 4           < 0 DM                   42           good
## 5           < 0 DM                   24           poor
## 6          unknown                   36           good
##                purpose amount savings_balance employment_duration
## 1 furniture/appliances   1169         unknown           > 7 years
## 2 furniture/appliances   5951        < 100 DM         1 - 4 years
## 3            education   2096        < 100 DM         4 - 7 years
## 4 furniture/appliances   7882        < 100 DM         4 - 7 years
## 5                  car   4870        < 100 DM         1 - 4 years
## 6            education   9055         unknown         1 - 4 years
##   percent_of_income years_at_residence age other_credit housing
## 1                 4                  4  67         none     own
## 2                 2                  2  22         none     own
## 3                 2                  3  49         none     own
## 4                 2                  4  45         none   other
## 5                 3                  4  53         none   other
## 6                 2                  4  35         none   other
##   existing_loans_count       job dependents phone default
## 1                    2   skilled          1   yes      no
## 2                    1   skilled          1    no     yes
## 3                    1 unskilled          2    no      no
## 4                    1   skilled          2    no      no
## 5                    2   skilled          2    no     yes
## 6                    1 unskilled          2   yes      no

2. How many rows and columns are in credit?

nrow(credit)

## [1] 1000

ncol(credit)

## [1] 17

dim(credit)

## [1] 1000   17

3. What are the names of the columns of credit?

names(credit)

##  [1] "checking_balance"     "months_loan_duration" "credit_history"      
##  [4] "purpose"              "amount"               "savings_balance"     
##  [7] "employment_duration"  "percent_of_income"    "years_at_residence"  
## [10] "age"                  "other_credit"         "housing"             
## [13] "existing_loans_count" "job"                  "dependents"          
## [16] "phone"                "default"

4. Add a new column to the dataframe called amount.eur that converts the loan amount in DM to euros. (1 DM is about .5 EUR)

credit$amount.eur <- credit$amount * 2
head(credit)

##   checking_balance months_loan_duration credit_history
## 1           < 0 DM                    6       critical
## 2       1 - 200 DM                   48           good
## 3          unknown                   12       critical
## 4           < 0 DM                   42           good
## 5           < 0 DM                   24           poor
## 6          unknown                   36           good
##                purpose amount savings_balance employment_duration
## 1 furniture/appliances   1169         unknown           > 7 years
## 2 furniture/appliances   5951        < 100 DM         1 - 4 years
## 3            education   2096        < 100 DM         4 - 7 years
## 4 furniture/appliances   7882        < 100 DM         4 - 7 years
## 5                  car   4870        < 100 DM         1 - 4 years
## 6            education   9055         unknown         1 - 4 years
##   percent_of_income years_at_residence age other_credit housing
## 1                 4                  4  67         none     own
## 2                 2                  2  22         none     own
## 3                 2                  3  49         none     own
## 4                 2                  4  45         none   other
## 5                 3                  4  53         none   other
## 6                 2                  4  35         none   other
##   existing_loans_count       job dependents phone default amount.eur
## 1                    2   skilled          1   yes      no       2338
## 2                    1   skilled          1    no     yes      11902
## 3                    1 unskilled          2    no      no       4192
## 4                    1   skilled          2    no      no      15764
## 5                    2   skilled          2    no     yes       9740
## 6                    1 unskilled          2   yes      no      18110

5. What is the median, mean, and standard deviation of the loan amounts in EUR?

median(credit$amount.eur)

## [1] 4639

mean(credit$amount.eur)

## [1] 6542.516

sd(credit$amount.eur)

## [1] 5645.474

6. What was the most common purpose for getting a loan? What was the least common reason?

table(credit$purpose)

## 
##             business                  car            education 
##                   97                  349                   59 
## furniture/appliances          renovations 
##                  473                   22

sort(table(credit$purpose))

## 
##          renovations            education             business 
##                   22                   59                   97 
##                  car furniture/appliances 
##                  349                  473

# I ran table and furniture/appliances are the most commmon purposes for getting a loan.

7. What percent of people got the loan for either education or a car?

mean(c(credit$purpose == "education" | credit$purpose == "car"))

## [1] 0.408

8. What percentage of the people had a good credit history?

table(credit$credit_history)

## 
##  critical      good   perfect      poor very good 
##       293       530        40        88        49

history.log <- c(credit$credit_history == "good")
#history.log
mean(c(history.log == "TRUE"))

## [1] 0.53

9. Of those people with a critical credit history, what was the median loan amount in EUR? What about for people with a good credit history?

critical.log <- c(credit$credit_history == "critical")
critical.data <- credit[critical.log,]
#critical.data
mean(critical.data$amount.eur)

## [1] 6176.075

good.log <- c(credit$credit_history == "good")
good.data <- credit[good.log, ]
mean(good.data$amount.eur)

## [1] 6081.917

10. Was there a relationship between a borrower’s age and their loan amount in EUR? Test this in two ways: once using a correlation, and once by comparing the average loan amount of people whose age is above the median age to those whose age is less than the median age.

names(credit)

##  [1] "checking_balance"     "months_loan_duration" "credit_history"      
##  [4] "purpose"              "amount"               "savings_balance"     
##  [7] "employment_duration"  "percent_of_income"    "years_at_residence"  
## [10] "age"                  "other_credit"         "housing"             
## [13] "existing_loans_count" "job"                  "dependents"          
## [16] "phone"                "default"              "amount.eur"

#1 correlation
cor(credit$age, credit$amount.eur)

## [1] 0.03271642

#2 comparison of the average loan amount of people whose age is above the median age to those whose age is less than the median age 

median(credit$age)

## [1] 33

over33.log <- c(credit$age > median(credit$age))
over33.data <- credit[over33.log, ]
#over33.data
mean(over33.data$amount.eur)

## [1] 6708.62

under33.log <- c(credit$age == median(credit$age))
under33.data <- credit[under33.log, ]
#under33.data
mean(under33.data$amount.eur)

## [1] 5318.667

setequal(mean(over33.data$amount.eur), mean(under33.data$amount.eur))

## [1] FALSE

11. Did people with a savings balance of less than 100 DM have a different default rate than those with a savings balance greater than 500 DM?

lessthan100.log <- c(credit$savings_balance == "< 100 DM")
lessthan100.data <- credit[lessthan100.log, ]