Homework1
2024-04-02
Boris Davidovic - Homework 1
2)/3)
athleteData <- read.table("./AthleteData.csv", header=TRUE, sep=",", dec=".")
head(athleteData)## S.NO Name Nationality Current.Rank Previous.Year.Rank Sport
## 1 1 Mike Tyson USA 1 n/a boxing
## 2 2 Buster Douglas USA 2 n/a boxing
## 3 3 Sugar Ray Leonard USA 3 n/a boxing
## 4 4 Ayrton Senna Brazil 4 n/a f1
## 5 5 Alain Prost France 5 n/a f1
## 6 6 Jack Nicklaus USA 6 n/a golf
## Year Earnings.USD
## 1 1990 28.6
## 2 1990 26.0
## 3 1990 13.0
## 4 1990 10.0
## 5 1990 9.0
## 6 1990 8.6tail(athleteData)## S.NO Name Nationality Current.Rank Previous.Year.Rank Sport
## 296 296 LeBron James USA 5 8 basketball
## 297 297 Stephen Curry USA 6 9 basketball
## 298 298 Kevin Durant USA 7 10 basketball
## 299 299 Tiger Woods USA 8 11 golf
## 300 300 Kirk Cousins USA 9 >100 NFL
## 301 301 Carson Wentz USA 10 >100 NFL
## Year Earnings.USD
## 296 2020 88.2
## 297 2020 74.4
## 298 2020 63.9
## 299 2020 62.3
## 300 2020 60.5
## 301 2020 59.1The data surrounds the highest paid athletes in the world in a given year from 1990-2020. There are a total of 301 observations. Each year has 10 records which indicate the top 10 highest paid athletes of that year, where they ranked in the previous year, what country they are from, what sport they played and what country they are from. Here is the variables the data described in more detail:
S.NO -> int value, indication of what row of data it is
Nationality -> chr value, what country the athlete is from
Current.Rank -> int value, what is their position in terms of pay of that specific year (i.e. 1 means they are the highest paid athlete of the year corresponding to the year value in the row)
Previous.Year.Rank -> chr value, where did they rank in terms of highest paid athletes in the previous year. The data consists of a mix of values: if they were in the top 10 the previous year, they have a clear indication of where they were ranked. If they were ranked, but outside the top 10, the data indicates approximately where they rank (i.e. >30, >40, etc.). Some of the data also has “not ranked” values which shows they were not ranked in the previous year. “n/a” is also a value used for values corresponding to 1990 as that is the first year of records
Sport -> chr value, what sport did the athlete play
Year -> int value, what year was the record in
Earnings.USD -> dbl value, how much the athlete made in that year in USD
The source of the data is from kaggle.com (https://www.kaggle.com/datasets/gaurav126/forbes-richest-athletes-1990-2021). The dataset on kaggle.com is extracted from topendsports.com
summary(athleteData[, -1])## Name Nationality Current.Rank Previous.Year.Rank
## Length:301 Length:301 Min. : 1.000 Length:301
## Class :character Class :character 1st Qu.: 3.000 Class :character
## Mode :character Mode :character Median : 5.000 Mode :character
## Mean : 5.449
## 3rd Qu.: 8.000
## Max. :10.000
## Sport Year Earnings.USD
## Length:301 Min. :1990 Min. : 8.10
## Class :character 1st Qu.:1997 1st Qu.: 24.00
## Mode :character Median :2005 Median : 39.00
## Mean :2005 Mean : 45.52
## 3rd Qu.:2013 3rd Qu.: 59.40
## Max. :2020 Max. :300.00Name -> categorical variable so cannot take descriptive statistics of it, insights would be better seen through graphs
Nationality -> also a categorical variable so cannot take descriptive statistics of it, insights would be better seen through graphs
Current.Rank -> this is technically a categorical variable as it is tied to rank, but the outputted descriptive statistics makes sense - the min is 1 and max is 10 which demonstrates it is a rank between 1 and 10
Previous.Year.Rank -> also a categorical variable so cannot take descriptive statistics of it, insights would be better seen through graphs
Sport -> also a categorical variable so cannot take descriptive statistics of it, insights would be better seen through graphs
Year -> also technically a categorical variable as it is tied to the year, but the outputted descriptive statistics makes sense - the min is 1990, max is 2020 and mean is 2005 which indicates the data is from 1990 to 2020
Earnings.USD -> the median of the earnings is 39 which means that more than half of these athletes make over $39 Million. The max is 300 which means there could be some large outliers in the data. Combined with the median being 39, it is likely the data is skewed to the right
library(ggplot2)
ggplot(athleteData, aes(y = Earnings.USD, x = Nationality)) + geom_boxplot() + theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
labs(x = "Nationality", y = "Earnings (USD)", title = "Earnings by Nationality")
This graph shows earnings by nationality. We can see that the highest earning country is the United States - there is also a lot of outliers with the US data, indicating that most of the highest earning athletes are from the US.
Many countries such as Austria, Serbia, Spain and more have only one black line in the plot, indicating the country has one athlete who has been in the top 10 of athletes in terms of earnings at any point in the data set. Some countries such as Argentina, Brazil and Philippines have the black line closer to the lower part of the box (the 25th percentile) which shows the earnings for that country is right skewed (a higher number of athletes earn more than the median). Countries such as Portugal, Switzerland and the UK have the black line closer to the top part of the box (75th percentile) indicating that the earnings for that country are left skewed (a higher number of athletes earn less than the median).
library(ggplot2)
ggplot(athleteData, aes(y = Earnings.USD, x = Sport)) + geom_boxplot() + theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
labs(x = "Nationality", y = "Earnings (USD)", title = "Earnings by Sport")
This graph shows earnings by sport. From this graph we can see that the
highest earning sport is boxing and has outliers that are far above some
other data points. From boxing, we can see the next highest earning
sports are soccer, tennis and golf with tennis and golf being fairly
similar in terms of earnings.
Most of the sports are slightly or fairly skewed towards the right in terms of earnings as many of the boxplots have the black line (median) towards the 25th percentile. This indicates that most sports have a few individuals that earn quite more than their counterparts which ultimately skews the data towards the right.
library(car)## Loading required package: carDatascatterplot(y = athleteData$Earnings.USD,
x = athleteData$Year,
ylab = "Earnings in Millions $USD",
xlab = "Year",
smooth = FALSE)
This is a scatter plot that shows the relationship between the year and
earnings. We can see from this data that there is a fairly strong
relationship that as time has progressed, the earnings of top athletes
has also increased. But there are some outliers in the data such as in
2015 and and 2018. In 2015, there a fight between Floyd Mayweather and
Manny Pacquiao where they earned incredible sums of money, creating the
outlier. It also looks like Mayweather had a fight in 2018 where he
earned just under 300M USD.
Also, the box plot on the y-axis shows a right skewed distribution which is consistent with the above graphs where most sports by earnings are right skewed. This can also be shown with more points above the line of best fit than below. Overall, this indicates that there are certain individual athletes that make much more than the rest which skews the data.
ggplot(athleteData, aes(x = Nationality)) +
geom_bar(fill = "skyblue", color = "black") +
theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
labs(x = "Nationality", y = "Frequency", title = "Frequency Plot by Nationality")
This is a graph showing the frequency of nationalities in this data set.
From this graph we can see that USA makes up a significantly large
portion of the athletes. This can likely be attributed to the USA having
the Big 4 North American sports leagues (NBA, NFL, NHL, MLB). Most of
the other countries have a small number of athletes or only an
individual athlete in this data set.