R Notebook

#Addition
2+3

[1] 5

5+10

[1] 15

7+7+7

[1] 21

#Subtraction
2-3

[1] -1

5-10

[1] -5

18-9-2

[1] 7

#Division
2/3

[1] 0.6666667

21/10

[1] 2.1

9/18

[1] 0.5

#Square root
sqrt(2)

[1] 1.414214

sqrt(5)

[1] 2.236068

sqrt(10)

[1] 3.162278

#Logarithms
log(2)

[1] 0.6931472

log10(5)

[1] 0.69897

log(6)

[1] 1.791759

#Question_1: Compute the log base 5 of 10 and the log of 10.

log(10, base = 5)

[1] 1.430677

log(10)

[1] 2.302585

#Batting Average=(No. of Hits)/(No. of At Bats)
#What is the batting average of a player that bats 29 hits in 112 at bats?
BA=(29)/(112)
BA

[1] 0.2589286

Batting_Average=round(BA,digits = 3)
Batting_Average

[1] 0.259

#Question_2:What is the batting average of a player that bats 42 hits in 212 at bats?
BA=42/212
BA

[1] 0.1981132

Batting_Average=round(BA,digits = 3)
Batting_Average

[1] 0.198

#On Base Percentage
#OBP=(H+BB+HBP)/(At Bats+BB+HBP+SF)
#Let us compute the OBP for a player with the following general stats
#AB=515,H=172,BB=84,HBP=5,SF=6
OBP=(172+84+5)/(515+84+5+6)
OBP

[1] 0.4278689

On_Base_Percentage=round(OBP,digits = 3)
On_Base_Percentage

[1] 0.428

#Question_3:Compute the OBP for a player with the following general stats:
#AB=565,H=156,BB=65,HBP=3,SF=7

OBP=(156+65+3)/(565+65+3+7)
OBP

[1] 0.35

3 == 8 #Does 3 equals 8?

[1] FALSE

4 == 6

[1] FALSE

4 == 4

[1] TRUE

3 != 8 #Is 3 different from 8?

[1] TRUE

3 != 9

[1] TRUE

3 != 3

[1] FALSE

3 <= 8 #Is 3 less than or equal to 8?

[1] TRUE

4 <= 9

[1] TRUE

4 <= 3

[1] FALSE

3>4

[1] FALSE

4>3

[1] TRUE

5>7

[1] FALSE

#Logical Disjunction (or)
FALSE | FALSE #False OR False

[1] FALSE

#Logical Conjunction (and)
TRUE & FALSE #True AND False

[1] FALSE

# Negation
! FALSE #Not False

[1] TRUE

#Combination of statements
2 < 3 | 1 == 5  #2<3 is True, 1==5 is False, True OR False is True

[1] TRUE

4 < 2 | 1 == 1

[1] TRUE

Total_Bases <- 6 + 5
Total_Bases*3

[1] 33

ls()

[1] "BA"                 "Batting_Average"    "OBP"                "On_Base_Percentage" "Total_Bases"       

rm(Total_Bases)

pitches_by_innings <- c(12, 15, 10, 20, 10) 
pitches_by_innings

[1] 12 15 10 20 10

strikes_by_innings <- c(9, 12, 6, 14, 9)
strikes_by_innings

[1]  9 12  6 14  9

#Question_4: Define two vectors,runs_per_9innings and hits_per_9innings, each with five elements. 

runs_per_9innings <- c(3, 5, 1, 4, 2)
runs_per_9innings

[1] 3 5 1 4 2

hits_per_9innings <- c(5, 8, 2, 6, 2)
hits_per_9innings

[1] 5 8 2 6 2

# replicate function
rep(2, 5)

[1] 2 2 2 2 2

rep(3, 5)

[1] 3 3 3 3 3

rep(1, 4)

[1] 1 1 1 1

rep(5, 10)

 [1] 5 5 5 5 5 5 5 5 5 5

# consecutive numbers
1:5

[1] 1 2 3 4 5

2:10

[1]  2  3  4  5  6  7  8  9 10

9:18

 [1]  9 10 11 12 13 14 15 16 17 18

#sequence from 1 to 10 with a step of 2
seq(1, 10, by=2)

[1] 1 3 5 7 9

seq(500, 1500, by=250)

[1]  500  750 1000 1250 1500

#add vectors
pitches_by_innings+strikes_by_innings

[1] 21 27 16 34 19

#compare vectors
pitches_by_innings == strikes_by_innings

[1] FALSE FALSE FALSE FALSE FALSE

#find length of vector
length(pitches_by_innings)

[1] 5

length(strikes_by_innings)

[1] 5

#find minimum value in vector
min(pitches_by_innings)

[1] 10

min(runs_per_9innings)

[1] 1

min(strikes_by_innings)

[1] 6

# find average value in vector
mean(pitches_by_innings)

[1] 13.4

mean(strikes_by_innings)

[1] 10

mean(runs_per_9innings)

[1] 3

pitches_by_innings

[1] 12 15 10 20 10

#If you want to get the first element:
pitches_by_innings[1]

[1] 12

#Question_5: Get the first element of hits_per_9innings.
hits_per_9innings[1]

[1] 5

pitches_by_innings[length(pitches_by_innings)]

[1] 10

runs_per_9innings[length(runs_per_9innings)]

[1] 2

#Question_6: Get the last element of hits_per_9innings.
hits_per_9innings[5]

[1] 2

hits_per_9innings[length(hits_per_9innings)]

[1] 2

pitches_by_innings [c(2, 3, 4)]

[1] 15 10 20

runs_per_9innings[c(3, 4, 5)]

[1] 1 4 2

player_positions <- c("catcher", "pitcher", "infielders", "outfielders")

data.frame(bonus = c(2, 3, 1),#in millions 
           active_roster = c("yes", "no", "yes"), 
           salary = c(1.5, 2.5, 1))#in millions 

sample(1:10, size=5)

[1]  7  8  9  1 10

sample(1:100, size = 10)

 [1] 43 30 32 18 74 87 40 89 81 90

bar <- data.frame(var1 = LETTERS[1:10], var2 = 1:10)
#Check data frame
bar

n <- 5

samplerows <- sample(1:nrow(bar), size = n)
#print sample rows
samplerows

[1] 8 6 3 5 2

#extract rows
barsample <- bar[samplerows, ]
#print sample
print(barsample)

  var1 var2
8    H    8
6    F    6
3    C    3
5    E    5
2    B    2

bar[sample(1:nrow(bar), n), ]

x <- c("Yes","No","No","Yes","Yes") 
table(x)

x
 No Yes 
  2   3 

sals <- c(12, .4, 5, 2, 50, 8, 3, 1, 4, 0.25)
#the average
mean(sals)

[1] 8.565

#the variance
var(sals)

[1] 225.5145

#the standard deviation
sd(sals)

[1] 15.01714

#the median
median(sals)

[1] 3.5

#Tukey's five number summary, useful for box-plots
# five numbers: min, lower hinge, median, upper hinge, max
fivenum(sals)

[1]  0.25  1.00  3.50  8.00 50.00

#summary statistics
summary(sals)

   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
  0.250   1.250   3.500   8.565   7.250  50.000 

#Function to find the mode, i.e. most frequent value
getMode <- function(x) {
     ux <- unique(x)
     ux[which.max(tabulate(match(x, ux)))]
 }

# Most frequent value in pitches_by_innings
getMode(pitches_by_innings)

[1] 10

getMode(hits_per_9innings)

[1] 2

#Question_7: Find the most frequent value of hits_per_9innings.
getMode(hits_per_9innings)

[1] 2

#Question_8: Summarize the following survey with the `table()` command:
#What is your favorite day of the week to watch baseball? A total of 10 fans submitted this survey.
#Saturday, Saturday, Sunday, Monday, Saturday,Tuesday, Sunday, Friday, Friday, Monday
game_day<-c("Saturday", "Saturday", "Sunday", "Monday", "Saturday","Tuesday", "Sunday", "Friday", "Friday", "Monday")

table(game_day)

game_day
  Friday   Monday Saturday   Sunday  Tuesday 
       2        2        3        2        1 

#Question_9: What is the most frequent answer recorded in the survey? Use the getMode function to compute results. 
getMode(game_day)

[1] "Saturday"