#Vectors have a linear structure.  Most of our datsets are rectangular.

#Different ways to import data:
#In R Studio, use the "Import Dataset" on the right

#rename the dataframe (unless you like the default name)
d <- john_data

#other methods allow more flexibility
#use the "file.choose" command
d <- read.csv(file.choose(), stringsAsFactors = FALSE, as.is=TRUE, strip.white=TRUE)

#load in the data yourself thorugh the console
d <- read.csv("c:\\directoryname\\filename.csv") #pc
d <- read.csv("~/Documents/Postdoc/Teaching/Intro to R/data files/john_data.csv") #mac

#look at the data to make sure it makes sense
names(d)    #look at the column names
rownames(d) #defaults to numbers if no names are assigned
head(d)     #look at the first six lines 
tail(d)     #look at the last six lines

#change the column names
names(d) <- c("Lineage", "Time", "DT", "Ploidy", "Enviro")
names(d)[5] <- "Env"


#As it reads your data, R will classify your variables into types:
#Columns with only numbers are make into numeric or integer variables
#Columns with non-numeric characters are made into factors unless otherwise specified ("stringsAsFactors=FALSE" in the read command)
str(d)      #check how R has classified the variables

#check on R's classification of just one variable
class(d$Ploidy)         #integer, character, numeric, etc.
is.numeric(d$Ploidy)    #TRUE or FALSE
is.character(d$Ploidy)  #TRUE or FALSE
is.integer(d$Ploidy)    #TRUE or FALSE

#convert bewteen character types
d$Ploidy <- as.character(d$Ploidy)
d$Ploidy <- as.factor(d$Ploidy)
d$Ploidy <- as.numeric(as.character(d$Ploidy))
d$DT <- as.numeric(as.character(d$DT))

#Useful data fram functions and operations
ncol(d)         #number of columns
nrow(d)         #number of rows
length(d)       #number of variables (NOT the length of the columns = nrow)

#Access aspects of a data frame
#Vectors can be refered to by name
d$Env           #the fifth (environment) vector
d$E             #the fifth (environment) vector (none of the other columns start with 'E')
d[,"Env"]       #the fifth vector

#Variables can also be accessed using square brackets.  Integers before the comma refer to rows, integers after the comma indicate columns: mydata[rows, columns]
d[,5]           #the fifth vector
d[1,]           #the first row
d[1:10,]        #the first ten rows
d[1,3]          #the third element in the first row
d[,3][1]        #the third element in the first row
d$DT[1]        #also the fifth element in the first row
d[1:4,c(3,5)]   #the second and fifth elements in the first four rows

#Change values within a data frame
d[,3][1] <- 2

#Create a new variable based on properties of an existing variable:
d$PloidyFac[d$Ploidy > 3.5] <- "tet"
d$PloidyFac[d$Ploidy < 3.5 & d$Ploidy > 2.5 ] <- "trip"
d$PloidyFac[d$Ploidy < 2.5 & d$Ploidy ] <- "dip"
is.factor(d$PloidyFac)
d$PloidyFac <- as.factor(d$PloidyFac)
str(d)

#Combine two vectors together
d$timeEnv <- paste(d$Time, d$Env, sep=".")

#saving this for later (since I use "d" a lot)
john <- d

#Combine vectors into a data frame.  Vectors need to be the same length, or multipliers of the length
tlo <- c("TLO1", "TLO2","TLO3", "TLO4", "TLO5",  "TLO7", "TLO8", "TLO9", "TLO10")
clade <- c("alpha", "beta", "alpha", "gamma", "gamma", "gamma", "alpha", "alpha", "alpha")
chr <- c("R","R","1","1","2", "3","3","4","4")
express <- c(10.2, 0.8, 0.9, 0.2, 0.3, 0.000001, 1.3, 2, 0.1)
loc <- "subtel"
tlos <- data.frame(tlo, clade, chr, express, loc, stringsAsFactors = FALSE)

tlo34 <- c("TLO34", "alpha", "1", "1", "internal")

#add a row
tlos <- rbind(tlos, tlo34)
str(tlos)

#add a column
env <- "YPD"
tlos <- cbind(tlos, env)

#can also do tlos$env <- "YPD"

#################################################################################
#################################################################################
#1) Oops, TLO8 is a gamma, not an alpha.  Correct that.
#2) Make a new vector that contains the log of expression values.
#2) We also collected expression data in fluconazole.  Add the information below into the existing dataframe without adding new columns (multiple ways to do this).
#expression in fluconazole:
# TL01 - 3
# TL02 - 0.4
# TL03 - 2.4
# TL04 - 0.9
# TL05 - 1.1
# TL07 - 0.23
# TL08 - 0.42
# TL09 - 4
# TL010 - 100
# TL034 - 43.2

#################################################################################
#################################################################################

#Sort and order
tlos.x <- tlos[order(tlos$clade),]      #by one column
tlos.x <- tlos[order(tlos$clade, tlos$express),]      #by two columns

#Obtain a subset of a dataframe using logical statements
atlo <- subset(tlos, clade=="alpha")
bgtlo <- subset(tlos, clade!="alpha")
aRtlo <- subset(tlos, clade=="alpha" & chr=="4")

#Write/save a data frame to a text file.  Include "row.names=FALSE" if you don't want the row names or number of the data frome included in the first column
write.csv(d, file='c:\\directoryname\\filename.csv', row.names=FALSE) #PC
write.csv(d, file="/directoryname/filename.csv", row.names=FALSE) #Mac