Introduction to R - Merging and Aggregating Data

#Below is the sample data we will be creating 2 dataframes  
key=(1:10)

#Here we are passing the row names and column names as a list. 
m<- data.frame(matrix(rnorm(40, mean=20, sd=5), nrow=10, ncol=4, dimnames=list((1:10),c("a","b","c","d"))))
m2<- data.frame(matrix(rnorm(40, mean=1000, sd=5), nrow=10, ncol=4, dimnames=list((1:10),c("e","f","g","h"))))

#This is one way of combining a vector with a dataframe. 
df<-  data.frame(key,m)
df2<- data.frame(key,m2)

#This is another way way of combining a vector with a dataframe. 
dfb<-  cbind(key,m)
df2b<- cbind(key,m2)

df
df2
dfb
df2b

# This manages the merge by an associated key.
df3 <- merge(df,df2,by="key")
# This just does a "column bind" 
df4<- cbind(df,df2)
df5<- data.frame(df,df2)
df3
df4
df5

#Here we can combine rows with rbind. 
df5<-df
#The make Row
df6<-rbind(df,df5)
df6
df7<-rbind(df,df5, make.row.names=FALSE)
df7

df7

iris=read.csv(file="../../input/iris.csv", header=TRUE,sep=",")
head(iris)

iris<-read.csv(file="../../input/iris.csv", header=TRUE,sep=",")

#Aggregate by Species  aggregate(x, by, FUN, ...)
iris.agg<-aggregate(iris[,1:4], by=list("species" = iris$species), mean)
print(iris.agg)

#Notice this gives us the same output but structured differently. 
by(iris[, 1:4], iris$species, colMeans)

iris<-read.csv(file="../../input/iris.csv", header=TRUE,sep=",")
iris$sum<-apply(iris[1:4], 1, sum) #This provides a sum across  for each row. 
iris$mean<-apply(iris[1:4], 1, mean)#This provides a mean across collumns for each row. 
head(iris)
apply(iris[1:4], 2, mean)

# create a list with 2 elements
sample <- list("count" = 1:5, "numbers" =5:10)

# sum each and return as a list. 
sample.sum<-lapply(sample, sum)

class(sample.sum)
print(c(sample.sum, sample.sum["numbers"],sample.sum["count"]))

# create a list with 2 elements
sample <- list("count" = 1:5, "numbers" =5:10)

# sum each and return as a list. 
sample.sum<-sapply(sample, sum)

class(sample.sum)
print(c(sample.sum, sample.sum["numbers"],sample.sum["count"],sample.sum[["count"]]))

#Note the differenece between #sample.sum[["count"]] and sample.sum["count"]

# We can also utilize simple 
square<-function(x) x^2
square(1:5)

# We can use our own function here.     
sapply(1:10, square)

#We can also specify the function directly in sapply.
sapply(1:10, function(x) x^2)

#Tapply example
#tapply(X, INDEX, FUN, …) 
#X = a vector, INDEX = list of one or more factor, FUN = Function or operation that needs to be applied. 
iris<-read.csv(file="../../input/iris.csv", header=TRUE,sep=",")
iris.sepal_length.agg<-tapply(iris$sepal_length, iris$species, mean)
print(iris.sepal_length.agg)