yabyzq

library(ggplot2)

#Get a small proportion to plot
dsample <- diamonds[sample(nrow(diamonds),500),]


#qplot(x, y, data)
qplot(carat,price,data=dsample)
#Add more measures
qplot(carat,price,data=dsample, color = color, shape =cut)

yabyzq

1. Search "ABS Housing finance"

2. Download data. TABLE 1. HOUSING FINANCE COMMITMENTS (Owner Occupation), By Purpose: Australia (Number, $000)

3. Clean up unused rows.

4. Pivoting

5. Switch it to month, Adjust Axis range, Add Color, Adjust line size

yabyzq

#Prepare training and test data
test_index <- which (1:length(iris[,1])%% 5 == 0)
iris_train <- iris[-test_index, ]
iris_test <- iris[test_index, ]
library(car)
test_index <- which (1:nrow(Prestige)%% 4 == 0)
prestige_train <- Prestige[-test_index, ]
prestige_test <- Prestige[test_index, ]

#Linear Regression --log(income) + education

yabyzq

library(mice)
library(missForest)
library(VIM)
library(Amelia)
library(Hmisc)
library(mi)

#Generate missing value using prodNA from missForest
data <- iris
summary(iris)

yabyzq

#Sampling method
#1. Undersampling
#2. Oversampling
#3. Synthetic - Smote

#Ensembling method
#1. BalanceCascade, Keep	removing	majority	class	examples	un4l	none	is	miss-classified	
#2. EasyEnsemble, ensemble different balanced model

#Cost-based method - Cost FN >> cost FP

yabyzq

#1 Numeric Variable
numeric_stats <- function (x, na.omit =FALSE){
  if(na.omit)
    x <-x[!is.na(x)]
  m <- mean(x)
  a <- median(x)
  s <- sd(x)
  min <- min(x)
  max <- max(x)
  return(c(min = min, mean = m, avg = a, max = max, sd = s))

yabyzq

library(dplyr)
library(nycflights13)

#look at the data
tbl_df(flights)
#filter
filter(flights, month == 1, day ==1,!is.na(month))
#arrange
arrange(flights, desc(year))

yabyzq

# Basic scatter plot of vocabulary (y) against education (x). Use geom_point()
ggplot(Vocab, aes(education, vocabulary))+geom_point()


# Use geom_jitter() instead of geom_point()
ggplot(Vocab, aes(education, vocabulary))+geom_jitter()


# Using the above plotting command, set alpha to a very low 0.2
ggplot(Vocab, aes(education, vocabulary))+geom_jitter(alpha = 0.2)

yabyzq

library(rpart)

#Generate binary Target
iris$isSetosa <- "N"
iris[iris$Species == "setosa",]$isSetosa <- "Y"
iris$isSetosa <- as.factor((iris$isSetosa))
levels(iris$isSetosa) <- c("N","Y")
head(iris)

#Create function

yabyzq

library(caret)

#looking at missing value
options(digits=2)
stats <- data.frame(missing = sapply(iris, function(x) sum(is.na(x))),
           mean = sapply(iris, function(x) if(is.numeric(x)) {mean(x, na.rm = T)} else names(table(x)[order(table(x), decreasing =  T)])[1])
)