Skip to content

03-Explore-01.md

Latest commit

 

History

History
260 lines (186 loc) · 7.51 KB

03-Explore-01.md

File metadata and controls

260 lines (186 loc) · 7.51 KB

03 Explore Data Part I: Statistic and Datasets

Agenda

  • Review Data Types & Data Structures
  • Review Descriptive Statistics
  • Install Packages in R
  • Factors (One of data type)
  • Explore the datasets in base R
  • Explore datasets from CSV

0. Prerequisites

  1. Open R Studio Desktop and setting working directory
  2. Open Slide 03 Explore Data Part I
  3. Create Script file name 03_63130500xxx.R and save in your working directory
  4. run command install.packages("DescTools") (This command will use in Topic 2)
  5. Open progress form Class 03 and checked on Study Check-In and 0 prerequisite

Review Week 2: Data Type and Data Structure

# Data type
varA <- 100             # Numeric ex 1, 1.0
varB <- TRUE            # Logical: TRUE, FALSE
varC <- "Hello World!"  # Character

# Data Structure
vec1 <- c()             # Vector
list <- list()          # List
world <- data.frame()   # Data Frame

# Useful Functions
length()          # number of elements or components
str()             # structure of an object
class()           # class or type of an object
summary()         # result summaries

1. Review Descriptive Statistics

  • In R Programming, we can find descriptive statistics by using: mean(), median(), min(), max(), quartile(), range(), sd(), var(), iqr()
  • When we using function we need to understand which function is appropriate for data.

Try to measures of central value (วัดค่ากลาง) of this sample data:

# Sample datasets
sample1 <- c(29,49,42,43,32,38,37,41,27,27,30)
sample2 <- c(1,2,3,5,10,500)
sample3 <- c("Male","Female","Male","Female","Female")

If you observe, you will find that in base R not has function to find mode. The mode() in R base is using for get the type mode of an object. You need to finding manually. If the data have more than 1 million. How should you do ? We wouldn't be so mean. I want to present 2 ways that you will get mode:

  1. Using function table() (enhanced version of data.frames)
  2. Install Packages
### Cross Tabulation and Table Creation
table(sample1)

2. Install Packages in R

A package is a suitable way to organize your own work and, if you want to, share it with others. Typically, a package will include code (not only R code!), documentation for the package and the functions inside, some tests to check everything works as it should, and datasets. Ref: DataCamp

# Install Package
install.packages(name)

# Loading/Attaching and Listing of Packages
library(name)

# Read description
help(package=name)

Let's try to install Tools for Descriptive Statistics package (If you have problem, you can use GUI to install package.)

# DescTools package
install.packages("DescTools")

library(DescTools)
help(package=DescTools)

Now you can finding mode by using Mode() function

Mode(sample1)

Try to using this function with marvel_movies and tell me the result.

3. Factors

Factors is one of data type in R. It's appropriate for nominal value (ข้อมูลเชิงปริมาณ หรือ ข้อมูลที่ไม่ใช่ตัวเลข) such as: gender (male / female), Class (1st / 2nd / 3rd). We can use factor() function to see levels of object

factor(sample3)

# treats gender as a nominal variable
gender <- factor(sample3)
summary(gender)

Ref: https://www.statmethods.net/input/datatypes.html

4. Explore the datasets in base R

In base R has built-in datasets. To see the list of datasets, you can using data() function. The popular used R demo datasets is mtcars, iris, ToothGrowth, PlantGrowth and USArrests.

data()

Explore overview dataset

head()      # See 6 observation (rows)
str()       # Structure of object
colnames()  # Column Name
rownames()  # Row Name
colMeans()  # Mean each column
rowMeans()  # Mean each row
colSums()   # Sum each column
rowSums()   # Sum each row

Data set: Motor Trend Car Road Tests

We will use mtcars datasets to explore data

View(mtcars)

str(mtcars)
head(mtcars)
summary(mtcars)
colMeans(mtcars)

Inspect inside dataset

We can select one column to calculate some thing by using df$col. Example, we want to finding average of miles per gallon in Motor Trend Car Road Tests.

# Find mean of miles per gallon (mpg)
mean(mtcars$mpg)  # 20.09062

Self Study:

Moreover, we can select row and column by using format: df[row,col]

mtcars[1,]                  # Select first row with all column
mtcars[,1]                  # Select all row with only first column
mtcars["Datsun 710","wt"]   # Select by specific row and column
mtcars[1:5,]                # Select row 1-5 with all column

5. Explore datasets from CSV

In R, we can import datasets from CSV (comma-separated values) files by using function read.csv(). If you import from local you need to set path correctly.

Hotels <- read.csv("https://raw.githubusercontent.com/safesit23/INT214-Statistics/main/datasets/Hotels.csv")
View(Hotels)

Extra

In R has library named readr that can read CSV file by using function read_csv()

Exercise:

  • How many observation ? (จำนวนข้อมูล)
  • What is the mean hotel room rate ?
  • What is the median hotel room rate ?
  • What is the mode ?

Dataset from: Chapter 3, page 92 – 93, Statistic for Business and Economics

Notes

Useful Function in week 3

View(obj)         # Invoke a Data Viewer
table(obj)        # Cross Tabulation and Table Creation
library(obj)      # Loading/Attaching and Listing of Packages
Mode(obj)         # Mode in descriptive stat
data()            # See all datasets in R
head(obj)         # See the First 6 observation of an Object
factor(obj)       # Category

Library in week 3

library(DescTools)    # DescTools package
library(MASS)         # Support Functions and Datasets for Venables and Ripley's MASS

Self Learning Activity:

A. Explore more datasets

1. Install and loading MASS package

install.packages("MASS")
library(MASS)

2. Explore this 2 dataset that given

3. Finding descriptive statistics and summary result in form of sentences/paragraph. For example,

  • From survey, most students use left hand writing.
  • The average of body weight of cats is xxx kg.

After you have done, you need to sharing and working with your group as Assignment 2.

B. What is NA Values ?

In the survey dataset, it has NA values. NA mean ‘Not Available’ / Missing Values. If you try to find mean(), it will get result NA because in that vector has NA value. You need to handle NA values by add parameter na.rm in function. For example:

nums <- c(1,2,3,4,NA)
mean(nums,na.rm = TRUE)

C. Pipes Operator in R

  • You need to install and loading dplyr package
  • f(x) can be rewritten as x %>% f
  • In short, this means that functions that take one argument, function(argument), can be rewritten as follows: argument %>% function(). Ref: Data Camp

Try to use %>% operator and see the result. Is it same ?

# In class
mean(mtcars$mpg)

# Use pipes operator
mtcars$mpg %>% mean()