Class 2 Notes

Question 1

R Studio Software

Answer 1

The console is where you can implement your commands.
However, if you simply write in commands there, it will not save them. It is thus generally better to write your commands into the Source window, highlight the code that you want to run, and click on CTRL and Enter at the same time to execute your commands.
The Environment panel is where you will look at your datasets, etc. You will use the
Files, Plots, Packages, Help, and Viewer panes a lot.

Question 2

Setting the Working Directory

Answer 2

Clear environment: rm(list=ls(all=TRUE))
Set the working directory setwd(“C:/Quant)

Question 3

Create a new project

Answer 3

File -> New Project: choose directory

Question 3

To enter a vector for country

Answer 3

country = c(“france”,”france”,”france”,”france”,”france”, “france”)

Question 3

Vector for year

Answer 3

year = c(2000,2005,2010,2015,2020,2025) or year2 <- seq(2000, 2025, 5)

Question 4

Inspect vector

Answer 4

print(vector name)

Question 4

Remove vector or data frame

Answer 4

rm(low_high)

Question 5

To summarize, we have created vectors of different classes/types, which we can check

Answer 5

class(country) - character
class(year) - numeric
class(poverty_levels) - factor

Question 6

Turn vectors into data frame

Answer 6

df = data.frame(country,year,poverty_rate,poverty_levels)

Question 7

Run a library

Answer 7

Library()

Question 8

Import dataframe

Answer 8

df = import(“data/france_poverty.csv”) # csv
df = import(“data/france_poverty.Rdata”) # R data file
df = import(“data/france_poverty.dta”) # Stata data file
df = import(“data/france_poverty.xlsx”, which=1) # Excel

Question 9

Inspect Data

Answer 9

View(df) # look at the whole thing
head(df) # get a quick view of the first few observations
dim(df) # check how many variable/dimensions (6 observations, 4 variables)
length(df$poverty_rate) # count the number of observations
unique(df$poverty_rate) # list unique values
table(df$country) # inspect variable

Question 10

Summary statistics

Answer 10

summary(df)

Question 11

To take into account that something has missing values (NA) by adding na.rm=TRUE

Answer 11

add na.rm=TRUE to the end of our command

Question 12

Mean

Answer 12

mean(df$poverty_rate) or mean(df)

Question 13

Standard Deviation

Answer 13

sd(df$poverty_rate) or sd(df)

Question 14

Variance

Answer 14

var(df$poverty_rate) or var(df)

Question 15

Calculate mean

Answer 15

sample mean = ̄ 𝑥 = (∑ 𝑥𝑖)/𝑁 = (2 + 4 + 6)/3 = 12/3 = 4

Question 16

Calculate sample variance

Answer 16

sample variance = (∑ (𝑥𝑖 − ̄ 𝑥)2)/(𝑁 − 1)

Question 17

Calculate sample standard deviation

Answer 17

sample standard deviation = √((∑ (𝑥𝑖 − ̄ 𝑥)2)/(𝑁 − 1))

Question 18

Drop NA

Answer 18

library(tidyverse)
df <-
df %>%
drop_na(poverty_rate)

Question 19

Piping

Answer 19

The pipe command, %>%, is telling R to use whatever is in this line of code as the basis for the next line of code.

Question 20

Nice table in R

Answer 20

library(modelsummary)
datasummary_skim(df) or
library(modelsummary)
datasummary_skim(df, output = “data.frame”) or
library(modelsummary)
datasummary(poverty_rate + year ~
NUnique + PercentMissing + Mean + SD + Var +
Min + Median + Max,
data = df)

Question 21

Every ggplot2 has

Answer 21

ggplot(data = , aes(x = , y = )) +
and geom_()

Question 22

Line graph

Answer 22

ggplot(data = df, aes(x=year, y=poverty_rate)) + geom_line()

Question 23

Fancy Line Graph

Answer 23

ggplot(data = df, aes(x=year, y=poverty_rate)) + geom_line() + labs( x = "Year", y = "Poverty Rate", title = "Poverty Rate in France") + theme(plot.title = element_text(hjust = 0.5))

Question 24

Bar Graph

Answer 24

ggplot(df, aes(x=year, y=poverty_rate)) + geom_bar(stat = "identity") + labs( x = "Year", y = "Poverty Rate", title = "Poverty Rate in France") + theme(plot.title = element_text(hjust = 0.5))

Question 25

Stat identity

Answer 25

The only major change would be stat=identity inside the parentheses to our geom_bar(). That stat=identity specification is necessary because, by default, geom_bar() just assumes that you want a count of the number of rows.

Question 26

Drop NA

Answer 26

df <- df %>% dplyr::select(-c(poverty_levels))

Question 27

Look at data set

Answer 27

print(df)

Question 28

Put Data into wide format

Answer 28

wide = df %>% pivot_wider(names_from = "year", values_from = c("poverty_rate"))

Question 29

Put data into long format

Answer 29

long <- wide %>% pivot_longer(cols = c(`2000`, `2005`, `2010`, `2015`, `2020`), names_to = "year", values_to = "poverty_rate")

Question 30

Summarize mean variable

Answer 30

collapse <- df %>% summarize(mean_poverty_rate = mean(poverty_rate)) print(collapse)

Question 31

Summarize

Answer 31

collapsed_group <- df %>% group_by(country) %>% summarize(mean_poverty_rate = mean(poverty_rate)) %>% ungroup()

Question 32

ifelse()

Answer 32

It presents a condition (e.g. 𝑥 = 1), and tells the computer to replace the value with one thing if the condition is true or something else if the condition is false.

Question 33

ifelse() example

Answer 33

df$decade2020s = ifelse(df$year >=2020,1,0) print(df)