poli380 - lecture 1-5

layout of r studio

r script = top left

environment = top right

r console = bottom left

help and plots = bottom right

r script

write and run code

r console

where r puts executed code and outputs (as well as errors)

what colour is executed code

blue

what colour is output of executed code

black

[1] thingy meaning

means the output to its right is the first output in this case

r script ___ code

saves

<

less than

<=

less than or equal to

!=

not equal to

==

must use both of these for conditionals!!!!!!

like “is this true or nah”

and symbol

&

or symbol

|

what’s the caveat for the or symbol

it won’t tell you which one is true

what does creating objects do

it’s how r stores info

like a box that contains something

object operator

←

sequence for creating an object

object_name ← object_content

how do i find the content of the object

object_name in R!!!!!

it’s basically saying what is inside this

where will the content of the object show up

environment!!!!!

rules for naming objects

cannot begin with a number/have spaces/special symbols

for textual content, use quotes

why do we use quotes for text

if you don’t it’ll assume the text is an object. and it' won’t work b/c the object hasn’t been named

when do we use quotes when writing code

not for names of objects, functions, arguments or special values

u use it for all other texts unless this choice is intentional

what’s this weird thing that r will do (hint: override)

it’ll override objects if you assign new content to existing object name

also r is case sensitive

what are vectors

objects with a series of values

“multiple elements of the same type”

what function do vectors use

c() COMBINE BABY

what are functions

actions you require r to perform on particular object or data

takes inputs → action → output

functions can also be applied to vectors

what if you have a combined function with “na”

no value so if u wanna calculate mean or smth u gotta deal w that

mean/median when you have na

mean(y, na.rm = TRUE)

median(y, na.rm = TRUE)

what is a name of a function followed by

function_name

PARANTHESIS

how many required arguments are needed (at least)

1, for most functions

optional arguments

are optional. not rlly necessary unless needed

what do you do if you have multiple arguments

function_name(argument 1, argument 2)

how do you specify arguments

put priority or include number of arguments in in specifics

if more than 1 req arg, put em in that order

okay what are the two formats we r using

function_name (required_argument, optional_argumentname = optional_argument)

function_name(required_argument)

packages

extend the functionality of r

what do we do with packages

install them + load

(installing is only done once tbh, load everytime)

sequence for installing packages

install.packages (“causaldata”)

loading packages

library(causaldata)

#

commenting code; r ignores everything after # until end of line

what are datasets/data

they capture characteristics of a particular set of individuals or entities

how are datasets organzied

as dataframes

rows =

observations

columns

variables

what are observations

information collected from a particular entity or individual in the study

unit of observation

defines the individual or entity that each observation in a dataframe represents; each observation is a row number, denoted as i

what are variables

values of changing characteristics for various individuals and entities in the study

how do we refer to variables

by its name

notation for defining a new variable

x = {10, 5, 8}

where x is the name of the variable

and the numbers are the content of the variables; multiple observations

each individual observation is

iiiiiiii

the observation number

what is data wrangling

allows us to make data in the form we want

tibble

how tidyverse stores dataframes

what are the three methods for datasets within r

some are built in, others are acquired via loading r packages, and others — we read em in ourselves

names()

gives us the names of columns

head()

first 6 rows and observations in a dataset

dim()

dimensions of a dataset

number of rows and columns

n(row)

number of rows

n(col)

number of columns

View() (w a capital)

contents of the entire dataset

what tf is dplyr

gives us a consistent set of verbs for data manipulation

how to get that dplyr

install.packages (“dplyr”)

library(dplyr)

pipe

used to write code sequentially %>%

base r vs the pipe

base r sqrt(sum(range(student_ages)))

pipe

student_ages %>%

range()%>%

sum()%>%

sqrt()

filter()

selects rows that satisfy the argument we provide it

how do we save as a new dataset

we must create a new object

subsets in base r

flights_AA2 ← flights [flights $ carrier == “AA”]

filtering on multiple conditions

flights%>%

filter(dest “LAS” | dest “LAX)

(w two equal signs)

notice how you put dest TWICE

new dataframe

new object

what does - do

removes columns

flights%>%

select(-carrier, -dep_delay)

you can combine dplyr functions

by pipe’in, filter, select

rename

(new_name = old_name)

what does mutate do

operates on the existing column

adds new row

(new_var = operation on existing)

in base r….

flights$distance_km. ← flights$distance * 1.6

summarize()

calculates the summaries of variables

like mean, max, min

it can also count the number of rows in a dataframe

group_by()

applies a function to a group of observations; divides data into groups based on variables

what’s the note about groupby function

it doesn’t change the data. subsequent functions will make that mess

what happens after you add summarize() after groupby function

you get a stack of descriptive statistics which is awesome

example

flights %>%

group_by (month) %>%

summarize(n_obs = n()))

steps for working directory

set wd to 380data

load dataset using read.csv

understand the data by using View and head

identify the number of observations and variables

read.csv()

reads csv files

the required argument is the name of the csv file in quotes

types of variables

character and numeric

type of character variables

categorical (2+ text categories)

types of numeric variables

binary - only two

non binary - more than two

binary

only two values, 1s and 0s

presence or absence of a trait

non binary

more than 2 values

character variables

contain text

often categorical; take on a limited number of values

binary numeric variables

numeric encodings tha could be coded as character variables with two categories

(1, 0) or (voted, didn’t vote)

avg/mean of a variable

sum of all values across all observations divided by total number of observations

x bar

avg of x

that weird e to the power of n and subscript 1 = 1 x i

sum of all xi (observations of x) from i = 1 to i = n, 1st observation of x to the last

xi

observation x where i = position of observation and n = total number of observations in the variable

non binary mean of a variable

as an average; same units as the variable (1 or 0)

binary mean of a variable

as a proportion in %; after multiplying by 100

why are binary means a proportion

bc mean of a binary variable = proportion to observations that have that trait

for categorical variables with more than 2 categories

there’s no straightforward interpretation

numeric/double variables

decimals

integer variables

whole numbers

factor variables

categorical

ggplot2 package (what does it do)

basically just the grammar of graphics framework

ggplot2

a statistical graphic; mapping of data variables to aesthetic attributes of gramatical objects

data

datasets w/ variables

geom

geometric objects we wish to plot (points + bars)