(1) DATA 180

data science

the process of extracting useful signal from noisy data using statistics, computer science, optimization, causal inference, and domain knowledge

signal

the useful information or pattern you are trying to identify in data

noise

variation or information in data that can obscure the useful signal

data-generating process

the underlying system that creates the data we observe

technical capability

the ability to use tools and methods, read and write code, and interpret outputs correctly

analytical judgment

the ability to recognize when a tool is appropriate, question assumptions, and think through the system being studied

data wrangling

turning raw, inconsistent data into usable information

data visualization

making patterns, comparisons, and uncertainty visible through graphs and other visual displays

machine learning

learning repeatable patterns from examples

R

a programming language widely used for statistical computing, data analysis, visualization, and reproducible research

RStudio

an integrated development environment used for working with R

IDE (integrated development environment)

a workspace that combines tools such as an editor, console, plots, files, help, and debugging

R Markdown

a format that combines writing, code, and results in one reproducible document

code chunk

a section of an R Markdown document where R code is run

assignment

giving a value a name so it can be reused later

object

a named value created in R that can be referenced or transformed later

variable

a named object or characteristic whose value can be used or changed in R

operator

a symbol used to perform an operation, such as +, -, *, /, or ^

vector

an ordered collection of values

collection

a group of values stored together

data type

the kind of value stored in an object or vector

continuous range

a sequence of consecutive integers created using a colon, such as 1:10

index

the position of an element within a vector

indexing

retrieving an element from a vector by its position

loop

a structure that repeatedly performs an operation

iteration

one repetition of a loop

iterable

the values that a loop visits one at a time

accumulation

repeatedly updating an object to build a result over multiple iterations

summation

a running total created by repeatedly adding values

product

a result created by repeatedly multiplying values

function

a reusable process that packages instructions for performing a task

argument

the input given to a function

function name

the name used to call a function

function body

the instructions that tell a function what to do

return value

the result that a function gives back

iteration solution

solving a problem by repeatedly performing steps in a loop

recursion

a process in which a function calls itself on a smaller version of the problem

recursive function

a function that calls itself

base case

the condition that stops a recursive function from making more calls

reproducible research

research in which the code, reasoning, and results can be recreated

print()

displays a value or message

c()

combines individual values into a vector

head()

returns the beginning of an object; in your Cars93 exercise, it is used to show the first six rows

return()

gives a function's result back to the caller

function()

creates a function

for

starts a loop that visits values one at a time

data-generating process

the underlying system that creates what we could observe

population

the whole collection of people, objects, or things connected to a data-generating process

sample

a subset of the population that we actually examine

data

recorded characteristics of the sample

unit

the person, object, place, or event we observe

observation

the recorded values for one unit

variable

a characteristic recorded across multiple units

dataset

the organized collection of units, variables, and their observed values

parameter

a numerical summary of a population

statistic

a numerical summary computed from a sample

measurement issue

a problem where the recorded values do not accurately capture the concept of interest

participation issue

a problem where people who participate differ systematically from people who do not

biased sampling

a selection process that gives some members of the population little or no chance of being included

representative sample

a sample that appropriately reflects the population it came from

statistical inference

using information from a sample to draw conclusions about a population

law of large numbers

the tendency for a sample mean to stabilize as the sample grows and each new observation has less influence

variable type

the classification describing what kind of information a variable contains

categorical variable

a variable whose values represent labels or groups

quantitative variable

a variable whose values represent numerical amounts or measurements

nominal categorical

categorical values that are unordered labels or groups

ordinal categorical

categorical values that have an order, but whose gaps are not measured numerically

discrete numeric

numeric values that are countable, often whole numbers

continuous numeric

numeric measurements that can vary anywhere within a range

matrix

a rectangular arrangement of data with rows and columns

data frame

the R structure used to represent a dataset, with rows representing units and columns representing variables

row

the part of a data frame representing one unit/observation

column

the part of a data frame representing one variable

dimension

the number of rows and columns in a data frame

filter

selecting observations that meet a specified condition

condition

a logical statement that determines whether an observation is selected

logical comparison

a comparison that produces TRUE or FALSE

logical operator

an operator used to combine or evaluate logical conditions

==

the operator used to check whether two values are equal

>

the operator meaning greater than

=

the operator meaning greater than or equal to

&

the operator meaning AND, requiring both conditions to be true

empty result

a result containing no rows because no observations met the filtering condition

data.frame()

creates a data frame from vectors of values

nrow()

returns the number of rows in a data frame

ncol()

returns the number of columns in a data frame

dim()

returns the dimensions of a data frame as its number of rows and columns

names()

returns the names of the variables/columns

$

accesses a named column or component of an R object

table()

counts how many observations occur in each category

prop.table()

converts counts into proportions

distribution

the way values of a variable are arranged or distributed

frequency

the absolute count of observations in a category

relative frequency

the count in a category divided by the total number of observations

center

a description of what value is typical in a distribution

spread

a description of how far apart the values are

skew

the direction in which a distribution is pulled, described by the direction of its longer tail

tail heaviness

how common extreme values are in a distribution

normal distribution

a symmetric distribution used as a useful baseline for comparing distribution shapes

mean

the arithmetic average of the values