Exploratory Data Analysis

Lesson 2: R Basics - Notes

Quick links

The Power of R
Intro to Lesson 2
Analyzing Tweets in Chicago
Quiz
Answers
Why R?
About R
ggplot2
Install RStudio on Windows
R Programming Language Installation
RStudio Installation
Install RStudio on a Mac
RStudio Layout
Quiz
Answer
Demystifying R
Answer
Getting Help
Read and Subset Data
R Markdown Documents
Answer
Factor Variables
Ordered Factors
Quiz
Answer
Setting Levels of Ordered Factors
Quiz
Answer
Data Munging
Advice for Data Scientists
Congratulations

Copyright © 2014 Udacity, Inc. All Rights Reserved.

The Power of R
Intro to Lesson 2
Welcome to lesson two. In this lesson, we'll learn about the
advantages of using R to perform data analyses. We'll install R Studio.
And we'll learn the basic data types and commands of R.

Analyzing Tweets in Chicago

Before we get started, I want to share with you how one data scientist used R to
improve the health outcomes for the citizens
Data of Chicago, Illinois.
scientist Corey Nissen used the R
programming language to analyze tweets
from people in the city of Chicago. Corey
analyzed tweets in real time by searching
for the words food poisoning in each of
them. I want you to read more about his
analysis and his work by clicking on the
link in the instructor notes. Try to find the
answers to these next three questions as
you read through the article.

Quiz
What was the name of the R package used by Corey? How many Tweets per day did
the system flag? (lower and upper limits) What did you find most interesting about
the article?

Answers
Textcat was the name of the R package that Corey used. We'll discuss R packages in
the next video and what they allow us to do. For the second question, the automated
system captures between ten tweets to 20 tweets a day. For each of the tweets, the
system will recommend that whoever sent the tweet file a report. Now, for this last
question, we accepted any text answer here.

I thought that the most interesting part was that Cory showed the open source R
code classifier on GitHub. GitHub is a repository for sharing code. You can learn more
about GitHub by following the links in the instructor notes. It's the most popular
repository for open source projects and it's really easy to get started.

Why R?
About R
Copyright © 2014 Udacity, Inc. All Rights Reserved.
R is the leading programming language for statistics and data analysis. One of the
main advantages of using R is that we can build up an analysis line by line in code.
We can save all of our work in a file and go back to see what we investigated at a later
date. Having R scripts allows you to easily share your work with others. And you can
see what others are doing with data. R also has over 2,000 user contributed packages
that increase its functionality. One example of this is the text analysis package,
TextCat, that you just saw.

ggplot2
Now, one of the main packages we'll use throughout this course is ggplot2. ggplot2
is a graphics package that lets us create plots and graphs with just a few commands.
We'll learn more about ggplot in the next lesson but just to give you a taste of
it, here's an example of a plot you can create. Don't worry about memorizing or
understanding all of this code. You'll have plenty of practice with this later in the
course. I just want to show you how a few lines of code, can create amazing graphics.

I'm going to load up the ggplot library and a color library. Then I'm going to load the
diamonds data set, and with this function I'll create a scatter plot. Let's check out

this plot in detail. This part shows the relationship between price and carat of almost
54,000 round cut diamonds. I'd say R is doing very well for such few lines of code. The
last thing I want to mention is that you can use R anywhere. It's free open source
software that works on any operating system. And as a result of this, R has a large
active and growing community of users.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Install RStudio on Windows

R Programming Language Installation

If you’re using windows, I'll show you how to install R and R Studio. To get started
you want to go to cran.rstudio.com. And then you'll click on the link to download R
for Windows. If you never installed the R programming language before, then click
on this first link. Since it's your first time downloading it. Now depending on your
version of Windows, you might need to see the frequently asked questions. I'm set for
my computer, so I'm going to go ahead and download R for Windows. I get a warning
message and I want to keep the file. Now this might take some time to download, so
just be patient. Then you want to open the file, and then run the program. A set up
wizard should appear and now you just want to walk through all the steps. I'll set my
language, and then I'll just continue installing by hitting next. I'm just going to accept

the defaults here and then install it to R. I'll leave these as is, so that way a desktop

Copyright © 2014 Udacity, Inc. All Rights Reserved.

icon will be created. And it's going along. Great. Now I'm finished with my setup.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

RStudio Installation
So that was just part one of the installation process. That just gave us the

programming language R. Now we need to go and install RStudio. RStudio is a

graphical user interface for programming in R. It's basically a point and click system
that lets us work with data really easily. So first let's just go to download now to get
RStudio. We're going to get RStudio for our desktop. So just click this first button.
Now there's going to be a lot of links here, and you really just want this first link. It's
going to be the version of RStudio that's recommended for your computer, and for

Copyright © 2014 Udacity, Inc. All Rights Reserved.

your operating system. Notice how this one's for Windows XP, Vista, Windows 7, or
Windows. Now that it's ready, so I'm going to open this up, and I get the wizard you
can see on the right. I'll follow the instructions, then install RStudio into this path.
And create a Start Menu folder as well. Great. I'm done. And this is what RStudio
looks like. Now yours might look slightly different than mine and that might be
because the file's open. So if I go to File > New File > RScript this opens a basic RScript

Copyright © 2014 Udacity, Inc. All Rights Reserved.

which allows me to start programming in R.

Install RStudio on a Mac

Copyright © 2014 Udacity, Inc. All Rights Reserved.

In this video, I'll show you how to install R Studio on a Mac, which is a graphical

interface for programming with the R

language. Simply open up any web
browser and go to RStudio.com then you
want to go to download now. You are
going to run R Studio from your desktop
so click on this link. Now here's where I
want you to be very careful. R studio requires R first, so if you don't already have the
programming language R installed, you need to download it here. Now this is a really

easy link to miss so make sure you click

here first to install R, before installing R studio. Once you're at cran.rstudio.com, you
want to select the download for the Mac OSX version. This will bring up another page
and then you can download the latest version of R. It may take some time to
download so you might want to take a break and then come back to the computer.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Once you've opened up the package, follow the instructions for installing R. You
should end up getting a successful message for the installation. Now you're ready to
go and install R studio. This website will recommend the best version of RStudio for
you so just click on whatever one comes up for you first. Once you've downloaded
this, you want to open that package. Now, just drag RStudio into your applications

and you'll be good to go.

RStudio Layout

Copyright © 2014 Udacity, Inc. All Rights Reserved.

This is R studio, it's an IDE or an Integrated
Development Environment and it's what we'll be
using to create R scripts and to create plots. It's a

point and click system that organizes

our workspace and it has some nice
features for creating and sharing work.
Here's a basic tour of it. The main
interface contains four panels that can
be resized as needed. Now, you may
only see three panels when you open R
studio. If so, just open a new R script
by clicking on the file, and then New
File > Rscript.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

So, in the top left here, we have Rscripts. These are the files in which we will type our
R commands. We'll also save these files, so we can alter them or share them at a
later time. In the bottom left, we have the R Console. You can type in R commands
here. However, these R commands will only be saved in a temporary history. The
commands won't be saved to a file like up in here that can be reviewed at a later
date.

In the
top
right of
R
studio,
we
have
two
tabs. One called Environment and one called History.
The Environment will contain all the objects,
functions and values that are in the current working
memory for R. History on the other hand will keep a
running log of any of the R commands that we run.
Now we can run R scripts directly from our files but
we can also type them into the console. No matter
which way we type them in the History will capture them.

And finally, in the lower right, we have the Files, Plots, Packages, Help and Viewer
section. You'll learn more about these later but this is where our plots or graphs will
first appear when we run an R command that creates a visualization.

Quiz
Now that we've gone over the layout of R Studio, I want you to match some actions
that you might do in each panel of R. Remember this first panel is for R files or
Rscripts. This second area is for the workspace or the history. This third space is

Copyright © 2014 Udacity, Inc. All Rights Reserved.

the Console where we get our output. And this fourth panel is for Files, Packages,
Plots, and Help. So for example, if I asked you to match the activity to one of these
panels, like save an R script that would correspond to box number one. You can save
them by clicking on this button. Now we could of course use the main menu, but I'm
actually include as any one of your options here. The main purpose of this quiz is just
to get you familiar with R studio. So here are your list of actions. And I want you to
pair it with the panel where you would do it in R studio. Put the number of the
corresponding panel in each of the boxes.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Answer
You can review the log of command entered by going to the history so that would be
panel two. Help documentation appears in the lower right hand side which would be
panel four. To clear the workspace you can either use panel two or panel three. You
can run a command in here to clear the workspace or you can use the icon located in
panel two. It should look like a sweeping broom. We can run a bunch of commands
from a file from panel one. Plots appear down here in the lower right, so that would
be panel four, and all of our objects and memory appear up here, in our environment
or our work space. So that's panel two. And finally to read the results from functions
or from calculations, we simply look at the console, where our output comes, panel
three.

Demystifying R
Now that you know the basic layout of our studio, you're
welcome to customize it, by going to Tools, and then
Options. From here you can choose the default directory if
you'd like. You can change formatting options for editing
code, or you can even change the font and the appearance.
Here, I'm using the tomorrow theme, but there are of
course some others. You can also make adjustments to the
panels so, feel free to play around with this to get a
configuration that you like. Now, you might not have any
idea what you like first so just wait until you play with r a
bit, and then you'll find a configuration that suits you. I'm going to leave this as is.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Alright, we've got R set up, so let's start using the programming language. To get
started, I want you to download the R script (demystifying.R) and open it here. When
you open the file, it should look like this. This R script contains both comments as
plain text and R code. You can run R commands by highlighting a chunk of the code
and then hitting Cmd+Enter on a Mac, Or Ctrl+Enter on a PC. The lines of text that
start with a hash or a pound symbol are comments. There lines are plain text that we
don't want to run as R code. It's the standalone text that you really want for the R
code.

Now, your task is to read through this file and run and write code when you're
prompted to do so. When you get to the end of the file, you'll get to a question, and
you'll need to answer this question to move on to the next video in this lesson. Now,
if you're already familiar with R, I suggest trying to answer the questions, and then
you might want to go back to the files to see if you missed anything.

We know you learn best by doing, so it's important that you execute and run this
code as you go, and that you make sense of the output. Any output will appear
down here in the console. I also hope that you have moments of inspiration. So if
something does come to mind I want you to code it up and just run it in R. You won't
break anything and the worse thing that would happen would be an error message or
a warning message in this box. If you do happen to get stuck at any point post in the
discussions so that way your peers or one of the instructors can help.

Quiz
Download the demystifying.R file and open it in R Studio. As you read through it,
read and write code when prompted. When you get to the end of the file answer this
question: what is the average mpg (miles per gallon) for all of the cars in the mtcars
dataset?

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Answer
This question should have been pretty simple. You just needed to run this bit of code,
“mean(mtcars$mpg)”, and that would have given you the value to enter.

Getting Help
As you go through this course, you might encounter commands that you've never
seen before. And we don't expect you to know everything, but we do want you to
keep the following in mind when you encounter any sort of challenges or difficulties.
First, take an active role in problem solving and be aware of the resources at your
disposal. You can type a question mark and the name of any function to bring up help
documentation in R.

So for example I can look up the

documentation for the function
mean. Here it will tell me any
parameters that mean takes. And it
will give me examples usually at
the bottom of the documentation. I
can copy and paste these into the
console. And press enter to see
what the function does. Here, I
created something called x, and
then if I hit x into the console, I can
see what this is. This is just the vector of numbers zero to ten, and then the number
50. I can take the mean of x, and print it out. And finally, I have some sort of trend

Copyright © 2014 Udacity, Inc. All Rights Reserved.

parameter. Maybe you can figure
out what this does.

Now, if you can't solve your

problem using the documentation I
would suggest Googling it.
StackOverflow is also another great
site to browser for help and for an
FAQ. One of my personal favorites
when I learned R was Quick-R, or
statmethods.net. This site has
excellent examples and short pieces of code that can refresh your memory and get
you
andgoing on any data analysis. I've linked to a couple other websites like the R Cookbook
R-bloggers, and you can also find these in the course materials under R Resources on
the course wiki.

Read and Subset Data

In the next three lessons, we will explore a data set that resembles user activity on
Facebook but before we get to that data set I need to show you how to read in data
that doesn't come with R and how to subset that data. It turns out that not all data
sets can be easily loaded like the mtcars data set. You can find a link to a new data
set here. This data set is going to contain information on the fifty states in the US.
Most of the data comes from about 1970.

Before we can read in the data, we need to set our current working directory. So,
to figure out what directory you're in now, you can type in getwd(). We can run
Copyright © 2014 Udacity, Inc. All Rights Reserved.
this command and see the output to the console. It looks like I'm already in the
downloads file, and that also appears here in the top of my console. So I don't need
to change my directory, but maybe you do. To change your directory, you can type
setwd(‘directory’). This will take a string which will be the file path to whatever
directory you want to go to. My guess is that your data set is in your downloads file.
So I would probably run this command (setwd(‘~/Downloads’)). Now it's important
to note that whether or not you are on a Mac or Windows machine you still need to
separate your paths or your folders with a forward slash. Also be sure that you use
quotes around your path.

Now in order to load up the data, we can use the read.csv(‘filename’) command.
This command takes a string, which is the name of the file. And here we're going to
pass it to a variable called statesInfo. statesInfo is going to save all of our data into a
data frame. When I run this code, I can see that states info appears in my
environment. I can double click on the data frame in the workspace, and this will let
me see the table of values in R Studio.

Now, let's say I wanted to get information on states located in only the Northeast.
Those states would be states like Connecticut and they have a state region of one.
I'm going to go back to my R-script and write a command that pulls in this data. This
subset command would look like this. Here I'm passing the data frame states info to
subset and I'm asking for it to retrieve any states that have a state.region equal to 1.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

When I run this code I
can see the output
down below. We have
Connecticut, Maine,
and many others.
Now, there is another
way to subset this
data frame. I didn't
want to show you it,
just so that way, you don't get confused later if you
see it. It uses bracket notation, where we have the
name of our dataSet, followed by two brackets.
And we'll have a comma in between. This first spot is for the rows of our data set that
we want to keep. And the second spot is for the columns that we want to keep.

So if I want only the states in the Northeast, I would write this code. The name of the
data set is states info and then I want the rows that have a state region equal to one.
Now I can't just use state region here, I need to access the actual variable, so I have
to put states info and the dollar sign. This gives me the actual variable value and I can
see if it's equal to one. If it is equal to one, I want to return every single column in the

data frame. So for example, with Connecticut if it's state region is equal to one. I want
to return every single column in this row.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

To return all of the columns, I'll just leave this blank. So this code searches for rows
that have a state region equal to one. And then it takes all the columns in that row.
And all of this will be sent to the console as a new data frame. Now it might not be
so helpful to have this output just in the console. So we can save these subsets into
new variables. I'm also going to include some functions to print out the first two rows
of each data frame, and also their dimensions. Hopefully, I've convinced you that
they're the same data set.

Now, I really want you to pay careful attention to the syntax in both of these
examples. Throughout this course, we tend to make use of the subset command, but
there might be instances where we use the other method. Just know that both
methods produce the same result. Now, I recommend that you try subsetting this
data frame for other regions of the country on your own. You could also try finding
out which states have an illiteracy rate of 0.5%, or which states have high school
graduation rates above 50%. Feel free to play around.

R Markdown Documents
Let's get some more practice working with data frames. You downloaded an R script
earlier and saw how we can save our work and run an R code from it. For your next
task, you're going to download an RMD file and run code in it. The file will look
something like this. Notice how this file is slightly different from the file that you saw
before. An R script can only have R code and comments. This file, however, the RMD
file, allows us to do so much more. It's an R Markdown document, or RMD.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Now, you might be wondering what Markdown is. And basically, it's a simple
formatting language that lets us author web pages. If you want to know all you can
do with Markdown then go to Help and then go to Markdown Quick Reference. This
would bring up the documentation for Markdown, and you can scroll through some
of the examples in the formatting. We've also included a video that will go through
Markdown.

So far this file only contains text that will be formatted using Markdown. Let's add
some R code to this. We can do this by clicking on Chunks and then going to Insert
Chunk. Now if you're friendly with the keyboard, you can use the shortcut
Cmd+Option+I. There's many other shortcuts in here and you can see them here in
this menu. Here I've added some code to see what this data set is. This is the cars
data set that also comes with R, and it contains 50 observations of speeding and

Copyright © 2014 Udacity, Inc. All Rights Reserved.

stopping distances for cars in the 1920s. I'm going to add this text as the title to this
document, so that way it's more descriptive of my file. I'll also run the str command
on this data frame, so that way I can see what's inside of it. It looks like I have two
variables, speed and distance, which are measured in miles per hour and feet. Not
only can we run R code and get the output from these files, we can also embed
images. This plot function comes with a base graphics package in R. But don't worry
learning more about it. We'll be using the ggplot2 package which is a powerful
graphics language in the coming lessons.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

When I run this code - plot(cars) - I
get a simple plot of speed versus
distance. Now, what's really amazing
about this document is that we can
click the Knit HTML button. This will
generate a web page that includes
both content, as well as the output of
any embedded R code. First, you’ll
need to install and load the package
knitr in order to use the Knit HTML
button. Run the commands to the
right in RStudio console install and
load knitr. This simple button allows us to easily share and publish our work. That's
enough about RMD files. Let's have you work in one now. Download the second RMD file
if you haven't already, and run and write code when you're prompted to do so. At the
end of the file, you'll come across a question. I want you to answer that question for
this next quiz. And then you can
continue on with the rest of the lesson.
Good luck and have fun.

Answer
The cars that satisfy either this
condition or this condition were the
Fiat 128, the Honda Civic, the Lotus
Europa, and the Toyota Corolla. Let's
see how we can figure this out
using "R". Your task was to determine
which cars in the mtcars data set have
MPG greater than or equal to 30 and
an HP less than 60. To answer this
question you needed to subset the
data frame. I'll show you two ways of
doing this. In the first method, I'll use
the subset command on empty cars,
and I want to get the cars which have
an mpg greater than or equal to 30 or
whose horse power is less than 60. That would be the Fiat, the Honda Civic, the
Toyota Corolla, and the Lotus Europa. Using the bracket notation, the coats and
tacks would look like this, and there's the same output.

Factor Variables
Copyright © 2014 Udacity, Inc. All Rights Reserved.
Now that you're familiar with the basic R commands, let's look at some more data.
This time we'll be looking at data collected from a survey of Reddit users. Reddit's
a social and entertainment website where users can post links and comments
about trending news. This survey asks users about demographic information such
as gender, age, nationality and employment status. It even asks users what type of
cheese they would be. And whether they prefer dogs, cats, or turtles. Download the
data set from the instructor notes and load it into R. Once you've done that, take a
look at the data by using the str(data) function.

Now when I try to read in the file, sometimes I might get an error, and this is pretty
common, so I would suggest looking at your current working directory to figure out
the problem. Often times your directory isn't where your file is stored. Alright, so I've
set my directory, and now I'm going to try this code up here again. And there we go,

there's our data. Running the str(reddit) command, we can see that there's lots of
data here. Most of these variables have a type of factor.

Now, a factor is a categorical variable that has different flavors or levels to it. An
example of this would be employment status. This variable has many different levels
such as employed full time or employed part time or not working. One thing we
might be interested in is how many people are in each group of employment status.
We can table that variable to see the number in each of these groups. Running this
code - table(reddit$employment.status) - I can see the table.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

We can also get these counts and other data points by running the summary(reddit)
function on our data frame. I encourage you to check out the output of this for
yourself. In addition to factor variables like employment status, the R programming
language also has other data types like list and matrices, but we really won't be
working with them in this course. We've included a reference for data types, so if you
would like to learn more about those, check it out.

Ordered Factors
Let's look more closely to these factor variables. For now I want to draw your

attention to the age.range variable right here. Notice that it says that we have a
factor variable with seven different levels. We can examine the levels of a variable, by

Copyright © 2014 Udacity, Inc. All Rights Reserved.

typing in the command levels(factor variable) and then putting in the variable to the
argument. In the console we can see the seven levels of the age.range variable. Now,

instead of creating a table of the age.range variable, let's create a plot that shows
how many users are in each bin. That is, we want to figure out how many surveyed
respondents are between the ages of 18 and 24, 25 and 34, and so on. I’m gonna
create this plot using the ggplot2 package, and the qplot function that comes with it.

Again, don't worry about understanding this code too much, we'll have practice with
Copyright © 2014 Udacity, Inc. All Rights Reserved.
this in the next lesson. When I run this code, I get my plot over here. I want you to
notice that the age groups appear to be in order. This is true for everyone except the
survey takers who are under the age of 18. Now, it would be really helpful if the under
18 bar was really on the left of the 18-24 bar. That way we could make comparisons
across the groups more easily. Now this is why we would want to have ordered
factors.

The variable age.range just contains factors with seven levels, but these levels aren't
arranged in any particular order. Sometimes you want to introduce order into our
data set. So that way we can make more readable plots. So, knowing a little bit about
ordered factors, let's see if you can answer this next question.

Quiz
If you haven't already done so, download the Reddit survey data and look at its
structure. After you looked at the structure of the variables, try and answer this
question. Which of these variables in the data set could also be converted to an
ordered factor? Just like age.range. Check any of the variables that apply.

Answer

Copyright © 2014 Udacity, Inc. All Rights Reserved.

The answer here is just income level. Gender would not be considered an ordered
factor, since it doesn't make sense to rank male and female groups. The military
service and children variables are binary variables with yes or no values, so these
really wouldn't make sense to order either. Using the same thinking, it doesn't really
make sense to order the levels of country and state variables, since we wouldn't
rank those either. We wouldn't really say one location is better than another. And
finally, cheese falls into the same bucket as the country and state variables since
they wouldn’t necessarily be ranked either. The income levels actually have a natural
order to them where we can group the buckets from lowest to highest, or from
highest to lowest. That's what make it an ordered factor.

Setting Levels of Ordered Factors

So here's our problem. If we try to see the amount of users in each age group, it's not
in order and it's hard to compare immediately. This problem is even more noticeable
on a different plot. If we try to create the same plot for users in each income bracket,
we see a very similar problem. And here it's much worse. These first two bins are for a
100,000 and over 150,000. And this last bin is for people who make under 20,000. Our
eyes tend to read pages from left to right , so this graph is pretty hard to interpret. Or
even better said, it'd be very hard to make comparisons naturally with our eye. We
have to always scan down to the bottom to figure out what group we're in. So let's
order the factors in our age.range variable.

Quiz
What I want you to do is to look up the documentation for the factor function. Or you
can read through the example here. Once you're ready, try to write the code in order
Copyright © 2014 Udacity, Inc. All Rights Reserved.
to order the levels of the age.range variable. And just as a reminder, the level should
take on the following values, 18-24, 25-34, 35-44, 45-54, 55-64, 65 and over, and under
18.

Answer

Copyright © 2014 Udacity, Inc. All Rights Reserved.

For this programming assignment, you needed to have rearranged the levels of the
age.range variable, so that way, under 18 appeared first. One way to do this is to use
the ordered function on the variable. We'll use the ordered function on age.range,
and then we'll set the levels. I can run all this code to set the levels, and then, I can
make my plot again. Pretty nice, huh? Another way to achieve the same result would
be to use this code. Here, we're using
the factor function. We're taking our
age.range variable, setting the levels,
and then making the order be true
since we want these ordered. I'll run
this code. And then, I'll make sure
that the graph is the same output.
And yes, it is. Now, if you want
another challenge, try reordering the
levels for the income range variable,
and see if you can produce this same
plot.

Data Munging

It's important to note that many of the data sets that we've used so far in this course
are what I would call tidy data sets. What I mean is that these data sets were

Copyright © 2014 Udacity, Inc. All Rights Reserved.

manipulated into a specified format of rows and columns. You can learn even more
about tidy data here. This is what allowed us to so easily import these data sets into
R. You should know that not all data sets are going to be so nice. Sometimes, you
may be pulling data from different sources, like webpages, audio files, or even PDFs.
Other times, you may need to reshape or rearrange your data into different formats.
We're going to cover this later in lesson 5.

The important thing to know is that this is a necessary step prior to conducting EDA
and it's called data munging. There are plenty of tools for doing this kind of work,
and if you're interested in learning more about how to wrangle and adjust data,
check out our data munging course. Now, this course EDA won't cover data
munching, but techniques for doing so are vital for any data scientist.

Advice for Data Scientists

Before we wrap up this lesson, let's hear from Eytan Bakshy and
Sean Taylor, who are data scientists at Facebook. They'll offer
their advice for how to work with data.

Eytan: My advice to data scientists, or future data scientists is

find data sets that you're interested in, and,
and work with them, and play with the data.
And just developed experience with playing with data.

Copyright © 2014 Udacity, Inc. All Rights Reserved.

Sean: Good data science comes from good questions, not from
fancy techniques, or from, you know, having the right data. It
comes from motivating your research with an idea that you care
about, and that you think other people will care about. It's almost,
you almost have to think about it like a journalist would. You have
this audience and they're going to consume the work that you
produce, and you want them to care about it, and be interested. And so, starting with
the right questions, in particular, something that motivates you as a person is really
the right starting point, not, like, you know, some fancy technique that you want to
learn.

Congratulations
Congratulations on finishing lesson two. In this lesson you
learned how to use R Studio and you learned about the basic
commands in R. If you found something particularly helpful or if
you have ways that we can improve the course, let us know by
posting in the forum. In the next lesson, you'll learn how to visualize and summarize
single variables within a data set. We hope to see you there.

Copyright © 2014 Udacity, Inc. All Rights Reserved.