Don’t be scaRed of R
Red of R
Wyclife Agumba Oluoch
2022-07-28 16:19:35
| Contacts | : \(\downarrow\) \(\downarrow\) \(\downarrow\) \(\downarrow\) \(\downarrow\) \(\downarrow\) |
| wyclifeoluoch@gmail.com | |
| https://twitter.com/WYCLIFEAGUMBA | |
| RPubs | https://rpubs.com/Wyclife/ |
| GitHub | https://www.github.com/Wycology |
| Code | https://github.com/Wycology/scaRy_R |
Note: Some of the above links you may need to copy and paste in your browser then run.Introduction
This post is reflection of my journey in learning R as a
statistical programming language. I will use R version 4.2.1
(2022-06-23 ucrt). I intend it to be read by seasoned
R users (to remind them of what they should not overlook
when teaching juniors) as well as new R users (to tell them
what should not discourage them) whom have not even downloaded the
program.
Learning R, I think just like any other programming
language (though I have no much experience in those ‘other’ programming
languages), is normally received by learners with zero
background in programming with awe. Why? Because the training normally
comes from seasoned users, who have used the program for a long time
that they know, but overlook, basics that a new user should grasp while
making their initial step into R.
A seasoned R user will start teaching R by
statements like:
Scope
- We cannot teach you everything in
R🥶.
Well, to a new learner, do not think that what you will cover is too little, it will still be enough to get you started. So, let that statement not scare you. It is like a lecturer teaching Island Biogeography and then referring students to other literature to learn more into that concept. So, this is normal. Carry on!
Version
- Make sure you have the latest
version of
RandRStudio⏰.
How do you even start to make sure? Anyway, since you will be
downloading R for the first time, you are likely to
download the latest version. Now, to give you some background, since the
first R programme was created, efforts have been made to
make it better. Each time it is made better, a version is released.
Think of yourself as someone who is learning R with a
better version than what even your trainer used to learn it! With that,
most things are more simplified and more user friendly to you. So, it is
to your advantage to have the latest version of R. Do not
be scared. You will say the same to those whom you will be training one
day.
Package compilation
- You may also need to install
RTools🧰.
I know, this can be too much to expect from newbie. What it does is
that it helps to install packages that need compilation. An analogy to
this is doing a Masters degree programme in Germany. Now, German
students can be admitted without any language course requirements.
However, if you come from non-German speaking country, you may need a
‘compiler’ to train you on speaking German language then you get
enrolled into your course of choice. Now some packages, forget about
what packages are now, may need compilation before they are usable in
R and that is the role of RTools. However, as
a new user, most packages you will need can run without compilation:
They speak same language as R! So, keep going!
Slack channel
- Kindly share your RMarkdown via
our
Slackchannel 🛸.
“Hey! we are here to learn R not slack!”,
you may say. Slack is actually making your work easier. It
is where you can share your progress and get help when stuck. What we
are trying to avoid here is sending e-mail to everyone while asking for
help. Instead, just post it on slack and anyone in the
channel will be able to see and suggest possible help. Even if you
succeed in what is tormenting others, you can also share it there. So,
it is your site for help and sharing. Sharing is caring!
Slack is to your advantage.
GitHub
- All the scripts are available on GitHub repository 🚀.
This is another scary word that can come from a seasoned trainer.
Whatever GitHub means to
someone who has just installed R for the first time or
generally new to programming and version control is obscure. However, as
a trainee, see this as something to your advantage that all running
codes are stored somewhere online where you can get them forever, copy
and paste in R, and use them repeatedly without every time
sending e-mail to the trainer requesting for the scripts. You should
think of GitHub as Google Drive for your
data or documents. In the world of programming it is far much more than
that.
Personal work
- Now you can fetch your own data and rerun the scripts 🌍.
When being introduced to R, the trainer may use data
already available in the program or associated with loaded packages.
However, the data which you will analyse for your own good will be from
somewhere else. Its format and size, generally its structure, will most
likely be completely different. This is the same as research methods
course. We are all trained on same interview protocols, focus group
discussion procedures, randomized complete block designs, but using them
in the field is research specific. For example I will interview 30
people, another person will interview 50 people in 10 cities. So,
consider this as equipping yourself with the right tools and skills to
go and explore your own data. Keep going!
RStudio cloud
- There is also option to use
RStudio Cloud☁️.
Cloud computing, maybe you are hearing it for the first time, is the
bread and butter of data scientists and related fields moving forward.
Just like Microsoft Word in your desktop and Google docs online, there
is also R you can run on desktop and R you can
interact with online via RStudio Cloud. When you spill
coffee on your PC, it may be difficult to recover programmes and data in
it. However, if you were having everything on the so called
cloud, then you have no worry at all, except for the
hardware lost. You can access everything anywhere. On Mac, Linux,
Windows, absolutely any platform provided you have internet coverage.
Think about it.
Error messages
- Follow my instructions otherwise
it will throw an
Error☢️.
This is programming! Every single mistake, forgetting comma,
forgetting }, can lead to an annoying Error. When faced with
errors, do not give up, they are one of the best ways to learn. Read the
error message, sometimes you will get an answer by reading it. However,
as a new user of R, It is not expected that you know how to
interpret all error messages. So ask colleagues on the
slack channel. When asking, do not say “I ran the code and
got an error!” No one can help you with that kind of question. There is
something called reproducible example, you need to show the
code, data (or part of them), and error message. Sometimes, you may need
to state the version of OS you are using as well as that of
R and RStudio and packages among others.
Depending on the nature of the error. It is like going to a lawyer. To
increase chances of the lawyer winning your case, you need to give the
lawyer the complete and true details of the case. So, reveal everything
that can help the community to help you. Otherwise it may not be
possible. Be open, R itself is
Open Source/Access.
Google! Google!
- Google is your best friend, the
Rcommunity is huge 💻.
Yes, Google. Google it, copy the error message and paste it in the
browser and add R. You are not alone facing the error. Many
people, including your trainer, got errors before. So, you most likely
won’t miss an answer in Google. Even when you will be
perfect in R, you will still get errors or
bugs. At this point, I will encourage you to use slack
channel first because Google will throw to you everything and you might
need to be more experienced to separate bones from flesh of sardines.
Within slack channel, we are working on quite closely
related topics and as such you will get help faster and more specific.
Once it is solved on slack, other members will also benefit
from it, should they face the problem with their data too. On a related
note, do not single out one person whom you think knows it all and
bombard them with all your questions. No. Ask on the slack
channel. Others too want to learn. The problem is not personal, neither
should the answer be. Slack! Slack!
Slack!
Terms and names
- read.csv, there is no import.csv? 👽
Words used in R can sometimes mislead a new ‘fresh’
user. Read csv does not mean you will listen to the computer reading out
loud every line of the csv file for you. Actually, as a new user, you
would expect the function to say import csv. However, that is how the
function was named and it just picks the file from where it is located
and puts it within R environment so that you can analyze
it. It will take sometime before you appreciate the choice of words used
in R like library is not a collection of books like where
you read the books but it is a function to load a package. So, do not
take too much time linguistically judging the names like run, click,
enter, of functions and packages. For the start, concentrate on
what they do. Later, you will know how they do
it. Possibly question why they do what they do and you will
learn to make your own functions and packages for yourself
and others. It is possible, others have done it, you can do it.
Calling Functions
- Now we use, group_by() and then summarize() 🖥️
How on earth will you know which function to call and when to call
it? Trainer will be hitting keyboard with names of functions and running
them and producing awesome results! From where? Is there inner
communication between the trainer and R that you are not
aware of? How many functions are there in one package/library alone?
When and how do you know that what you have typed is an object name and
not a function? How risky is it to store an object with same name as a
function in R without you knowing? Get it from me, don’t
worry. If you ever traveled from your hometown to a new city to stay
for, say, more than 3 months then it is the same experience. You get to
know where all supermarkets are, where parks are, where health
facilities are, and where train/bus/taxi/tram stations are. You will not
know the door of a health center while looking for a taxi, will you? It
is always easy to Google like “How to get sum of elements in R vector?”
and you will see solutions like sum(x), where x is the vector containing
elements to be summed. Another insight, there are over 18000 packages on
CRAN alone as of 2022-07-28 16:19:37. You will, most likely, need less
than 20 for regular use. Within those 20, you will only need few
functions from each. It is a marathon, not sprint, with time, you will
get it. Keep the spirit.
Project directory
- Remember to save your work in the working directory or project folder 📁
Just like text editor or spreadsheet, after typing something and you
want to close the work, you should save them. You should also know where
you have saved them so that next time you can access the file easily. In
the case of R, there is no difference, you need to know
where your file is stored and save it whenever you make changes. Having
something called project folder is a way to help you
organize files. Think of a house where bed is near the door, pillow is
on the window, blanket is in the freezer, duvet is in bedroom, and
bedsheet is up on the fan rotating. When you want to sleep you have to
walk around the room to get everything together. However, if everything
is in the bedroom, it would be easier. That is your project folder in
R. Have it and you will thank yourself!
Docker
- You can then run everything in Docker container 🚢.
Just like GitHub above, forget about this for now. When
you will need it, you will learn it. Have you seen those people who lift
the front-wheel of bicycle and ride on hind-wheel only? Is that what you
want to get to on your first day of training to ride a bicycle? I bet
not. Under the version section, I said that R versions keep
improving. With such improvements, some old codes may not run well in
new versions. However, with Docker, you can run the oldest code ever
written in R. That is enough for a beginner. Sorry, Docker
is out of scope for a new R learner. Forget it now! Again,
not knowing Docker does not mean you don’t know R.
Exploration
- From the basics we have given you, it is now upon you to explore more and learn 🗺️.
Just like driving course, you will be trained how to drive at a
certain place. However, once you master the rules of driving, the
trainer will not follow you throughout your life to show you how to
drive everywhere you go. In fact, you can train others. In that same
spirit, once you are introduced into R, keep exploring and
teaching others. Today you are being trained, tomorrow you will train
others. Do not feel left alone. It is the best way to learn
R. Your current trainer was also trained not long ago.
A pro in Excel, SPSS, STATA, and other related graphical user
interface based applications are sometimes hugely discouraged by these
statements. I wish I could know the population of potential
R users who dropped the program the first day because they
were discouraged by one or more of the sentences listed above. They are
so many out there. Do not add to their statistics because now you know
the wealth in the positive sides of the statements! The trainer does not
mean anything bad with those statements. All are to your advantage.
Enjoy learning R.
Conclusion
You are the only person who can stop you.