Actually, functions from the tidyverse-packages cover just a small proportion of all things you can do in R.

When I teach R to newcomers, I start with some base R basics, like data types (vector, factor, data.frame, list) or indexing (like [, ] or [[]] etc.). I think some basics are necessary, because most help files and search results rely on base R.

But after a short introduction, I quickly switch to teaching the tidyverse-packages (mainly dplyr and tidyr, and my sjmisc-package), because when it comes to "data wrangling" and preparation, the philosophy of "one verb - one function" and consistent function design / api makes it easy for newcomers to remember how to use a package's functions.

To summarize, from my experience, I would recommend learning some base R basics, but then switch to "tidyverse". After that, there are a lot of things, where you need "base R" again (like modelling).

The same applies to functional programming. I would start teaching if/else, for etc. to give newcomers an impression of conditional programming. Then you can switch to the purrr-package. This helps R users understand the "magic" behind the functions.

Ricardo take a look in this post : http://varianceexplained.org/r/teach-tidyverse/

It probably depends on what you want to do. I recommend the following book to people starting:

https://www.amazon.com/Everyone-Advanced-Analytics-Graphics-Addison-Wesley/dp/013454692X/ref=sr_1_1_sspa?ie=UTF8&qid=1533422007&sr=8-1-spons&keywords=jared+lander+r+for+everyone&psc=1

This covers both Base R and Tidyverse

See also the attached. Best, David Booth

Dear Ricardo,

I recommend you to start with base R. Although a lot of tasks (e.g. data manipulation - dplyr, tidying data - tidyr, visualizing - ggplot2) can be done using the tidyverse universe of Hadley Wickham, not all. Tidyverse is easy to use if you are familiar with it, the codes might be more readable than those written in base R, but you will surely encounter problems that you have to solve in base R. Or that are easier to solve, or that are easier to find solution for.

Tidyverse should be the next step, after you gain experience in base R. Then you'll definitaly appreciate the advantages of tidyverse.

This is only an opinion, but hope this helps,

Ákos

It depends upon the foundation of mainly computer science and foundations of mathematics as well as of statistics you have earned.

Based on your foundation, you are to opt.

I suggest you start learning RStudio, it is a complete developing environment in R and much more.

It depends on if you are a programmer who interfaces with biologists (for example) or are a biologist who needs to know a bit of programming to accomplish some task (analyze data, make publishable graphs). If you are a programmer, then learning S is the obvious answer (R is the current form of the S programming language which is what I assume you mean by "base R"). As a biologist, you should learn enough S to be able to effectively use packages like tidyverse. In this case you learn both base R and tidyverse at the same time because to a great extent you do not need to be fluent in all aspects of either. You need enough R to get output that you have to manipulate to get a publishable figure/table, and only the parts of tidyverse that are relevant to changing base R output into the publishable figure/table is needed.

Hi,

Best, Sergio

BTW: I'm not a tidyverse user

This is always a tough question to answer; where to start first? It sounds like you need complete some data analysis tasks right away. Tidyverse will certainly help you there, but you'll undoubtedly run into issues and troubles and base R (as well as other R packages) can help get you out of a jam. Why not learn both at the same time? I wish I could give you a syllabus with a recommended learning sequence, but perhaps the best recommendation is to use this project to guide your learning. It will certainly take longer to complete the project, but you'll know what you need to learn based on the tasks required of the project. And certainly continue to reach out to experts online, like you're doing here, for help when you're stuck! We'll certainly help get you on your way.

There is a popular misconception that one has to have good programming skills to use R. It is certainly true that more skills get more results. However there is a package, called Rcommander,, which is a point and click GUI that works with R just as Analyze(Isn't this the point and click part?) does with SPSS. Given the cost(0) and the almost infinite power, I think one would not want to pass up R. There are many books(as well as courses, YouTube videos,etc) to get you started. I recommend:

https://www.amazon.com/Everyone-Advanced-Analytics-Graphics-Addison-Wesley/dp/013454692X/ref=sr_1_1?ie=UTF8&qid=1533778190&sr=8-1&keywords=lander+jared

the one in this link and the attached. With a $30 investment you can have the whole range of statistical computing and a complete object oriented language). BTW Python has the same cost. See: https://www.google.com/search?client=firefox-b-1&ei=BZtrW7idOIO-tQXqqq-ICw&q=python+for+beginners+pdf&oq=python+fo&gs_l=psy-ab.1.2.35i39j0j0i67j0i20i264j0i67j0l4j0i67.42860.45691..49475...0.0..0.133.1081.7j4......0....1..gws-wiz.......0i71.9e9f1fp98dg

You just can't beat these. Definitely

Can't beat 'em.Happy Computing, David Booth

I recommand R first.

And from the answers i saw RStudio for the first time, seems interesting

Actually, functions from the tidyverse-packages cover just a small proportion of all things you can do in R.

When I teach R to newcomers, I start with some base R basics, like data types (vector, factor, data.frame, list) or indexing (like [, ] or [[]] etc.). I think some basics are necessary, because most help files and search results rely on base R.

But after a short introduction, I quickly switch to teaching the tidyverse-packages (mainly dplyr and tidyr, and my sjmisc-package), because when it comes to "data wrangling" and preparation, the philosophy of "one verb - one function" and consistent function design / api makes it easy for newcomers to remember how to use a package's functions.

To summarize, from my experience, I would recommend learning some base R basics, but then switch to "tidyverse". After that, there are a lot of things, where you need "base R" again (like modelling).

The same applies to functional programming. I would start teaching if/else, for etc. to give newcomers an impression of conditional programming. Then you can switch to the purrr-package. This helps R users understand the "magic" behind the functions.

https://www.youtube.com/watch?time_continue=6&v=dT5A0sAWc2I

R is more fundamental and functional: data analysis, plotting graphs, various charts, 3D, 2D, animations, statistical analysis, create networks, etc. thereare a lot of packages and libraries. R is not that much difficult to master it if learning step-by-step. You can take a look at my codes as tutorials and re-use them in your work applied to your need and specific tasks. The scripts are uploaded in my RG profile, and on my github:

https://github.com/paulinelemenkova?tab=repositories

https://www.researchgate.net/profile/Polina_Lemenkova/publications

Do feel free to re-use them.

To give my two cents.

I started self teaching a few months ago and picked up things in a rather haphazard manner. So would be great if people could learn from my mistakes.

I would say that as a start learning base R is a good place to start ESPECIALLY if the person you are teaching has no programming/coding background such as I did. Being confident using logical operators and if else statements is also key component in the fundamentals.

Learning such basics are quite important as I think it begins to drill home the philosophy of coding in R... and general?.

Alongside this very important to teach data structures(vectors, matrices, data frames,arrays) and make sure that the 'student' has gotten this down to a T especially understanding different data types.

Then important to be able to do basic manipulations of these data types before moving into Tidyverse. (gotta love that %>% operator)

While I'm new to this it would seem that ggplot2 for a start and plotly and others later supersede the need for learn base R graphics.

I first touched R about 5 years ago and have been using it almost exclusively for statistical analysis & data science tasks for most of the past 3 years. Coming from a SAS & SQL background, the syntax looked complex and was intimidating at first. I had some free time open up and really dove into it last year with a few Udemy tutorials and some more complex projects, with the goal of sharpening my skills, and have never looked back.

My recommendation is to learn some very basic, base R first, just to get a feel for what the language is capable of out-of-the box - which is really quite impressive. Tidyverse is a great next step and provides a very well-maintained API for common data science tasks (management, analysis, data visualization, and reporting). There's also something called pipes (%>%) that can be used in base R to "nest" functions, and these are also compatible with functions available in the Tidyverse framework. These are very common in combination with functions the 'dplyr' package (where I first encountered them in numerous posts) and can greatly enhance readability when multiple steps are required.

Feel free to follow me at @biolabanalytics if you want to hear more about what the #rstats community on the Twitterverse is saying about this programming language. I have honestly been surprised (and impressed) by how inclusive and informative this online community has been, and I often Re-Tweet things from more seasoned experts.

Happy coding and welcome to R !

I start using the R program and this discussion is quite useful to me. Thanks all

It mainly depends upon the expertization in basic knowledge, understanding & application in

(1) computer science (language, programmings etc.),

(2) mathematics,

(3) Statistics & statistical methods

and (4) Computational Methods.

Dear Jagathnath Krishna K M ,

It may be so in your case but may not be so in the case of others.

It is due to the fact that the line of thinking of every people may not be the same.

I agree with Daniel Lüdecke tidyverse has just a small portions of things that you can do using R. So it's nice to start learn R than move to tidyverse

The questioner seems to be suggesting both should learned, and is just wondering the sequence.

Matloff (editor of the R journal, author of a great book on R) wrote in his blog yesterday comparing R and Python ( https://github.com/matloff/R-vs.-Python-for-Data-Science ) how having the two systems is very problematic, both trying to use both and having a for profit company pushing something very different than the ethos of R being community lead.

Here is some of what he says:

By contrast, R is rapidly devolving into two mutually unintelligible dialects, ordinary R and the Tidyverse. Sadly, this is a conscious effort by a commercial entity that has come to dominate the R world, RStudio. I know and admire the people at RStudio, but a commercial entity should not have such undue influence on an open-source project.

It might be more acceptable if the Tidyverse were superior to ordinary R, but in my opinion it is not. It makes things more difficult for beginners. E.g. the Tidyverse has so many functions, some complex, that must be learned to do what are very simple operations in base R. Pipes, apparently meant to help beginners learn R, actually make it more difficult, I believe. And the Tidyverse is of questionable value for advanced users.

I agree with Deniel Wright that both are to be learned.

It does not matter which is to be started first.

For some learners it may be convenient to start with R while for other learners it may not be so. It is due to the fact that the line of thinking of every people may not be the same and that every people do not have the same background of knowledge & understanding..

Dhritikesh Chakrabarty , the reference to Matloff (and my belief too, but I am not as knowledgeable as him, so trust him more!) is NOT saying that. He is saying that it is problematic that the two do not always work well together. I have seen Hadley speak a few times and he uses his stuff for almost everything and the discussion in the audiences tends to be either use his stuff for lots or not. Reading between the lines of Matloff's blog and listening to chat among S/R users, I think people think the two will separate more, and one from a company that needs to make money out of its users and therefore with a different philosophy, and trying to mix the two political ideologies can be difficult. Similar worries exist with the microsoft version ( https://mran.microsoft.com/open ). I think people wonder about the business plans for these.

The opinion of Matloff may be true for some learners but may not be for any learner.

I learned some base R first and the tidyverse. It all depends on what your use cases are in the end, but you can get on with quite little base R in the beginning and the switch to tidyverse. As long as you know what assignment, functions, variables, arguments, basic logic operators, difference between list/vector and matrix/dataframe is, you can probably move on.