How to upgrade and learn the R programming / analytics language?

Question

data mining

How to "pump" and learn the programming / analytics language R?
What topics to choose, in what order, what to pay attention to?
What is more promising from data analysis in modern analytics systems and what will be promising in the future? Forecasts?
What topics in mathematics to pull up and catch up for more effective use with R?
What resources, websites, youtube channels, video tutorials, webinars are there on this topic?
Of course, first of all, Russian speakers, but English ones will also work.
How to become a big data analysis monster with mathematical skills and analytical karma, R guru?

Answer 1

I download DataCamp for free courses. I started with the Try R course from CodeSchool.com.
Further Data Science Specialization on coursera.org, everything can be done for free.

Answer 2

I don’t know how much of a monster I am of all this, but I’ll tell you how I got something.
First, about five years ago, probably, I tried to run R and do something, but I could not open the file. A year and a half later, a friend from some American university came to visit us at the psychology department of St. Petersburg State University and was going to teach us the statistics used in clinical psychology. However, most of the lesson we learned how to open a file in R, which helped me a lot in the future.
In 2012, I could already do some things in R, but I felt extremely insecure and did everything very, very slowly. After I quit my job at SPbU (after all, blogging "SPbU News" was too risky) I no longer had an academic license for SPSS, and I completely switched to R and started fulfilling commercial orders.
In one of them, I had to quickly prepare about fifteen reports on the same geomarketing research in different parts of the Moscow region. And I hired a colleague who was much more advanced in R than me. On my order, he wrote a script that automatically opened all xls files in a folder, processed them and put the pictures drawn in ggplot into new folders.
In my subsequent work, I used fragments of this code for about a year and a half. Finally, in July 2014, I took a job as a researcher at Wargaming, hoping that regular, daily practice would allow me to significantly strengthen my skills. Still, working constantly is not the same as joining a project for 3-4 days a month. And I did not lose in the sense that I no longer need to look into the function guide to solve ordinary tasks.
I highly recommend that, first of all, not to master some super-complex fashionable data processing methods, but to achieve a complete and clear understanding of the data "shaping" commands. These are all aggregate, cast, melt, rbind.fill, apply, lapply, recode, merge ... Because while this understanding is not there, when working with any other methods, 70-90% of the time is spent on understanding how prepare the data of the type you need.
Alas, the decision to get a job led me to the complete impossibility of improving my skills on the courser, etc., because there is simply no time for this in principle. Is that if you somehow plan and spend your vacation or New Year holidays on this.

Answer 3

https://www.coursera.org/course/rprog

Answer 4

I’ll add that there is also a pack of courses on data analysis on the courser:
https://www.coursera.org/jhu
I started programming in R myself, you can also look at udacity - they are not very bad, there is also a course on pandas - - a sort of mixture of python and R.
As a practice, you can start writing articles on habr on this topic. I try to collect all the data in the same place for public use: https://github.com/SergeyParamonov/HabraData
People on Habré perceive this more or less positively, you can also get feedback there (in the form of comments from knowledgeable comrades)
habrahabr.ru /company/dmlabs/blog/219679
habrahabr.ru/post/236759
habrahabr.ru/company/dmlabs/blog/218607

Answer 5

Practical tasks can be solved on sites like Kaggle (there are analogues in Russian). Theory (as already written) on Coursera.

Answer 6

I recommend using the blog r-analytics.blogspot.ru/.
The guys have recently released a manual on the entire blog material: r-analytics.blogspot.ru/2014/12/r.html

Answer 7

swirlstats.com