What’s a good good goal for knowing that you are an advanced R programmer:
Build shiny app
Read the book advanced R.
Implement a more solide functional approach (more use of the pipe operator, be more purely fonctional)
Read a good book on how to write packages on R.
Learn to write api with rplumber
As i do more stats because i am working with structured data. Doing data science and ML are more about working with non structural data and non parametric analysis universe. I don’t think you will gain deep knowledge in R. What’s your take?
Data-scientist who loves to use #datascienceforgood, especially in ecology, energy and the environment. Bonsai, gardening, bikes and music when I'm not at a keyboard.
Unstructured data is definitely part of ml, but if you look at some of @juliasilge
posts about tidymodels, there is a decent amount of work before the modeling to 'structure' the data at least a little. E.g data cleaning/pre-processing. Check out her post on the cocktails tidy Tuesday data set.
There is something else I would add, which is to 'deploy' R 'into production'. A hurdle that I often see is moving the work out of the users local machine and into the cloud. Writing about doing that really demonstrates an understanding of the language, purely because so few cloud providers make this simple, unlike things like PHP, python, java and JavaScript.
Something that really helped me understand functional approaches was reading the vignettes in the purrrr package, and a lot of the blog posts by Jenny Bryan.
Data-scientist who loves to use #datascienceforgood, especially in ecology, energy and the environment. Bonsai, gardening, bikes and music when I'm not at a keyboard.
What’s a good good goal for knowing that you are an advanced R programmer:
As i do more stats because i am working with structured data. Doing data science and ML are more about working with non structural data and non parametric analysis universe. I don’t think you will gain deep knowledge in R. What’s your take?
Those are all great project ideas 👍
Unstructured data is definitely part of ml, but if you look at some of @juliasilge posts about tidymodels, there is a decent amount of work before the modeling to 'structure' the data at least a little. E.g data cleaning/pre-processing. Check out her post on the cocktails tidy Tuesday data set.
PCA and UMAP with cocktail recipes 🥃🍸🍹
Julia Silge ・ May 27 ・ 6 min read
There is something else I would add, which is to 'deploy' R 'into production'. A hurdle that I often see is moving the work out of the users local machine and into the cloud. Writing about doing that really demonstrates an understanding of the language, purely because so few cloud providers make this simple, unlike things like PHP, python, java and JavaScript.
Something that really helped me understand functional approaches was reading the vignettes in the
purrrrpackage, and a lot of the blog posts by Jenny Bryan.jennybc.github.io/purrr-tutorial/
Thank you, i will the articles and i will take a look at purrrr. You really should be a mentor lol. Take care, i am sending all my love from Haiti.
Likewise from Wales, UK :)