I use R for everything from quick and dirty exploratory analysis, to refined visualizations using Shiny.
These programs are both prominent in different areas where data science is employed, and although they are not my go-to tools, Tableau produces unrivaled fluidity in data visualization, and Stata is a standard when interacting with the social sciences.
Python is my main toolset for Machine Learning, including using the wealth of libraries such as Keras and TensorFlow which are available for it.
Node.JS is an invaluable data mining tool by virtue of how it allows native interaction with html elements, as well as a wonderful platform to build out light weight browser based visualizations when nescessary.
A model or an analyis must be grounded in solid theory. I have taken classes including Theory of Statistics, Applied Statistics, and classes deeper into mathematical theory like Measure Theory, and Riemannian Geometry.
Data Science follows the old adage garbage in, garbage out, so a key component is design of expirments and research to collect data. Between coursework like Research Methods for Science and practical instruction from research with the measurement & evualation focused Data4Peace group, research design is a vital aspect of my skillset