Based on your experience, would you mind to advising me to choose between the learning R or Matlab as required software to analyze the datasets of my studies in Banking and finance.
If you want to be a good researcher in the field of Accounting and Finance, you should better learn Eviews & R software as well as Python programing. Matlab is more useful for engineering disciplines and I do not recommend it for you. Eviews and R are best choice for you.
In my opinion R (or alternatively Python) is more robust for Finance disciplines because just like the foundations of R and Python, finance has a statistical foundation. In addition, R offers packages that can be interactively used with Python as long as you have a good interface such as R-Studio or Anaconda Navigator. I also find R easily Customizable since you can share coding ideas with peers and experts via platforms such as Github. You can also call some Matlab functions from R or Python.
Unlike R and Python which are mainly founded on statistical computing and graphics, Matlab is founded on technical and mathematical computing thus more suitable for Engineers and Physical scientists.
R is a preferred software due to its wide range of application software that is freely available. The other software Stata, Eviews and Matlab are fee paying licensed software.
1) Matlab is an academic tool. It requires a license, and is not widely used by in quant research outside of academia. That being said, a large fraction of academic research papers in finance use Matlab (including many of my own), but this number is shrinking.
2) While R is great for the statistics community, it is quickly waning in popularity in Finance, and is being actively phased out by many financial/ML research groups, including mine (the code is hard to maintain and debug, and encourages many bad coding practices that make life miserable later on). It has some great functionality, e.g. its data frame functionality/intuitive interface destroys pandas in python, and is typically faster, plotting is easier, etc.
3) Stata is used almost exclusively by econometricians and statisticians (it is the first "language" I learned, and the first one I ditched, for "R"). Its use beyond standard statistical analysis is rare, and its usefulness for finance in general is very limited.
4) Python is the horse I bet on for RnD in Finance. Its functionality out of the box is hard to match. It's paradigm is OOP (yet supports scripting, if you must), it is expressive and understandable, reasonably maintainable for a dynamically typed language, and has a massive community at its back. It's plotting functionality is tolerable, as is it's performance (for RnD). It is rapidly becoming the standard for Machine Learning research, mostly due to sklearn/pytorch/keras/tensorflow and its ability to support data-driven application development (ie 99% of Machine Learning). It is weaker than R in statistics, but unless you are on the bleeding edge of stats research, it's unlikely you will need much more than it already offers. Python has its flaws, but is a pretty good choice overall. Not to mention PyCharm is one of the best IDEs ever created (and it's free).
I think that the R language is more suitable for research in finance, especially in the area of option pricing. In addition to being free, R software also has several packages to simulate the stochastic processes involved in modelling the price of underlying assets, these simulations give an idea of the future behaviour of the financial market, which is very important for any investor. Currently, there are also R packages that incorporate parallel computing to solve difficult valuation problems such as American options or cumulative parisian options. For my part, R software has always accompanied my work in finance and stochastic processes.
Having used Matlab, Python & R, my preference for research is clearly R.
RStudio is amazing & lots of resources are compiled in the online book:
https://www.bigbookofr.com
The only issue is that for implementation, CS people prefer Python (more object orientated), so it's sometimes an issue between taste compatibility.
Nevertheless, I plan on learning Julia in the near future.
The reason is that it's inherently much faster than R & Python.
Its only (but critical) drawback is the current weak coverage of packages. My guess is that eventually, when Julia has a broader offer in data science + statistics & machine learning, people will start switching.
In finance, computation times are often an issue. Either for simulations for option pricing (which I code in C, then load in R), or for portfolio backtesting in high dimensions. R loops are too slow (even with functional programming (via purrr) and parallel computing (via furrr, parallel, foreach)). That's why I think the only relevant switch is towards Julia in the long run, not Python (see also https://venturebeat.com/2019/02/18/facebooks-chief-ai-scientist-deep-learning-may-need-a-new-programming-language/).