As a statistics major of University of Dhaka, I have had hands-on experience working with 2 programming languages R and Python.
About R: Firstly, as a beginner, I would always prefer R over any other programming languages. For core statistical problems, one should learn R. If someone wants to continue their career only with statistics, knowing R is enough for them. They must also know about two R libraries “dplyr” and “ggplot2”. “dplyr” is known as the grammar of data manipulation. “ggplot2” is a powerful tool to visualize data. It is called the grammar of graphics. If an individual wants to be an academician they might consider knowing R only. R has some complexity with brackets, parenthesis but is easy for mathematical and statistical operations.
Now on to Python: It is an industrial language in my point of view. The language has a lot of libraries and modules. One has to learn the basics first then they would be able to do the statistical works. This is a language for all types of professionals. But in terms of a statistician or data scientist they must know some extra libraries. Such as
1)Numpy: Used for vectors, matrices and arrays. Numpy has a lot of attributes, it works like a list and it is very fast as well. But it doesn’t have any direct vector and matrix commands like R.
2)Pandas: Basically the best tool for working with big data. Excel can’t operate with big data, in that case pandas is the best option for working with csv, txt, xlxs files etc. It basically creates a data frame or series. And one can add, subtract or edit values. They can edit the whole data set, sort them, describe them and do all kinds of operations they like to do.
3) Matplotlib: Best library for data visualization. The graphics are very clear and can be edited very easily. Better tool than ggplot2. But in order to draw a normal curve one has to import scipy as well. One might use seaborn for graphics as well but matplotlib is more easy to learn.
Python is overall a vast language. If any individual wants to learn machine learning and deep learning etc, they might consider learning scikit learn and TensorFlow etc. And as all over the world python is used more, so it is easy to attach Python codes in projects and group works. In order to work in real world industries, one should know python alongside R as a Data analyst or Data scientist. But as it is a very vast language, there are various ways of solving one particular problem. So a beginner might get frustrated too easily.
- কাজী শাফিন আলমhttps://www.thepapyrus.org/author/%e0%a6%95%e0%a6%be%e0%a6%9c%e0%a7%80-%e0%a6%b6%e0%a6%be%e0%a6%ab%e0%a6%bf%e0%a6%a8-%e0%a6%86%e0%a6%b2%e0%a6%ae/
- কাজী শাফিন আলমhttps://www.thepapyrus.org/author/%e0%a6%95%e0%a6%be%e0%a6%9c%e0%a7%80-%e0%a6%b6%e0%a6%be%e0%a6%ab%e0%a6%bf%e0%a6%a8-%e0%a6%86%e0%a6%b2%e0%a6%ae/
- কাজী শাফিন আলমhttps://www.thepapyrus.org/author/%e0%a6%95%e0%a6%be%e0%a6%9c%e0%a7%80-%e0%a6%b6%e0%a6%be%e0%a6%ab%e0%a6%bf%e0%a6%a8-%e0%a6%86%e0%a6%b2%e0%a6%ae/
- কাজী শাফিন আলমhttps://www.thepapyrus.org/author/%e0%a6%95%e0%a6%be%e0%a6%9c%e0%a7%80-%e0%a6%b6%e0%a6%be%e0%a6%ab%e0%a6%bf%e0%a6%a8-%e0%a6%86%e0%a6%b2%e0%a6%ae/
- কাজী শাফিন আলমhttps://www.thepapyrus.org/author/%e0%a6%95%e0%a6%be%e0%a6%9c%e0%a7%80-%e0%a6%b6%e0%a6%be%e0%a6%ab%e0%a6%bf%e0%a6%a8-%e0%a6%86%e0%a6%b2%e0%a6%ae/
