Why use pandas and numpy for data analysis?

Question

Python

I'm currently taking a data analytics course on Udacity. In particular, it discusses the use of Python, as well as the pandas and numpy libraries for analysis. The examples given there can be done just as well, and I suspect with much better performance on that SQL. Accordingly, the question arose - why use this rather slow interpreted language for things that are much better and more organically done by the database? Are there any special use cases, or reasons why Python is used?

Answer 1

Python is used because it can then be built into a working application. So does Google, Yandex, etc.
Numpy, etc. libraries are written in C so they are not as slow as they seem.
Where the speed of python libraries written in C is lacking, C++ is used rather than SQL.
SQL is suitable for some tasks, but it is impossible to draw a graph using SQL, but it is possible using the same python libraries. In data analysis, graphs are often built, vectors are calculated, etc. works, it is inconvenient to write such things in SQL.
If SQL or python is not used, then wolfram mathematica or mathcad, etc. are usually used. mathematical programs where it is convenient to count matrices, vectors, derivatives and it is convenient to build graphs.
Using python, they write software that uses the result of data analysis, so the data processing itself is written in python (or C ++).
For python, there are large machine learning libraries like google tensor flow https://www.tensorflow.org/ which again suggests that this language is the standard for writing applications related to machine learning, just like SQL is the standard for database work.

Answer 2

Python's joke is that in specific tasks it is suddenly not slow), both
in NumPy and in WEB applications,
so the phrase "pretty slow interpreted" works as a detector)

Answer 3

> with much better performance on the SQL volume. Accordingly, the question arose - why use
> this rather slow interpreted language for things that are much better and more organic
> done by the database?
In the popular OpenCV library , numpy is used for image processing and pattern recognition. Including on video.
I would like to see how you do it faster and more organically in SQL. :)

Answer 4

Python is used there to pull api written in C