Book Review: Practical Python Data Visualization for a Strong Foundation


Ashwin Pajankar
Convenient Python data visualization
An accelerated approach to learning data visualization with Python
Apress 2021
XV, 160 pages, 37.44 euros
ISBN: 978-1-4842-6454-6

Because of its simplicity, Python is often not the programming language of choice in circles of “professional” developers – but there is no doubt that Guido van Rossum’s programming system, which was once developed as a language of choice. teaching, has now acquired a solid place among the “great” programming languages. A decisive reason for this is certainly that Python is flanked by a vast treasure trove of libraries, which saves developers a lot of work, especially in math and artificial intelligence. With “Practical Python Data Visualization”, Apress-Verlag now presents a book of approximately 170 pages which demonstrates the advantages of using Python in the realm of data visualization tasks.

(Image: Apress)

Author Ashwin Pajankar expects computer literacy, but hands-on experience with Python is not required. This is particularly evident in the first two chapters, in which readers learn how to install Python on the workstation and how to use the “interactive mode” interpreter as a sort of pocket calculator for advanced functions. This is followed by the basics of working with Jupyter notebooks, which serve as a working environment for the first visualization tasks. The explanations on Jupyter notebooks are not an end in themselves, but the author uses them in the rest of the book to execute Python code and demonstrate the various libraries available.

The popularity and widespread use of Python in data science and machine learning has resulted in a wide range of different implementations for various purposes. Pajankar begins its data visualization journey with the Leather library. The relatively simple library performs well when rendering line and bar charts. The examples presented are limited to the creation of basic graphics, which are then supplemented by a legend and additional information. However, anyone who expects information about acquiring data from databases or via REST at this point will be disappointed.

For data visualization and scientific computing with Python, NumPy and MatPlotLib are two comprehensive libraries that you can’t ignore. Due to its above-average importance, the author devotes an entire set of chapters to this combination, which sheds light on libraries from several angles. Since NumPy typically prefers its own data structures for storing the information to be processed, the first act of the data visualization journey through the world of Python libraries is devoted to a brief but sufficient explanation of how to keep data in. NumPy. Next come the line diagram experiments that have already been done with Leather – which, given the more extensive customization options of MatPlotLib, also deal with much more sophisticated visualization options.

MatPlotLib is not limited to two-dimensional images, but can produce three-dimensional graphics as well. In the sixth chapter the author uses a didactic trick and first shows the “loading” of two-dimensional bitmaps and their mass processing. What may seem superfluous at first glance turns out to be valuable in practice – especially considering that various math operations (eg in shader programming) use bitmaps to provide larger data fields. In addition, the creation of three-dimensional diagrams of lines and structures as well as level tables, which are important in a scientific context, are not neglected. In the next chapter, Pajankar shows how graphics can be presented in an attractive form.

Virtually no current data science textbook does without direct references to the corona pandemic. “Convenient Python Data Visualization” is no exception and first briefly introduces the Pandas library. However, the author then also uses the other processed libraries to show how the pandemic information data provided in the Covid library can be viewed. However, it also doesn’t fit into finding information at this point in the book.

On nearly 170 pages, there is not enough room to fully cover the sometimes very extensive libraries. Pajankar doesn’t even try to do this, but focuses on bringing together the developers of those apps in which the different libraries can be used profitably. In doing so, he creates a solid base of basic knowledge, which can be individually extended through targeted further training. This approach makes the book highly recommended, especially for those interested who haven’t yet covered data visualization with Python in detail.

Hanna there
has been dealing with handheld computers and electronics since 2004. He currently focuses on interdisciplinary applications of information technology.


Disclaimer: This article is generated from the feed and is not edited by our team.

Source link


About Author

Leave A Reply