Some reasons to create a web site using Python
If you ask some astronomers about how to create a web page, most answers would mention HTML or PHP. Few would mention Python as a language to create a web page. However, the page that you are reading (and many others) has been created using Python. In this new post, I explain some of the advantages to create web pages using Python.
In the last few years, Python has become one of the most used languages in the academic and scientific world. Even several reasons can be put forward, Python has four essential ingredients:
It is a language that can be learned really fast.
It is installed by default in most platforms, or it can be easily installed in the others.
Since it is an interpreted language, code does not have to be compiled.
It has a great amount of libraries that are installed by default.
Updated on October 23, 2021
Link added to explain better what I mean by non-compiled code.
It is important to remark that a scientist will only use those tools that provide the fastest possible way to compute a complex calculation. Therefore, if a code takes slightly longer to execute, but can be learned in one afternoon, that will be chosen in most of the cases.
The reasons expressed above have stimulated the scientific community to develop and use a huge amount of programs and libraries written in Python. Some of these packages are really popular, like Numpy, Scipy, Matplotlib, IPython, Sympy and Pandas, to mention just a few. Astronomical community has not been an exception to this rule, developing and using programs and libraries like Astropy and Sunpy.
Updated on October 23, 2021
PyRAF is no longer supported. Sunpy has been included instead.
All these applications are used mainly locally. When we need a library, we download it to our computer and execute it in there. All the vitality that the scientific community has shown to develop new applications in Python, disappears when developing web applications. There are many reasons to explain this phenomenon, but I would highlight three largely extended beliefs:
A web application provides no great advantages to do the calculations that a scientist needs.
New programming languages have to be learned, like HTML, CSS, Javascript, …
None of these languages has the libraries that scientists need to create their applications.
The first one of the points above can be easily refuted. Most astronomers use technologies like the Virtual Observatory, the ADS, or the CDS. All these technologies are used, basically, to share information among the astronomers. In fact, the Virtual Observatory is much more than a web site: it is a set of protocols, agreed among the astronomical community, to share information through Internet. In addition, most professional observatories share their data using the Virtual Observatory.
The other two points above are not completely true, either, if we add Python to the list of languages. In fact, Python is the only language required if we use some applications like the IPython Notebook (currently known as Jupyter). Since the release of IPython Notebook 2.0, in April 2014, it represents an excellent tool to use Python in our browser, by knowing Python only. In addition, since we are using Python, we have access to all the libraries in Python, in the same way that we were working locally.
Updated on October 23, 2021
Ipython Notebook was split from Ipython during 2014-2015 and renamed as Jupyter. Link has been updated to the Jupyter site.
In case that we do not want to use IPython Notebook, it is rather common that scientists know some HTML and CSS, likely because it is relatively easy to learn. In addition, at the dawn of the Internet era, HTML and CSS were the only languages needed to create a web page. In fact, a web page can be created using only HTML and CSS even today, but there is something that we will never be able to do with those two languages: loading dynamic content. Dynamic content includes all those elements that can change, depending on the person or the moment that the web page is loaded.
Two types of programs can be used to load dynamic content: the ones loaded on the server that provides the web page, and the ones being executed on the client side, where we are loading the web page. On the server side, almost any code can be executed, but not all the languages can be perfectly integrated in a HTML document. In fact, this is the main reason because PHP has extended as one of the main programming languages to create web pages. Unfortunately, the number of astronomical (or scientific) libraries in PHP is almost non-existent, and we have to use other languages for that. It is precisely in this particular aspect that Python stands out. Python can be seamlessly integrated in a HTML document, without the need for any other programming language.
Regarding the codes that can be executed on the client side, they must be executed by our browser. Even several extensions can be installed to understand other languages, JavaScript is the only code that the main browsers understand by default, in addition to HTML and CSS. Once again, the number of astronomical libraries in JavaScript is rather small, even some interesting libraries are starting to be developed.
Updated on October 23, 2021
Link to Aladin Lite documentation updated.
Summarizing
Some of the main reasons to create web site using Python are:
It is a language that can be learned really fast.
It is installed by default in most platforms, or it can be easily installed in the others.
Since it is an interpreted language, code does not have to be compiled.
It has a great amount of libraries that are installed by default.
A large community is developing new scientific libraries, including astronomical ones.
It can be seamlessly integrated in a HTML document, allowing dynamic content to be loaded in a web site.
The web page that you are currently reading is a good example on how you can create a web page, knowing only HTML, CSS and Python.
Updated on October 23, 2021
The post on how to create a web site in Python was largely outdated and has been removed. The next post will no longer explain that.