I am working in the domain of big data and machine learning. I am helping companies with bringing their machine learning models to the production. In many projects there is a tendency to port Python, PySpark code to Scala and Scala Spark.

This yields to longer time to market and a lot of mistakes due to necessity to understand and re-write the code. Also many libraries/apis that data scientists/machine learning practitioners use are not available in jvm ecosystem.

Simply, refactoring (if necessary) and organising the code of the data scientists by following best practices of software development is less error prone and faster comparing to re-write in Scala.

Pipeline orchestration tools such as Luigi/Airflow is python native and fits well to this picture.

I have heard some arguments against Python such as, it is slow, or it is hard to maintain due to its dynamically typed language. However cost/benefit of time consumed porting python code to java/scala alone would be enough as a counter-argument. ML pipelines rarerly contains a lot of code (if that is not the case, such as complex domain and significant amount of code, then scala would be a better fit).

In terms of performance, I did not see any issues with Python. It is not the fastest runtime around but ML applications are rarely time-critical (majority of them is batch based).

I still prefer Scala for developing APIs and for applications where the domain contains complex logic.

This post is a bit of an obvious one, as we have a web application, we obviously need to have HTML and CSS in our stack. Though specifically though, we can talk a bit about backward compatibility and the specific approaches we want to enforce in our codebase.

HTML : Not much explanation here, you have to interact with HTML for a web app. We will stick to the latest standard: HTML 5.

CSS: Again if we want to style any of our components within he web, we have to use to style it. Though we will be taking advantage of JSS in our code base and try to minimize the # of CSS stylesheets and include all our styling within the components themselves. This leaves the codebase much cleaner and makes it easier to find styles!

Babel: We understand that not every browser is able to support the cool new features of the latest node/JS features (such as redue, filter, etc) seen in ES6. We will make sure to have the correct Babel configuration o make our application backward compatible.

Material UI (MUI): We need to make our user interface as intuitive and pretty as possible within his MVP, and the UI framework used by Google will provide us with exactly that. MUI provides pretty much all the UI components you would need and allows heavy customization as well. Its vast # of demos will allow us to add components quickly and not get too hung up on making UI components.

We will be using the latest version of create-react-app which bundles most of the above along many necessary frameworks (e.g. Jest for testing) to get started quickly.

For our front-end, React is chosen because it is easy to develop with due to its reusable components and state functions, in addition to a lot of community support. Because React is popular, it would be easy to hire for it here at our company MusiCore. Our team also has experience with React already. React can be written with ES6 and ES6 has a lot of popularity and versatility when it comes to creating classes and efficient functions. Node.js will be used as a runtime environment to compile the code. Node.js also has many different types of open-source packages that can help automate some of the tasks we want to do for the application. CSS 3 will be used to style components and is the standard for that.

We needed to incorporate Big Data Framework for data stream analysis, specifically Apache Spark / Apache Storm. The three options of languages were most suitable for the job - Python, Java, Scala.

The winner was Python for the top of the class, high-performance data analysis libraries (NumPy, Pandas) written in C, quick learning curve, quick prototyping allowance, and a great connection with other future tools for machine learning as Tensorflow.

The whole code was shorter & more readable which made it easier to develop and maintain.

Optimize-js I will not describe this tool a lot here, because it's already good done by author on github

I just want to mention that this tool wrap up all immediately-invoked functions or likely-to-be-invoked functions in parentheses what is do a great optimization a JavaScript file for faster initial execution and parsing (based on my experience).

The performance of application where I've introduced optimize-js improved on 20% in a common (tested in Chrome and IE11).

• Clarification on Readme to the optimize-js
• Some of Nolan thoughts on the virtues of compile-time optimizations can be found in "Parens and Performance" – counterpost

Is it maintaining now? - Unfortunately, no (but feel free to send PR)