The basics
Julia has a steeper learning curve than Python, but once you get past the basics, Julia is seen as much easier to use and understand. Python is good for beginners; however, it can be more difficult to understand complicated concepts because they are often unclear. Julia is also much faster than Python in terms of performance.
—
The following is my short answer to the question: “What are the basics of Julia?”. Before answering it, make sure you understand that neither Julia nor Python is a perfect programming language. Neither one is intended to be adopted as a primary language for serious scientific computing. If you have any prior experience with either Python or Julia, I suggest that you skip all this and go straight to the [[How To’s|How-To’s]]. There are several in-place code examples that use these languages in a straightforward manner.
—-
Geography and History: Julia is created by the same people who created MATLAB. You may have heard of them – they’re called Numerical Python. They and their new company are called Julia Computing. As of May, 2016, they’ve been funded by Red Hat.
Julia was developed in 2009. On the 25th Anniversary of the Mac OS X operating system, Apple released the first version of a programming language for it: Python 3. It’s called Python 3.0. The history of Julia is all in one place [[here|PyCon tutorial]] so I won’t repeat it here (see “The basics” at the start of this article).
Julia is a high-level programming language that runs on the JVM (Java Virtual Machine), the same platform used by Java. Julia’s goal is to make it easier to work with large data sets, spreadsheets and databases. It also has a small memory footprint, which means Julia programs don’t need so much memory. Julia is a JIT (Just-in-Time) language, in which the JVM compiler generates machine code to run the program at runtime. Julia has built-in support for parallel programming, making it easier to use multiple CPUs. It is also being developed as an open source project. Julia was designed by the team behind Revolution Analytics and has been in development since 2012.
Julia is a high-level programming language that runs on the JVM (Java Virtual Machine), the same platform used by Java. Julia’s goal is to make it easier to work with large data sets, spreadsheets and databases. It also has a small memory footprint, which means Julia programs don’t need so much memory. Julia is a JIT (Just-in-Time) language, in which the JVM compiler generates machine code to run the program at runtime. Julia has built-in support for parallel programming, making it easier to use multiple CPUs. It is also being developed as an open source project. Julia was designed by the team behind Revolution Analytics and has been in development since 2012.
Julia and Python have similar syntax (Python is very easy to learn from a Julia perspective). Both are used for scientific computing, and can be used for web development. However, Julia is a statically typed language, while Python is dynamically typed. Python has its own standard library, while Julia’s standard library is a subset of Python’s. Julia supports multiple dispatch, which means a function can have many versions, and it will be chosen based on the type of its arguments at the time it is called. This allows for more flexible ways to write functions; for example, you can have a generic function that works for many different types of objects. Julia also has generics, but they are much more restricted than Python’s multiple dispatch. Julia is intended as a more performant language than Python, especially in machine learning and scientific computing applications. This can be attributed to the fact that it has less type information, making it easier to optimize code written in it.
“How is Julia different from Python?” : “Julia is strongly typed, with classes and methods that take and return types. Julia has a unified type system, meaning you can pass any type to any method. Julia is missing overloaded functions, which are types that take multiple different types at the same time. Python has multiple dispatch, which allows for more flexible function definitions. You can define functions that work on many different types at once.”
“What is Julia? From Wikipedia, the free encyclopedia.” : “Julia is a high-level, high-performance dynamic programming language designed to suit technical computing and data science at scale. It provides a sophisticated compiler, distributed parallel execution, numerical accuracy, and an extensive mathematical function library.”
Development environments
Python can be programmed in almost any programming language, but Julia requires a Julia-specific compiler. You can use the julia command line to compile and run Python programs. Julia can be used in combination with PyPy, the Python virtual machine JIT compiler.
Python is a dynamically typed programming language. It employs duck typing.
Julia is a static typed programming language. It employs type inference and many-to-many relation between types and values, which makes the static typing unnecessary for most uses. This also allows Julia to have multiple dispatch, i.e., variables can have different types with the same name, as long as these types can be overloaded in the same way for any given operation.
Python was created by a Dutch programmer, Guido van Rossum. It was named “Python” in honor of the television show Monty Python’s Flying Circus.
Julia’s development environment is stronger in some ways, but Python’s is stronger in others. Python only needs one person to maintain the codebase, while Julia has some internal testing and a projet manager. Python has a more robust package system and tooling, whereas Julia’s package management is still being developed. Julia’s ecosystem is growing very quickly because of its high-level language approach that makes it easy to learn for many programmers. Python’s ecosystem is more mature, with more activity and a wider base of users.
To me (and many others), Python adds more value than Julia. I think it’ll take a lot for Julia to be really as powerful and flexible as Python, but I’m inspired by the work of the Julia team, and by all the exciting new projects that are starting up in Julia. Julia’s probably going to win in the long run, and I am excited to see it grow.
Side note: I have no idea how Julia can be better than Python in dealing with large, complex projects. In my opinion, the most important feature of Julia is the ability to write and debug high-level code using its dynamic language features. I find this much more useful than any language-specific tooling. If you want syntactic sugar for subtyping and metaprogramming, you can use C++ or any other object-oriented language.
Julia vs Python
