slice(stop)slice(start, stop, [step])<ul><li>Standard indexing relies on slice(): a[start:stop:step] calls a.__getitem__(slice(start, stop, step)).</li><li>slice() creates a shallow copy of the selected elements from the original list.</li><li>Whenever you slice a list, the Python interpreter creates a new list and populates it with references to the elements you selected from the original list. Given that a reference needs 8 bytes on 64-bit systems, the overhead of copying references becomes noticeable when working with large lists. For example, slicing 1000 elements allocates almost 8kb worth of references.</li></ul>

slice

islice(iterable, stop)islice(iterable, start, stop, [step])<ul><li>Returns an iterator that produces the selected elements from the iterable. It skips start elements from the iterable, and then produces values until stop is reached, or until the iterator is exhausted. This affects execution speed when it needs to skip lots of elements.</li><li>It's memory efficient because it doesn't allocate extra memory for the iterated elements, and it can be used to slice both iterables and iterators. Standard indexing (a[start:stop]) doesn't work on iterators because they don't implement __getitem__().</li></ul>

itertools.islice

Choosing between slice() and islice() is a trade-off between memory usage and iteration speed. slice() is faster, but requires more memory, whereas islice() allocates a constant amount of memory, but is slower.Use slice() when:<ul><li dir = "auto">you need fast iteration speed, and you afford the extra memory usage</li></ul>Use islice() when:<ul><li>you need to slice iterators, or</li><li>memory efficiency is important to you</li></ul>

slice or islice?

Software Development

slice or islice? When to use which

Returns an infinite iterator that produces evenly spaced values starting with number start.It's often used in combination with map() to generate consecutive data points, and with zip() to add sequence numbers.

itertools.count(start=0, step=1)

Assume you have three lists and you want to turn them into a table-like structure, where each row has a numeric ID.You can combine count() with zip() to add sequence numbers to the resulting tuples. See code in image.

Add sequence numbers

Python's Infinite Count Generator

A generator function is any function that has the yield keyword in its body. When called, generator functions return a generator object that wraps the body of the function.Generator objects can be consumed like any other iterator. When we call next() on them, execution advances to the next yield statement in the function body, and the yielded value is returned. Finally, when the function body returns, the generator object returns StopIteration.See example in image: gen_AB() returns a generator that can be iterated over. next() is called on it 3 times in total.

How generator functions work

Generator Functions in Python

Idiomatic refers to the most common way to write something in a language. Writing idiomatic code means following the conventions of the language, and solving problems the way they are meant to be solved in that language, not by porting programming patterns from other languages.Idioms vary widely across languages.

Idiomatic code

Learning and internalizing the idioms of a language greatly improves your reading abilities. You are able to focus on understanding what the code does instead of wondering why the code is written in an awkward way.Projects that use idiomatic code benefit from:<ul><li>Higher maintainability - Idiomatic code is clear, maintainable, and bug-free</li><li>Faster code reading &amp; Lower cognitive load - People are no longer stopping at every line, trying to figure out what it does, while struggling to keep in mind what came before</li><li>Faster debugging - It's easier to spot bugs in idiomatic code</li></ul>

Benefits of idiomatic code

Writing idiomatic code implies relying on the standard library to solve problems, and not reimplementing functionality that is already maintained by core developers.This has two main benefits:<ol><li>Others will find it easier to read and maintain your code because you use packages familiar to them</li><li>You save time by not having to implement functionality from scratch</li></ol>

Use the standard library

Write Idiomatic Code

<ul><li>Python is a general-purpose, object-oriented programming language.</li><li>It emphasises code readability by using white space.</li><li>It is easy to learn.</li><li>It is a favourite of programmers and developers.</li><li>Python is very well suited for use in machine learning at a large scale.</li><li>Its suite of specialised deep learning and machine learning libraries includes tools like scikit-learn, Keras and TensorFlow. It enables data scientists to develop sophisticated data models that plug directly into a production system.</li></ul>

Python

<ul><li>R is also is an open-source programming language.</li><li>It's optimised for statistical analysis and data visualisation.</li><li>R has a rich ecosystem with complex data models and practical tools for data reporting.</li><li>R is popular among data science scholars and researchers.</li><li>R provides various libraries and tools for cleansing and prepping data, creating visualisations, and training and evaluating machine learning and deep learning algorithms.</li><li>R is used within RStudio.</li><li>R applications can be used directly on the web via Shiny.</li></ul>

They approach data science differently: <ul><li dir = "auto">R is used for statistical analysis.</li><li dir = "auto">Python provides a more general approach to data.</li></ul> <ul><li dir = "auto">R is built by statisticians and leans heavily into statistical models and specialised analytics.</li><li dir = "auto">Python is a multipurpose language that can provide data analysis or machine learning in scalable production environments.</li></ul> <ul><li>You might use R for customer behaviour analysis or genomics research.</li><li>You might use Python for developing a machine learning application.&nbsp;</li></ul>

The main difference between R and Python

Data collection<ul><li dir = "auto">Python supports all kinds of data formats.</li><li dir = "auto">R is designed for data analysts to import data from Excel, CSV and text files.</li></ul> Data exploration<ul><li dir = "auto">In Python, you explore data with Pandas, the data analysis library for Python.</li><li dir = "auto">With R, you can build probability distributions, apply different tests, use standard machine learning and data mining techniques.</li></ul> Data modelling<ul><li>Python has standard libraries.</li><li>With R, you'll sometimes have to rely on packages outside of R's core functionality.</li></ul>

Other key differences

The language to choose depends on your situation. Points to consider:<ul><li>Do you have programming experience? Python has a learning curve that's linear and smooth. With R, novices can Run data analysis tasks in minutes, but it takes longer to develop expertise.</li><li>What do your colleagues use? Academics, engineers and scientists use R. Python is used in a wide range of industries.</li><li>What problems are you trying to solve? R is used for statistical learning, and Python for machine learning and large-scale applications.</li></ul>

Python vs. R: which one to choose

Explore the basics of these two open-source programming languages, the key differences that set them apart and how to choose the right one for your situation.

python

slice

slice or islice? When to use which

itertools.count(start=0, step=1)

Python's Infinite Count Generator