4 Working with collections in Julia

In chapters 2 and 3 you learned basic elements of the Julia language. We have mostly used scalar types (like numbers) in all the examples. However, in data science you will typically work with data collections, that is, groupings of some variable number of data items. One collection type we have already introduced in chapter 2 is a vector.

In this chapter you will learn how to use several fundamental collections that are most used in practical scenarios: arrays, dictionaries, tuples, and named tuples.

4.1 Working with arrays

In this section you will learn the basics of working with arrays in Julia: their creation, indexing into arrays and most common operations you can expect to perform when working with arrays. Arrays are commonly used collections in data science. Most of the machine learning algorithms expect data stored in arrays as their inputs. In Julia, as opposed to, for example, Python, arrays are part of the language specification so they are equipped with a convenient syntax, there is only one set of rules for working with them you will need to learn, and they are fast.

4.1.1 Getting the data into a matrix

4.1.2 Computing basic statistics of the data stored in a matrix

4.1.3 Indexing into arrays

4.1.4 Performance considerations of copying vs. making a view

4.1.5 Calculation of correlations between variables

4.1.6 Fitting a linear regression

4.1.7 Plotting the Anscombe's quartet data

4.2 Mapping key-value pairs with dictionaries

4.3 Putting structure to your data using named tuples

4.3.1 Defining named tuples and accessing their contents

4.3.2 Analyzing Anscombe's quartet data stored in a named tuple

4.3.3 Understanding composite types and mutability of values in Julia

4.4 Summary