Variables in Python: Usage and Best Practices – Real Python

Creating Variables With Assignments

The primary way to create a variable in Python is to assign it a value using the assignment operator and the following syntax:

In this syntax, you have the variable’s name on the left, then the assignment (=) operator, followed by the value you want to assign to the variable at hand. The value in this construct can be any Python object, including strings, numbers, lists, dictionaries, or even custom objects.

Here are a few examples of variables:

In this code, you’ve defined several variables by assigning values to names. The first five examples include variables that refer to different built-in types. The last example shows that variables can also refer to custom objects like an instance of your SomeCustomClass class.

Setting and Changing a Variable’s Data Type

Apart from a variable’s value, it’s also important to consider the data type of the value. When you think about a variable’s type, you’re considering whether the variable refers to a string, integer, floating-point number, list, tuple, dictionary, custom object, or another data type.

Python is a dynamically typed language, which means that variable types are determined and checked at runtime rather than during compilation. Because of this, you don’t need to specify a variable’s type when you’re creating the variable. Python will infer a variable’s type from the assigned object.

For example, consider the following variables:

In this example, name refers to the "Jane Doe" value, so the type of name is str. Similarly, age refers to the integer number 19, so its type is int. Finally, subjects refers to a list, so its type is list. Note that you don’t have to explicitly tell Python which type each variable is. Python determines and sets the type by checking the type of the assigned value.

Because Python is dynamically typed, you can make variables refer to objects of different data types in different moments just by reassigning the variable:

Now, age refers to a string, and subjects refer to a set object. By assigning a value of a different type to an existing variable, you change the variable’s type.

Expressions

In Python, an expression is a simple statement that Python can evaluate to produce and return a value. For example, consider the following expressions that compute the circumference of two different circles:

Each of these expressions represents a specific computation. To build the expressions, you’ve used values and the multiplication operator (*). Python evaluates each expression and returns the resulting value.

The above expressions are sort of rigid. For each expression, you have to repeat the input values, which is an error-prone and repetitive task.

Now consider the following examples:

In this example, you first define variables to hold the input values. Then, you use those variables in the expressions. Note that when you build an expression using variables, Python replaces the variable by its value. As shown in the example, you can conveniently reuse the values in different expressions.

Another important point to note is that now you have descriptive names to properly identify the values used in the expressions.

To summarize, variables are great for reusing values in expressions and running computations with data that varies over time. In general, variables let you name or label objects so that you can reference and manipulate them later in the program. In the following sections, you’ll see use cases of variables in practice.

Object Counters

A counter is an integer variable that allows you to count objects. Counters typically have an initial value of zero, which increments to reflect the number of times a given object appears. To illustrate, say that you need to count the objects that are strings in a given list of objects. In this situation, you can do something like the following:

In this example, you create the str_counter variable by initializing it to 0. Then, you run a for loop over a list of objects of different types. Inside the loop, you check whether the current object is a string using the built-in isinstance() function. If the current object is a string, then you increment the counter by one.

At the end of the loop, str_counter has a value of 3, reflecting the number of string objects in the input list.

You reuse the expression str_counter += 1 in each iteration of the loop to increment the value of str_counter, making it change over time. This dynamic updating is a key feature of variables. As the name variables suggests, they are designed to hold values that can vary over time.

Accumulators

Accumulators are another common type of variable used in programming. An accumulator is a variable that you use to add consecutive values to form a total that you can use as an intermediate step in different calculations.

A classic example of an accumulator is when you need to compute the sum of numeric values:

In this example, the loop iterates over a list of numbers and accumulates each value in total. You can also use accumulators as parts of larger computations, for example to calculate the mean of a list of numbers:

Then, you take advantage of total to compute the average using the built-in len() function. You could have used an object counter instead of len(). Similarly, Python comes with several accumulator functions that you can often use instead of explicit accumulators. For example, you can use sum() instead of calculating total as above.

Temporary Variables

Temporary variables hold intermediate results that you need for a more elaborate computation. A classic use case for a temporary variable is when you need to swap values between variables:

In this example, you use a temporary variable called temp to hold the value of a so that you can swap values between a and b. Once you’ve done the swap, temp is no longer needed.

For a more elaborate example of using temporary variables, consider the following function that calculates the variance of a sample of numeric data:

The expression that you use as the function’s return value is quite involved and challenging to understand. It’s also difficult to debug because you’re running multiple operations in a single expression.

To make your code easier to understand and debug, you can take advantage of an incremental development approach that uses temporary variables for intermediate calculations:

In this alternative implementation of variance(), you calculate the variance in several steps. Each step is represented by a temporary variable with a meaningful name, making your code more readable.

Boolean Flags

Boolean flags help you manage control flow and decision-making in your programs. As their name suggests, these variables can be either True or False. You can use them in conditionals, while loops, and Boolean expressions.

Suppose you need to perform two different actions alternatively in a loop. In this case, you can use a flag variable to toggle actions in every iteration:

Every time this loop runs, the conditional checks the value of toggle to decide which course of action to take. At the end of the loop, you change the value of toggle using the not operator. The next iteration will run the alternative action.

Flags are also used as function arguments. Consider the following toy example:

In this example, the verbose argument is a Boolean variable that lets you decide which greeting message to display.

You’ll find that a few Python built-in functions use flag arguments. The sorted() function is a good example:

The sorted() function takes an iterable as an argument and returns a list of sorted objects. This function has a reverse argument that is a flag and defaults to False. If you set this argument to True, then you get the objects sorted in reverse order.

In practice, you’ll find that Boolean variables are often named using the is_ or has_ naming pattern:

In this example, the is_adult variable is a flag that changes depending on the value of age. Note that the is_ naming pattern allows you to clearly communicate the variable’s purpose. However, this naming convention is just a common practice and not a requirement or something that Python reinforces.

Loop Variables

Loop variables help you process data during iteration in for loops and sometimes in while loops. In a for loop, the variable takes the value of the current element in the input iterable each time you go through the loop:

In this example, you iterate over a list of colors using a for loop. The loop variable, color, holds the current color in each iteration. This way, you can do something with the current item while you iterate over the data.

Python’s for loops can have multiple loop variables. For example, say that you want to map each color with its index. In this case, you can do something like the following:

In this example, you use the built-in enumerate() function to generate indices while you iterate over the input data. Note how this loop has two variables. To provide multiple loop variables, you use a comma-separated series of variables.

You can also use variables to control while loops. Here’s a toy example:

In this example, the loop variable works as a counter that defines the number of iterations. Inside the loop, you use the augmented subtraction operator (-=) to update the variable’s value.

Data Storage Variables

Data storage variables allow you to work with containers of values, such as lists, tuples, dictionaries, or sets. For example, say that you’re writing a contact book application that uses a list of tuples to store the information of your contacts:

This contacts variable allows you to manipulate your data using a single and descriptive name. You can use the variables in a loop, for example:

The first loop iterates over the items in the contact list and prints them as tuples. The second loop uses three loop variables to process every piece of data individually.

Note that inside the loop, you can use the variables as needed. You don’t need to use all the variables or use them in the same order. However, when you have an unused variable, then you should name it using an underscore to point out that it’s a throwaway variable.

So, you could rewrite the loop header to look like this: for name, phone, _ in contacts:. In this case, the underscore represents the email variable, which you aren’t using in the body of the loop.

Rules for Naming Variables

Variable names in Python can be any length and can consist of uppercase letters (A-Z) and lowercase letters (a-z), digits (0-9), and the underscore character (_). The only restriction is that even though a variable name can contain digits, the first character of a variable name can’t be a digit.

All of the following are valid variable names:

These variables follow the rules for creating valid variable names in Python. They also follow best naming practices, which you’ll learn about in the next section.

The variable name below doesn’t follow the rules:

The variable name in this example starts with a number, which isn’t allowed in Python. Therefore, you get a SyntaxError exception.

It’s important to know that variables are case-sensitive. Lowercase and uppercase letters aren’t treated the same:

In this example, Python interprets the names as different and independent variables. So, casing is something to consider when you’re creating variable names in Python.

Nothing stops you from creating two different variables in the same program called age and Age, or, for that matter, agE. However, this practice isn’t recommended because it can confuse people trying to read your code and even yourself after a while. In general, you should use lowercase letters when creating your variable names.

The use of underscore characters is also significant. You’ll use an underscore to separate multiple words in a variable name:

In these variable names, you use the underscore character as a separator for multiple words. This is a way to improve the readability of your code by substituting the space character with an underscore. To illustrate this, consider how your variables would look without the underscores:

Even though these names are technically valid, they can be challenging to read and understand at a glance. The lack of separation makes it harder to grasp the meaning of each variable quickly, and they require more effort to interpret. Using underscores improves the clarity of your code and makes it more maintainable.

Best Practices for Naming Variables

You should always give a variable a descriptive name that clearly explains the variable’s purpose. Sometimes, you can find a single word to name a given variable:

Variables always refer to concrete objects, so their names should be nouns. You should try to find specific names for your variables that uniquely identify the referred object. Names like variable, data, or value may be too generic. While these names can work for short examples, they’re not descriptive enough for production code.

In general, you should avoid single-letter names:

Single-letter names may be hard to decipher, making your code difficult to read, especially when you use them in expressions with other similar names. Of course, there are exceptions. For example, if you’re working with nested lists, then you can use single-letter names to identify indices:

It’s common to use letters like i, j, and k to represent indices, so you can use them in the right context. It’s also common to use x, y, and z to represent point coordinates, so these are also okay to use.

Using abbreviations to name variables is discouraged, in favor of using the complete name:

It’s best practice to use a complete name instead an abbreviated name because it’s more readable and clear. However, sometimes abbreviations are okay when they’re widely accepted and used:

In these examples, cmd is a commonly used abbreviation for command and msg is commonly used for message. A classic example of a widely used abbreviation in Python is the cls name, which you should use to identify the current class object in a class method.

Sometimes, you need multiple words to build a descriptive variable name. When using multi-word names, you can struggle to read them if there isn’t a distinguishable boundary between words:

This variable name is difficult to read. You have to pay close attention to figuring out the words’ boundaries so that you can understand what the variable represents.

The most common practices for multi-word variable names are the following:

• Snake case: Lowercase words are separated by underscores. For example: number_of_graduates.
• Camel case: The second and subsequent words are capitalized to make word boundaries easier to see. For example: numberOfGraduates.
• Pascal case: Similar to camel case, except the first word is also capitalized. For example: NumberOfGraduates.

The Style Guide for Python Code, also known as PEP 8, contains naming conventions that list suggested standards for names of different object types. Regarding variables, PEP 8 recommends using the snake case style.

When you need multi-word names, it’s common to combine an adjective as a qualifier with a noun:

In these examples, you create descriptive variable names by combining adjectives and nouns, which can dramatically improve your code’s readability. Another point to consider is to avoid multi-word names that start with my_, like my_file, my_color, and so on. The my_ part doesn’t really add anything useful to the name.

Flag variables are another good example of when to use a multi-word variable name:

In these examples, you’ve used an underscore to separate the words, making their boundaries visible and quick to spot.

When it comes to naming lists and dictionaries, you should use plural nouns in most situations:

Using plural nouns in these examples makes it clear that the variable refers to a container that stores several objects of similar types.

When naming tuples, you should consider that they’re commonly used to store objects of different types or meanings. So, it’s okay to use singular nouns:

Although these tuples store multiple objects, they represent a single entity. The first tuple represents an RGB (red, green, blue) color, while the second represents a row in a database table or some other tabular data.

Public and Non-Public Variable Names

A widely used naming convention for variables in Python is to use a leading underscore when you need to communicate that a given variable is what Python defines as non-public. A non-public variable is a variable that shouldn’t be used outside its defining module. These variables are for internal use only:

In this module, you have a non-public variable called _timeout. Then, you have a couple of functions that work with this variable. However, the variable itself isn’t intended to be used outside the containing module.

Restricted and Discouraged Names

Python reserves a small set of words known as keywords that are part of the language’s syntax. To get the list of Python’s keywords, go ahead and run the following code:

In most cases, you won’t be able to use these words as variable names without getting an error:

This behavior is true for most keywords. If you need to use a name that coincides with a keyword, then you can follow PEP 8’s recommendation, which is to add a trailing underscore to the name:

In this example, you’ve added an underscore character at the end of the keyword, which enables you to use it as a variable name. Even though this convention works, sometimes it’s more elegant to do something like the following:

Now, your variable’s name is way more descriptive and specific, which improves your code’s readability.

There are also soft keywords, which are keywords only in specific contexts. For example, the match keyword is only considered a keyword in structural pattern matching.

Because match is a soft keyword, you can use it to name variables:

In this example, you import the re module to use regular expressions. This example only searches for a basic expression in a target text. The idea here is to show that match can be used as a valid variable name even though it’s a keyword.

Another practice that you should avoid is using built-in names to name your variables. To illustrate, say that you’re learning about Python lists and run the following code:

In this example, you’ve used list as the name for a list object containing numeric values. This shadows the original object behind the name, which prevents you from using it in your code:

Now, calling list() fails because you’ve overridden the built-in name in your code. So, you’re better off avoiding built-in names to define your variables. This practice can make your code fail in different ways.

Note that this recommendation is also valid for names that refer to objects defined in third-party libraries that you use in your code.

Variables Hold References to Objects

What happens when you create a variable with an assignment? This is an important question in Python because the answer differs from what you’d find in many other programming languages.

Python is an object-oriented programming language. Every piece of data in a Python program is an object of a specific type or class. Consider this code:

When presented with the statement 300, Python does the following operations:

1. Creates an integer object
2. Gives it a value of 300
3. Displays it on the screen

You can see that an integer object is created using the built-in type() function:

A Python variable is a symbolic name that refers to or points to an object like 300. Once an object is assigned to a variable, you can refer to it by the variable’s name, but the data itself is still contained within the object.

For example, consider the following variable definition:

This assignment creates an integer object with a value of 300 and makes the variable n point to that object. The diagram below shows how this happens:

In Python, variables don’t store objects. They point or refer to objects. Every time you create an object in Python, it’s assigned a unique number, which is then associated with the variable.

The built-in id() function returns an object’s identifier or identity:

In CPython, the standard Python distribution, an object’s identity coincides with its memory address. Therefore, CPython variables store memory addresses. Through these memory addresses, variables access the concrete objects stored in memory.

You can create multiple variables that point to the same object. In other words, variables that hold the same memory address:

In this example, Python doesn’t create a new object. It creates a new variable name or reference, m, which points to the same object that n points to:

Next, suppose you do something like this:

Now, Python creates a new integer object with the value 400, and m becomes a reference to it:

Finally, say that you run the following statement:

Now, Python creates a string object with the value "foo" and makes n a reference to that:

Because of the n and m reassignments, you no longer have a reference to integer object 300. It’s orphaned, and you have no way to access it again.

When the references to an object drop to zero, the object is no longer accessible. At that point, its lifetime is over. Python reclaims the allocated memory so it can be used for something else. In programming terminology, this process is known as garbage collection.

Variables Have Dynamic Types

In many programming languages, variables are statically typed, which means they’re initially declared to have a specific data type during their lifetime. Any value assigned to that variable during its lifetime must be of the specified data type.

Python variables aren’t typed this way. In Python, you can assign values of different data types to a variable at different moments:

In this example, you’re making the value variable refer to or point to an object of another type. Because of this feature, Python is a dynamically typed language.

It’s important to note that changes in a variable’s data type can lead to runtime errors. For example, if a variable’s data type changes unexpectedly, you may face type-related bugs or even get an exception:

In this example, the variable type changes during the code’s execution. When value points to a string, you can use the .upper() method to convert the letters into uppercase. However, when the type changes to float, the .upper() method isn’t available, and you get an AttributeError exception.

Variables Can Use Type Hints

You can use type hints to add explicit type information to your variables. To do this, you can use the following Python syntax:

The square brackets aren’t part of the syntax. They denote that the enclosed part is optional. Yes, you can declare a Python variable without assigning it a value:

The variable declaration on the first line works and is valid Python syntax. However, this declaration doesn’t really create a new variable for you. That’s why when you try to access the number variable, you get a NameError exception. Even though number isn’t defined, Python has recorded the type hint:

When it comes to basic data types such as numbers and strings, type hints may look superfluous:

If you’re familiar with Python’s built-in types, then you won’t need to add type hints to these variables because you’ll soon know that you have a string, integer, and floating-point value, respectively. So, in this situation, you’re okay to skip the type hint. Also, your static type checker or linter won’t complain.

The story changes when you use collection types, such as lists, tuples, dictionaries, and sets. With these types, it makes sense to provide type hints for their contained data.

To illustrate, say that you have the following dictionary of colors:

In this case, it’d be great to have information about the data types of keys and values. You can provide that information using the following type hint:

In this updated code, you’re explicitly communicating that your colors dictionary will store its keys and values as strings.

Why is this type hint important? Say that in another part of your code, you have another color dictionary defined as shown below:

At some point, you might be working with both dictionaries and won’t have an unambiguous way to know which dictionary is needed. To work around the issue, you can add a type hint to the second dictionary as well:

The type hint in this example is a bit more elaborate, but it clearly states that your dictionary will have string keys and tuples of three integers as values.

It’s important to note that type-hinting variables that refer to container data types is especially useful when you initialize the variable with an empty container:

In these examples, you have two empty lists. Because they have type hint information, you can quickly figure out the data type of the content. Now you know that the first list will contain strings, while the second list will hold tuples, each consisting of three items: a string, an integer, and another string value.

Note that in this example, your static type checker or linter will explicitly complain about the type hint of your lists if you don’t provide any type information.

Later in your code, you can append items to each list using the correct data type:

First, you use the .append() method to add new fruits as strings to the end of fruits, and then you add new rows to the end of rows.

Parallel Assignment

Python also allows you to run multiple assignments in a single line of code. This feature makes it possible to assign the same value to several variables simultaneously:

The parallel assignment in this example initializes three different but related variables to False simultaneously. This way of creating and initializing variables is more concise and less repetitive than the following:

By using parallel assignments instead of dedicated assignments, you make your code more concise and less repetitive.

Iterable Unpacking

Iterable unpacking is a cool Python feature also known as tuple unpacking. It consists of distributing the values in an iterable into a series of variables. In most cases, the number of variables will match the number of items in the iterable. However, you can also use the *variable syntax to grab several items in a list.

You can use iterable unpacking to create multiple variables at a time using an iterable of values. For example, say that you have some data about a person and want to create dedicated variables for each piece of data:

If you didn’t know about iterable unpacking, then your first approach might be to distribute the data into different variables manually, as shown below:

This code works. However, using indices to extract the data may lead to an error-prone and hard-to-read result. Instead of using this technique, you can take advantage of iterable unpacking and end up with the following code:

Now, your code looks cleaner and more readable. So, when you find yourself creating variables from iterables using indices, consider using unpacking instead.

A great use case for unpacking is when you need to swap the values between two variables:

In the highlighted line, you swap the values of a and b without using a temporary variable, as you saw before. In this example, it’s important to note that the iterable to the right of the equal sign is a tuple of variables.

Assignment Expressions

Assignment expressions allow you to assign the result of an expression used in a conditional or a while loop to a name in one step. For example, consider the following loop that takes input from the keyboard until you type the word "stop":

This loop works as expected. However, this code has the drawback that it unnecessarily repeats the call to input().

You can rewrite the loop using an assignment expression and end up with the following code:

In the expression (line := input("Type some text: ")), you create a new variable called line to hold a reference to the input data. This data is also returned as the expression’s result, which is finally compared to the "stop" word to finish the loop. So, assignment expressions are another way to create variables.

To learn more about assignment expressions, check out The Walrus Operator: Python’s Assignment Expressions.

Global, Local, and Non-Local Variables

Global variables are those that you create at the module level. These variables are visible within the containing module and in other modules that import them. So, for example, if you’re using the Python REPL, then the current global scope is defined in a module called __main__.

This module defines your current global scope. All the variables defined in this module are global, so you can use them at anytime during an interactive session run:

In this code snippet, you define the value variable and call the built-in dir() function to check the list of names defined in your current global scope. At the end of the list, you’ll find the 'value' entry, which corresponds to your variable.

Now, say that you use this same interactive session to run several pieces of code while you try out some cool features of Python. After all those examples, you need to use the value variable again:

You’ll notice that the variables will still be available for you. That’s because this value variable is global to your code.

You can also import variables defined outside your current module by using the import statement. A typical example of this practice is when you have a module that defines a variable to hold some configuration parameters. You can import this variable into your global scope and use it as you’d use any other global variable.

As a quick example, say that you have the following config.py module:

This file defines a dictionary that contains several configuration parameters for a hypothetical application. You can import this variable to your app’s global scope and use the settings parameters as needed:

With the import in the first line, you’ve brought the settings variable into your current global scope. You can use the variable from any part of your code.

Local variables are those that you define inside a function. These variables can help you store the results of intermediate computations under a descriptive name. They can make your code more readable and explicit.

Consider the following example:

Local variables are only visible within their defining function. Once the function returns, the variables disappear. That’s why you can’t access integer in the global scope.

Similarly, non-local variables are those that you create in a function that define inner functions. The variables local to the outer function are non-local to the inner functions. Non-local variables are useful when you’re creating closure functions and decorators.

Here’s a toy example that illustrates how global, local, and non-local variables work and how to identify the different scopes in Python:

In this example, you first create a global variable at the module level. Then, you define a function called outer_func(). Inside this function, you have nonlocal_variable, which is local to outer_func() but non-local to inner_func().

In inner_func(), you create another variable called local_variable, which is local to the function itself.

The calls to print() are intended to show how you can access variables from different scopes inside a function.

Here’s how this function works:

In summary, global variables are accessible from every place in your code. Local variables are visible in their defining function. Non-local variables are visible in their defining or enclosing function and in any inner function that lives in the enclosing function.

Class and Instance Variables (Attributes)

You can create variables inside custom Python classes when using object-oriented programming tools. These variables are called attributes. In practice, you can have class and instance attributes.

Class attributes are variables that you create at the class level, while instance attributes are variables that you attach to instances of a given class. These attributes are visible only from within the class or its instances. So, classes define a namespace, which is similar to the scope.

To illustrate how class and instance attributes work, say that you need a class to represent your company’s employees. The class should keep information about the employee at hand, which you can store in instance attributes like .name, .position, and so on. The class should also keep a count of how many employees are currently in the company. To implement this feature, you can use a class attribute.

Here’s a possible implementation of your class:

In this Employee class, you define a class attribute called count and initialize it to 0. You’ll use this attribute as a counter to keep track of the number of Employee instances. Class attributes like .count are common to the class and all its instances.

Inside the initializer, .__init__(), you define three instance attributes to hold the name, position, and salary of the employee. The last line of code in .__init__() increments the counter by 1 every time you create a new employee. To do this, you access the class attribute on the class object itself.

Finally, you have a method to display the employee’s profile according to their current information. This method shows that you need to use the self argument to access instance attributes within a class.

Here’s how you can use this class in your code:

In this code, you first create two instances of Employee by calling the class constructor with appropriate arguments. Then, you call the display method on both employees and get the corresponding information. Finally, you access the .count attribute on Employee to get the current number of employees.

It’s important to note that you can access instance attributes like .name or .position using the dot notation on the target instance:

Instance attributes are specific to one instance, so you can’t access them through the class. In contrast, class attributes are common to the class and all its instances:

To access a class attribute, you can either use an instance or the class itself. However, to change a class attribute directly, you need to use the class. Go ahead and give it a try with the following example:

In this example, you try to update the count attribute using an instance, john. This action results in you attaching a new instance attribute to john instead of updating the value of Employee.count. To update the class attribute, you need to use the class itself.

Deleting Variables From Their Scope

In Python, you can explicitly remove variables or, more generically, names from a given scope using the del statement:

In this code snippet, you first create a new variable called city in your current global scope. Then, you use the del statement to remove the variable from its containing scope. When you try to access the variable again, you get a NameError exception.

You should know that while del removes the reference to an object, it doesn’t necessarily free the memory immediately. Python’s garbage collector claims the memory once there are no more references to the object.