Expressions, and decoding code

Expressions, and decoding code#

The expressions page introduced the idea of an expression as:

a recipe that returns a value.

When Python evaluates an expression, it follows the recipe to return the value.

In this page we go into a little more detail on expressions, and we consider sub-expressions.

We do this to get used to the more general idea in programming, of building complex things from simple building blocks, according to simple rules.

A large part of learning how to program is understanding what the computer is doing as a result of the code you type. You can do this by looking at the code, in order to break it down into simpler building blocks. This will help you understand how the parts are combined.

We call this process decoding the code.

Simple building blocks#

Let’s start with some simple building blocks. Here is a simple expression in Python.

10.50

10.5

Yes, the text you typed - 10.50 - is a simple expression in Python.

When you run this cell, Python analyzes the text 10.50 - and detects that this is an expression.

In fact the expression is what Python terms a “float literal”, meaning that the text 10.50 should be evaluated to a floating point number.

It evaluates the “float literal” expression, to give a value, which is its own internal representation of the floating point number 10.50.

Finally, because this is the last and only line in this cell, Jupyter shows us a text representation of the value in the output part of the cell.

This is a general feature of Jupyter - if the last line of the cell is an expression, Jupyter shows us a representation of the value that results from the expression.

Compound expressions#

We can build up more complex expressions, by combining simple expressions. For example, this expression consists of two simpler expressions (the “float literals” 10.50 and 9.25) combined with addition +. An expression that combines simple expressions is a compound expression.

# Combining two simple expressions into a compound expression.
# Jupyter shows us the result, when we execute the cell.
10.50 + 9.25

19.75

As you will see, we can also say that the simple expressions 10.50 and 9.25 are sub-expressions of the compound expression 10.50 + 9.25.

Decoding the code#

Now consider this compound expression:

(10.50 + 9.25) * 0.15

2.9625

Let us work out the process that Python will go through, when evaluating this expression.

To work out what it will do, we need to consider the rules of precedence. One helpful mnemonic for operator precedence is BODMAS, referring to these operations:

  • Brackets

  • Order

  • Division / Multiplication

  • Addition / Subtraction

“Order” refers to powers, as in \(2^3\).

Python first goes through and works out what is in the expression. Here’s the expression again:

(10.50 + 9.25) * 0.15

Python sees:

  • A left bracket (.

  • The “float literal” 10.50

  • The addition operator +.

  • The “float literal” 10.50

  • A right bracket ).

  • The multiplication operator +.

  • The “float literal” 0.15

Python knows BODMAS; it knows it should start with things between brackets. So, it starts by evaluating the part of the expression between brackets. This involves:

  1. Evaluating the expression 10.50 to give the computer representation (CR) of 10.50;

  2. Evaluating the expression 9.25 to give the CR of 10.50;

  3. Evaluating the compound expression 10.50 + 9.25 to give the CR of 19.75;

Now it has evaluated the expression between brackets down to its result - 19.75

  • its problem has reduced to the part outside the brackets. Python:

  1. Evaluates 0.15 to give the CR of

  2. Multiplies the result of the bracketed expression (19.75) by 0.15 to give the CR of the number 2.96250.

When we run the cell, Python does this work, and Jupyter shows us a text representation of the result:

(10.50 + 9.25) * 0.15

2.9625

Sub-expressions#

A compound expression is an expression made up of smaller sub-expressions.

A sub-expression is a smaller part of the compound expression, that Python will evaluate in the process of evaluating the compound expression.

We have already considered this compound expression:

10.50 + 9.25

19.75

The two sub-expressions are:

  • 10.50: an expression that returns the CR of the number 10.50;

  • 9.25: an expression that returns the CR of the number 9.25.

Here is the more complicated compound expression:

(10.50 + 9.25) * 0.15

2.9625

There are four sub-expressions here:

  • 10.50 as above;

  • 9.25 as above;

  • 0.15 returns the CR of the number 0.15;

  • (10.50 + 9.25) is a compound expression that is itself a sub-expression of the full expression above. It returns the CR of the number 19.75.

Each of these sub-expressions will be evaluated in the process of evaluating (10.50 + 9.25) * 0.15.

What do you think? Is 9.25 * 0.15 a sub-expression in this compound expression? [1]

Now consider:

10.50 + 9.25 * 0.15

11.8875

Be careful - remember BODMAS.

Python will:

  • evaluate the expression 10.50 to gives the CR of 10.50;

  • evaluate the expression 9.25 to gives the CR of 10.50;

  • (because of the precedence rules) evaluate 9.25 * 1.15 to give the CR of 1.3875;

  • evaluate 10.50 * 1.3875 to give the CR of 11.8875.

So, these are the sub-expressions:

  • 10.50;

  • 9.25;

  • 0.15;

  • 9.25 * 0.15.

Note that 10.50 + 9.25 is not a sub-expression, because Python does not do this calculation in the course of evaluating the expression, because of BODMAS. The value that 9.25 is involved in is 9.25 * 1.3875, where 1.3875 is the value that comes back from 9.25 * 0.15 above.

Call subexpressions#

Let’s say I have imported my cos function (see the functions page):

# Get the cos function from the numpy library.
from numpy import cos

Now consider:

cos(0)

1.0

cos(0) is an expression, because it is a recipe that returns a value. In fact, it is a call expression.

There are two sub-expressions to this expression. They are:

  • cos returns the CR of the recipe (function), and

  • 0 returns the CR of 0.

It may not have been obvious that cos on its own (without the parentheses) is an expression. cos is just the name for a thing — where the thing is the CR of the recipe. Here you can see that cos on its own is an expression, because, as for any expression, Jupyter shows us the value of the expression, when the expression is the last line of code in the cell:

# Show the value of the "cos" expression.
cos

<ufunc 'cos'>

Think of the name “cos” as an expression that refers to a value, where the value is the CR of the recipe.

Call expressions are often sub-expressions of compound expressions:

cos(0) + 2

3.0

This has four sub-expressions:

  • cos returns the CR of the recipe (function);

  • 0 returns the CR of 0;

  • cos(0) applies the recipe to 0, and returns the CR of the number 1;

  • 2 returns the CR of 2.

Variables in expressions#

Finally, let’s consider variables.

a = 10.50

Here is an expression:

a

10.5

It is an expression because it is a recipe that returns a value - in this case, the CR of the number 10.50. We can see it returns a value because we see a text representation of the value, when Jupyter executes this cell.

We could also define the variable b:

b = 9.25

b

9.25

Here is another compound expression:

(a + b) * 0.15

2.9625

Python:

  • Evaluates the expression a to get the CR of 10.50.

  • Evaluates the expression b to get the CR of 9.25.

  • Evaluates the expression (a + b) to get the CR of 19.75. [2]

  • Evaluates the expression 0.15 to get the CR of 0.15.

  • Evaluates the result of the bracketed expression multiplied by 0.15, to give the CR of 2.9625

Therefore the expression above has four sub-expressions:

  • a

  • b

  • 0.15

  • (a + b)

Contents