Question

In this part of the Python programming tutorial, we will cover Python dictionaries in more detail.

Python dictionary is a container of key-value pairs. It is mutable and can contain mixed types. A dictionary is an unordered collection. Python dictionaries are called associative arrays or hash tables in other languages. The keys in a dictionary must be immutable objects like strings or numbers. They must also be unique within a dictionary.

Creating dictionaries

First, we will show how to create Python dictionaries.

#!/usr/bin/python

weekend = { "Sun": "Sunday", "Mon": "Monday" }
vals = dict(one=1, two=2)

capitals = {}
capitals["svk"] = "Bratislava"
capitals["deu"] = "Berlin"
capitals["dnk"] = "Copenhagen"

d = { i: object() for i in range(4) }

print weekend
print vals
print capitals
print d

In the example, we create four dictionaries. In four different ways. Later we print the contents of these dictionaries to the console.

weekend = { "Sun": "Sunday", "Mon": "Monday" }

We create a weekend dictionary using dictionary literal notation. The key-value pairs are enclosed by curly brackets. The pairs are separated by commas. The first value of a pair is a key, which is followed by a colon character and a value. The "Sun" string is a key and the "Sunday" string is a value.

vals = dict(one=1, two=2)

Dictionaries can be created using the dict() function.

capitals = {}
capitals["svk"] = "Bratislava"
capitals["deu"] = "Berlin"
capitals["dnk"] = "Copenhagen"

This is the third way. An empty capitals dictionary is created. Three pairs are added to the dictionary. The keys are inside the square brackets, the values are located on the right side of the assignment.

d = { i: object() for i in range(4) }

A dictionary is created using a dictionary comprehension. The comprehension has two parts. The first part is the i: object() expression, which is executed for each cycle of a loop. The second part is the for i in range(4) loop. The dictionary comprehension creates a dictionary having four pairs, where the keys are numbers 0, 1, 2, and 3 and the values are simple objects.

$ ./create_dict.py 
{'Sun': 'Sunday', 'Mon': 'Monday'}
{'two': 2, 'one': 1}
{'svk': 'Bratislava', 'dnk': 'Copenhagen', 'deu': 'Berlin'}
{0: <object object at 0xb76cb4a8>, 1: <object object at 0xb76cb4b0>, 
2: <object object at 0xb76cb4b8>, 3: <object object at 0xb76cb4c0>}

Basic operations

The following examples will show some basic operations with Python dictionaries.

#!/usr/bin/python

basket = { 'oranges': 12, 'pears': 5, 'apples': 4 }

basket['bananas'] = 5

print basket
print "There are %d various items in the basket" % len(basket)

print basket['apples']
basket['apples'] = 8
print basket['apples']

print basket.get('oranges', 'undefined')
print basket.get('cherries', 'undefined')

We have a basket with different fruits. We perform some operations on the basket dictionary.

basket = { 'oranges': 12, 'pears': 5, 'apples': 4 }

The basket dictionary is created. It has initially three key-value pairs.

basket['bananas'] = 5

A new pair is created. The 'bananas' string is a key, the 5 integer is the value.

print "There are %d various items in the basket" % len(basket)

The len() function gives the number of pairs in the dictionary.

print basket['apples']

The value of the 'apples' key is printed to the terminal.

basket['apples'] = 8

The value of the 'apples' key is modified. It is set to number 8.

print basket.get('oranges', 'undefined')

The get() method retrieves the value of a specified key. If there is no such a key, the second parameter of the method is returned.

print basket.get('cherries', 'undefined')

This line returns 'undefined'. There are no cherries in the basket.

$ ./basics.py
{'bananas': 5, 'pears': 5, 'oranges': 12, 'apples': 4}
There are 4 various items in the basket
4
8
12
undefined

Example output.

The next example will present two dictionary methods: the fromkeys() and the setdefault() method.

#!/usr/bin/python

basket = ('oranges', 'pears', 'apples', 'bananas')

fruits = {}.fromkeys(basket, 0)
print fruits

fruits['oranges'] = 12
fruits['pears'] = 8
fruits['apples'] = 4

print fruits.setdefault('oranges', 11)
print fruits.setdefault('kiwis', 11)

print fruits

The fromkeys() method creates a new dictionary from a list. The setdefault() method returns a value if a key is present. Otherwise it inserts a key with a specified default value and returns the value.

basket = ('oranges', 'pears', 'apples', 'bananas')

We have a list of strings. From this list a new dictionary will be constructed.

fruits = {}.fromkeys(basket, 0)

The fromkeys() method creates a new dictionary, where the list items will be the keys. Each key will be initiated to 0. Note that the fromkeys() method is a class method and needs the class name, which is {} in our case, to be called.

fruits['oranges'] = 12
fruits['pears'] = 8
fruits['apples'] = 4

Here we add some values to the fruits dictionary.

print fruits.setdefault('oranges', 11)
print fruits.setdefault('kiwis', 11)

The first line prints 12 to the terminal. The 'oranges' key exists in the dictionary. In such a case, the method returns the its value. In the second case, the key does not exist yet. A new pair 'kiwis': 11 is inserted to the dictionary. And value 11 is printed to the console.

$ ./fruits.py 
{'bananas': 0, 'pears': 0, 'oranges': 0, 'apples': 0}
12
11
{'kiwis': 11, 'bananas': 0, 'pears': 8, 'oranges': 12, 'apples': 4}

We receive this output, when we launch the fruits.py script.

The next code example will show, how to add two Python dictionaries.

#!/usr/bin/python

domains = { "de": "Germany", "sk": "Slovakia", "hu": "Hungary"}
domains2 = { "us": "United States", "no": "Norway" }

domains.update(domains2)

print domains

We have two dictionaries. They are joined with the update() method.

domains.update(domains2)

The domains2 dictionary is added to the domains dictionary with the update() method.

$ ./domains.py
{'sk': 'Slovakia', 'de': 'Germany', 'no': 'Norway', 
'us': 'United States', 'hu': 'Hungary'}

The result shows all values from both dictionaries.

Now we will show, how to remove a pair from a dictionary.

#!/usr/bin/python

items = { "coins": 7, "pens": 3, "cups": 2, 
  "bags": 1, "bottles": 4, "books": 5 }

print items  

items.pop("coins")
print items

del items["bottles"]
print items

items.clear()
print items

The items dictionary has 6 key-value pairs. We will delete pairs from this dictionary.

items.pop("coins")

The pop() method removes a pair with a specified key.

del items["bottles"]

The del keyword deletes a "bottles": 4 pair from the items dictionary.

items.clear()

The clear() method clears all items from the dictionary.

$ ./removing.py 
{'bags': 1, 'pens': 3, 'coins': 7, 'books': 5, 'bottles': 4, 'cups': 2}
{'bags': 1, 'pens': 3, 'books': 5, 'bottles': 4, 'cups': 2}
{'bags': 1, 'pens': 3, 'books': 5, 'cups': 2}
{}

This is the example output.

Keys and values

A Python dictionary consists of key-value pairs. The keys() method returns a list of keys from a dictionary. The values() method creates a list of values. And the items() method returns a list of key-value tuples.

#!/usr/bin/python

domains = { "de": "Germany", "sk": "Slovakia", "hu": "Hungary",
  "us": "United States", "no": "Norway"  }

print domains.keys()
print domains.values()
print domains.items()

print "de" in domains
print "cz" in domains

We demonstrate the above mentioned methods. We also check if a key is present with the in keyword.

print domains.keys()

We print the list of keys of a domains dictionary with the keys() method.

print domains.values()

We print the list of values of a domains dictionary with the values() method.

print domains.items()

And finally, we print the list of key-value tuples of a domains dictionary using the items() method.

print "de" in domains
print "cz" in domains

With the in keyword, we check if the "de", "cz" keys are present in the domains dictionary. The return value is either True or False.

$ ./keys_values.py
['sk', 'de', 'no', 'us', 'hu']
['Slovakia', 'Germany', 'Norway', 'United States', 'Hungary']
[('sk', 'Slovakia'), ('de', 'Germany'), ('no', 'Norway'), 
('us', 'United States'), ('hu', 'Hungary')]
True
False

Output of the example.

Looping

Looping through the dictionary is a common programming job. This can be done with the for keyword.

#!/usr/bin/python

domains = { "de": "Germany", "sk": "Slovakia", "hu": "Hungary",
  "us": "United States", "no": "Norway"  }

for key in domains:
  print key
  
for k in domains:
  print domains[k]
  
for k, v in domains.items():
  print ": ".join((k, v))

In the example, we traverse the domains dictionary to print the keys, values and both keys and values of the dictionary.

for key in domains:
  print key

This loop prints all the keys of the dictionary.

for k in domains:
  print domains[k]

The second loop prints all values of the dictionary.

for k, v in domains.items():
  print ": ".join((k, v))

In the final loop, all keys and values are printed.

$ ./looping.py
sk
de
no
us
hu
Slovakia
Germany
Norway
United States
Hungary
sk: Slovakia
de: Germany
no: Norway
us: United States
hu: Hungary

Output of the example.

Sorting

Python dictionaries are orderless. This also implies that they cannot be sorted like a Python list. Programmers can create sorted representations of Python dictionaries. In this section, we will show several ways to create a sorted output.

Programmers might want to sort the data in a normal or reverse order. They could sort the data by keys or by values.

#!/usr/bin/python

items = { "coins": 7, "pens": 3, "cups": 2, 
  "bags": 1, "bottles": 4, "books": 5 }
  
kitems = items.keys()
kitems.sort()

for k in kitems:
  print ": ".join((k, str(items[k])))

The first example provides the simplest solution to have the data sorted by the keys.

kitems = items.keys()
kitems.sort()

A list of keys is obtained from the dictionary. The list is sorted with the sort() method.

for k in kitems:
  print ": ".join((k, str(items[k])))

In the loop we print the sorted keys together with their values from the dictionary.

$ ./simplesort.py
bags: 1
books: 5
bottles: 4
coins: 7
cups: 2
pens: 3

The items dictionary is sorted by its keys.

More efficient sorting can be done with the built-in sorted() function.

#!/usr/bin/python

items = { "coins": 7, "pens": 3, "cups": 2, 
  "bags": 1, "bottles": 4, "books": 5 }
  
for key in sorted(items.iterkeys()):
  print "%s: %s" % (key, items[key])

print "####### #######"  
  
for key in sorted(items.iterkeys(), reverse=True):
  print "%s: %s" % (key, items[key])

In the example we print sorted data by their keys in ascending and descending order using the sorted() function.

for key in sorted(items.iterkeys()):
  print "%s: %s" % (key, items[key])

In this for loop, we print the pairs sorted in ascending order. The iteritems() function returns an iterator over the dictionary’s (key, value) pairs.

for key in sorted(items.iterkeys(), reverse=True):
  print "%s: %s" % (key, items[key])

In the second for loop, the data is sorted in descending order. The order type is controlled by the reverse parameter.

$ ./sorting.py
bags: 1
books: 5
bottles: 4
coins: 7
cups: 2
pens: 3
####### #######
pens: 3
cups: 2
coins: 7
bottles: 4
books: 5
bags: 1

Output of the sorting.py script.

In the next example, we are going to sort the items by their values.

#!/usr/bin/python

items = { "coins": 7, "pens": 3, "cups": 2, 
  "bags": 1, "bottles": 4, "books": 5 }
  
for key, value in sorted(items.iteritems(), 
  key=lambda (k,v): (v,k)):
    
  print "%s: %s" % (key, value) 

print "####### #######"  
  
for key, value in sorted(items.iteritems(), 
  key=lambda (k,v): (v,k), reverse=True):
     
  print "%s: %s" % (key, value)  

The example prints the data in ascending and descending order by their values.

for key, value in sorted(items.iteritems(), 
  key=lambda (k,v): (v,k)):

Dictionary pairs are sorted by their values and printed to the console. The key parameter takes a function, which indicates, how the data is going to be sorted.

$ ./sorting2.py
bags: 1
cups: 2
pens: 3
bottles: 4
books: 5
coins: 7
####### #######
coins: 7
books: 5
bottles: 4
pens: 3
cups: 2
bags: 1

From the output we can see that this time the pairs were sorted according to their values.

Views

Python 2.7 introduced dictionary view objects. Views provide a dynamic view on the items of a dictionary. They bear similarity to SQL views. When the dictionary changes, the view reflects these changes. The dict.viewkeys() , dict.viewvalues() and dict.viewitems() methods return view objects.

A view is a virtual read-only container. A view does not make a copy of a dictionary.

#!/usr/bin/python

fruits = { 'oranges': 12, 'pears': 5, 'apples': 4, 'bananas': 4 }

vi = fruits.viewitems()
vv = fruits.viewvalues()
vk = fruits.viewkeys()

for k, v in vi:
  print k, v

for v in vv:
  print v
  
for k in vk:
  print k

Three view objects of the dictionary's items, dictionary's keys and dictionary's values are created. We traverse the view with the for loops.

vi = fruits.viewitems()

The viewitems() creates a view of the dictionary's items.

for k, v in vi:
  print k, v

We traverse the created view and print the keys and values in the for loop.

$ ./views.py
bananas 4
pears 5
oranges 12
apples 4
4
5
12
4
bananas
pears
oranges
apples

Output of the views.py script.

In the next example we show that a view reflects dictionary changes.

#!/usr/bin/python

fruits = { 'oranges': 12, 'pears': 5, 'apples': 4, 'bananas': 4}

vi = fruits.viewitems()

for k, v in vi:
  print k, v

fruits.pop('apples')
fruits.pop('oranges')

print "########### ##########"

for k, v in vi:
  print k, v

A view is created on the fruits dictionary. Two items are deleted from the dictionary. Then we traverse the view to see if the changes are reflected.

vi = fruits.viewitems()

A view is created on the fruits dictionary.

fruits.pop('apples')
fruits.pop('oranges')

Two items are deleted with the pop() method.

for k, v in vi:
  print k, v

We loop through the view of the fruits.

$ ./views2.py
bananas 4
pears 5
oranges 12
apples 4
########### ##########
bananas 4
pears 5

From the output we can see that the changes were reflected in the view.

In this part of the Python tutorial, we have written about Python dictionaries.