Python Data Structures: Dictionaries

Dictionaries

A dictionary is a map between key-value pairs.

For example:

A house can have 3 bedrooms.
‘bedrooms’ is the key, and the number of bedrooms is the value.

How does this look in terms of a data structure?

house = {'bedrooms': 3}
house

{'bedrooms': 3}

Dictionaries are used in different languages to look up definitions of words.

Python has a data structure by the same name that replicates this “lookup” action.

For example:

A house can have 3 bedrooms. Bedroom is called the key, and the number of bedrooms is called the value.

How does this look in terms of a data structure?

We use curly brackets and a colon that separates the key and its value.

house = {'bedrooms': 3, 'bathrooms': 2, 'city': 'Vancouver', 'price': 2499999, 'date_sold': (1,3,2015)}
house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 2499999,
 'date_sold': (1, 3, 2015)}

condo = {'bedrooms' : 2, 
         'bathrooms': 1, 
         'kitchens' : 1, 
         'city'     : 'Burnaby', 
         'price'    : 699999, 
         'date_sold': (27,8,2011)
        }
condo

{'bedrooms': 2,
 'bathrooms': 1,
 'kitchens': 1,
 'city': 'Burnaby',
 'price': 699999,
 'date_sold': (27, 8, 2011)}

This works with many different data types for the keys and values.

Here, we see the keys being of type str and the values being of type int, str, and even tuples.

The keys, which are the elements on the left of the colon, are unique and cannot be duplicated since they are used to look up the value. The values, which are the elements on the right of the colon, are not unique.

We can see that in our condo dictionary where the keys bathrooms and kitchen both have values of 1.

house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 2499999,
 'date_sold': (1, 3, 2015)}

house['price']

2499999

house['price'] = 7
house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 7,
 'date_sold': (1, 3, 2015)}

We can access, or look up, a value of a dictionary with the key in square brackets.

In this example, we obtain the value corresponding to the price key.

And since dictionaries are mutable, we can change the values. Let’s change the existing price to 7. That’s a bargain!

house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 7,
 'date_sold': (1, 3, 2015)}

house['bed monster'] = True
house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 7,
 'date_sold': (1, 3, 2015),
 'bed monster': True}

We can add to the dictionary in the same way as we edit them, but using a new key name.

Here we add to the house dictionary a new key that represents whether or not a bed monster exists. We are giving this a value of True.

Keys are not limited to strings and can be many different data types, including numerical values and tuples (but they cannot be dictionaries or lists).

Values can be of any type.

Here you can see that we have a key set as 9999 with a value of type list ['age', 'old'].

You have a lot of possibilities when making dictionaries.

house[9999] = ['age', 'old']
house[('trees', 'flower', 'vegetables')] = {'Garden': 3}
house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 7,
 'date_sold': (1, 3, 2015),
 'bed monster': True,
 9999: ['age', 'old'],
 ('trees', 'flower', 'vegetables'): {'Garden': 3}}

We can add to the dictionary in the same way as we edit them, but using a new key name.

Here we add to the house dictionary a new key that represents whether or not a bed monster exists. We are giving this a value of True.

Keys are not limited to strings and can be many different data types, including numerical values and tuples (but they cannot be dictionaries or lists).

Values can be of any type.

Here you can see that we have a key set as 9999 with a value of type list ['age', 'old'].

You have a lot of possibilities when making dictionaries.

house

{'bedrooms': 3,
 'bathrooms': 2,
 'city': 'Vancouver',
 'price': 7,
 'date_sold': (1, 3, 2015),
 'bed monster': True,
 9999: ['age', 'old'],
 ('trees', 'flower', 'vegetables'): {'Garden': 3}}

house.items()

dict_items([('bedrooms', 3), ('bathrooms', 2), ('city', 'Vancouver'), ('price', 7), ('date_sold', (1, 3, 2015)), ('bed monster', True), (9999, ['age', 'old']), (('trees', 'flower', 'vegetables'), {'Garden': 3})])

We can access all of the key-value pairs in a dictionary with the verb .items().

Let’s try it on our house dictionary.

Now we see tuples contained in a list, where each tuple contains the key and its corresponding value.

Dictionaries to Dataframes

	name	height	diameter	flowering
0	Cherry	7	12	True
1	Oak	20	89	False
2	Willow	12	30	True
3	Fir	16	18	False

data = { 'name': ['Cherry', 'Oak', 'Willow', 'Fir'], 
         'height': [7, 20, 12, 16], 
         'diameter': [12, 89, 30, 18], 
         'flowering': [True, False, True, False]}
         
forest = pd.DataFrame.from_dict(data)
forest

	name	height	diameter	flowering
0	Cherry	7	12	True
1	Oak	20	89	False
2	Willow	12	30	True
3	Fir	16	18	False

What about making dataframes from dictionaries?

We are lucky enough to have two ways of making data from a dictionary using the verb pd.DataFrame.from_dict().

First, let’s try making this table into a dataframe where we insert our data column-wise.

We can use the dictionary keys to represent the column names and the dictionary values for the column values stored in a list.

	name	height	diameter	flowering
0	Cherry	7	12	True
1	Oak	20	89	False
2	Willow	12	30	True
3	Fir	16	18	False

data = {0: ['Cherry', 7, 12, True],
        1: ['Oak', 20, 89, False],
        2: ['Willow', 12, 30, True],
        3: ['Fir', 16, 18, False]}
column_names = ['name', 'height', 'diameter', 'flowering']

forest = pd.DataFrame.from_dict(data, orient='index', columns= column_names)
forest

	name	height	diameter	flowering
0	Cherry	7	12	True
1	Oak	20	89	False
2	Willow	12	30	True
3	Fir	16	18	False

The other option is to create the dataframe row-wise.

In this case, we assign each row index as a dictionary key and the row values as the dictionary’s values.

We use the argument orient to explain the keys are the index and the argument columns to label our columns.

Let’s add what we learned to our table

Data Structure	preserves order	Mutable	Symbol	Can contain duplicates
str	✓	☓	'' or ""	✓
list	✓	✓	[]	✓
tuple	✓	☓	()	✓
set	☓	✓	{}	☓
dict	✓	✓	{:}	keys: ☓ , values: ✓

Let’s add the dictionary data structure to our data structure summary table.

Let’s apply what we learned!