Another built in python data structure is the dictionary. Dictionaries creating mappings between key value pairs like so:
dict1 = {keyX : valueX, keyC : valueC, keyQ : valueQ}.

you could also create the same dictionary like this:
dict1 = dict(keyX : valueX, keyC : valueC, keyQ : valueQ).

Unlike strings or lists, dictionaries do not have an index or a specific order. There are ways to iterate through a dictionary, but the order that the items are returned is not gauranteed to be in a specific order. Instead, dictionaries are good for finding the value of a specific item rather then an item in a specific place.

Let's look at a couple of examples in practice:

As usual, let's start by import pandas and some data that we would be working with.

import pandas as pd

df = pd.read_csv('lego_sets.csv')
df.head()

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

A common use case of dictionaries is to rename the values in a column. For example, let's say that we wanted to rename the review_difficulty naming convention to use a quantitative scale.

# Get previous values
df.review_difficulty.unique()

array(['Average', 'Easy', 'Challenging', 'Very Easy', nan,
       'Very Challenging'], dtype=object)

Notice the nan value above which represents null or blank values. We could potentially translate these difficulty ratings into a quantitative scale like this:

diff_dict = {'Very Easy' : 1, 'Easy' : 2, 'Average' : 3, 'Challenging' : 4, 'Very Challenging' : 5}

We could then create a new column (or update the current column) using that dictionary:

df['Difficulty_Rating'] = df.review_difficulty.map(diff_dict)
df.head() #Preview changes

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

You can also quickly create dictionaries from another dataset. Let's say we want the full name of countries listed under the country column.

df.country.value_counts()[:5]

US    817
CA    815
GB    576
NL    576
DN    575
Name: country, dtype: int64

#Pull in a new dataset from online
countries = pd.read_csv('Country_Codes.csv')
countries.head()

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

#Create a dictionary
#The zip method is a neat little tool for pairing each entry from two columns together (like a zipper!)
#We then just wrap that in the dict() function.
country_dict = dict(zip(countries['A2 (ISO)'], countries['COUNTRY'])) 

#Map it to our original dataset (you can also do this with pd.merge() for joining multiple fields
df['Country_Full_Name'] = df.country.map(country_dict)
df.head(2)

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

df.Country_Full_Name.value_counts()[:5]

United States     817
Canada            815
Netherlands       576
United Kingdom    576
Austria           575
Name: Country_Full_Name, dtype: int64

We can also use dictionaries along with the pandas groupby method to apply different aggregations to different columns:

import numpy as np

agg_dict = {'ages' : 'max',
            'Difficulty_Rating' : [np.mean, np.std],
           'Country_Full_Name' : lambda x: x.value_counts().index[0],
           'num_reviews' : ['mean', 'max']}
df.groupby('theme_name').agg(agg_dict)

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead tr th {
    text-align: left;
}

.dataframe thead tr:last-of-type th {
    text-align: right;
}

Create a dictionary that rebins the age column to the following age ranges: Under 5, 5-8, 8-12, 12+

*If there is a conflict in age bin, default to the higher age bin.

# Your code here

Create a bar graph depicting the number of lego sets for the original age range column. Then create a second bar graph for the new age column you created. How do they compare?

# Your code here