Sorting Dictionaries for Fun and Profit
Python dictionaries don’t have an inherent order. You can iterate over them just fine but there’s no guarantee that iteration returns the dictionary’s elements in any particular order (although this is changing with Python 3.6).
意思就是字典是乱序的
However, it’s frequently useful to get a sorted representation of a dictionary to put the dictionary’s items into an arbitrary order based on their key, value, or some other derived property. Suppose you have a dictionary xs with the following key/value pairs:
>>> xs = {'a': 4, 'c': 2, 'b': 3, 'd': 1}
To get a sorted list of the key/value pairs in this dictionary, you could use the dictionary’s items() method and then sort the resulting sequence in a second pass:
>>> sorted(xs.items())
[('a', 4), ('b', 3), ('c', 2), ('d', 1)]
The key/value tuples are ordered using Python’s standard lexicographical ordering for comparing sequences.
To compare two tuples, Python compares the items stored at index zero first. If they differ, this defines the outcome of the comparison. If they’re equal, the next two items at index one are compared, and so on.
Now, because we took these tuples from a dictionary, all of the former dictionary keys at index zero in each tuple are unique. Therefore, there are no ties to break here.
在比较元组的时候我们是从序列的第一项开始进行比较的，如果第一项就不同那么我们就得出了相应的结果，如果第一项相同那么我们就比较第二项，以此类推。但是字典因为键之间各不相同，所以我们直接在第一项就可以区分出来。
In some cases a lexicographical ordering might be exactly what you want. In other cases you might want to sort a dictionary by value instead.
也许你想按照值来对字典的内容进行排序。
Luckily, there’s a way you can get complete control over how items are ordered. You can control the ordering by passing a key func to sorted() that will change how dictionary items are compared.
A key func is simply a normal Python function to be called on each element prior to making comparisons. The key func gets a dictionary item as its input and returns the desired “key” for the sort order comparisons.
Unfortunately, the word “key” is used in two contexts simultaneously here—the key func doesn’t deal with dictionary keys, it merely maps each input item to an arbitrary comparison key.
Now, maybe we should look at an example. Trust me, key funcs will be much easier to understand once you see some real code.
Let’s say you wanted to get a sorted representation of a dictionary based on its values. To get this result you could use the following key func which returns the value of each key/value pair by looking up the second element in the tuple:
>>> sorted(xs.items(), key=lambda x: x[1])
[('d', 1), ('c', 2), ('b', 3), ('a', 4)]
See how the resulting list of key/value pairs is now sorted by the values stored in the original dictionary? It’s worth spending some time wrapping your head around how key funcs work. It’s a powerful concept that you can apply in all kinds of Python contexts.
就是我们根据我们的需要对sorted的key值进行设定，这样我们就可以根据自己的需要对值进行排序。
In fact, the concept is so common that Python’s standard library includes the operator module. This module implements some of the most frequently used key funcs as plug-and-play building blocks, like operator.itemgetter and operator.attrgetter.
Here’s an example of how you might replace the lambda-based index lookup in the first example with operator.itemgetter:
>>> import operator
>>> sorted(xs.items(), key=operator.itemgetter(1))
[('d', 1), ('c', 2), ('b', 3), ('a', 4)]
Using the operator module might communicate your code’s intent more clearly in some cases. On the other hand, using a simple lambda expression might be just as readable and more explicit. In this particular case, I actually prefer the lambda expression.
Another benefit of using lambdas as a custom key func is that you get to control the sort order in much finer detail. For example, you could sort a dictionary based on the absolute numeric value of each value stored in it:
>>> sorted(xs.items(), key=lambda x: abs(x[1]))
If you need to reverse the sort order so that larger values go first, you can use the reverse=True keyword argument when calling sorted():
>>> sorted(xs.items(), key=lambda x: x[1],reverse=True)
[('a', 4), ('b', 3), ('c', 2), ('d', 1)]
Like I said earlier, it’s totally worth spending some time getting a good grip on how key funcs work in Python. They provide you with a ton of flexibility and can often save you from writing code to transform one data structure into another.
Key Takeaways
- When creating sorted “views” of dictionaries and other collections, you can influence the sort order with a key func.
- Key funcs are an important concept in Python. The most frequently used ones were even added to the operator module in the standard library.
- Functions are first-class citizens in Python. This is a powerful feature you’ll find used everywhere in the language.
