按每个元组的第一个元素对元组列表进行排序-Python 2.7

Question

I have a list of tuples:

lst = [('Q4-2005', 327.93), ('Q1-2005', 133.05), ('Q3-2005', 500.95), ('Q2-2005', 254.22)]

I want to sort this list by the first element of each tuple. Thus my resulting, sorted list should look like:

[('Q1-2005', 133.05), ('Q2-2005', 254.22), ('Q3-2005', 500.95), ('Q4-2005', 327.93)]

I tried doing this using sorting() and lambda, but the resulting list is not sorted. Im thinking its because my 'dates' are actually strings:

sorted(lst, key = lambda x: x[0])

So i guess I have 2 questions..

First, How do I sort the list of tuples so that they are in chronological order?

Second, what is the best way to make python realize that 'Q1-2005' is a date? Eventually I want to plot this data where x-axis is the 'date' and y-axis is the number associated with each date?

Answer 1

You said that you want Python to interpret Q1-2005 as a date. If you convert the list to such a format, then sorting becomes trivial, and plotting will also be easier. Here's one way to do it using datetime.date (it encodes the quarters as a date representing the first date of the quarter).

from datetime import date

# date() takes in year, month, day args
date_lst = [(date(int(q[3:]), 3 * int(q[1]) - 2, 1), v)
            for q, v in lst]

This will result in something like this:

[(datetime.date(2005, 10, 1), 327.93),
 (datetime.date(2005, 1, 1), 133.05),
 (datetime.date(2005, 7, 1), 500.95),
 (datetime.date(2005, 4, 1), 254.22)]

After that, sorting chronologically is as simple as sorted(date_lst) .

Answer 2

The best way to do this, redefining a sort, would be to make a new class or a new function. The default sort algorithms in Python use traditional lexicographical sorting. Calling sort(tuple) will sort them in alphabetical/alphanumeric order, but not necessarily the order you want.

I will note, sorting by default will sort the tuples based on their index [0] value, only going to index [1] if two first indices are the same.

Answer 3

不要命名您的变量列表！！！

sorted(the_list, key=lambda x: list(reversed(map(int,x[0][1:].split("-")))))

Answer 4

Using an object with __lt__ and/or __cmp__ would be good, but if you want to use a key function:

#!/usr/local/cpython-3.3/bin/python

list_ = [('Q4-2005', 327.93), ('Q1-2005', 133.05), ('Q3-2005', 500.95), ('Q2-2005', 254.22), ('Q1-2006', 123.45), ('Q2-2007', 678.90), ]

def key(element):
    subresult1 = element[0]
    subresult2 = subresult1.split('-')
    subresult2.reverse()
    subresult2[0] = int(subresult2[0])
    return subresult2

list_.sort(key=key)
print(list_)

Which out for the years - they're more significant than the quarters.