Path: csiph.com!news.swapon.de!eternal-september.org!feeder.eternal-september.org!mx02.eternal-september.org!.POSTED!not-for-mail
From: DFS <nospam@dfs.com>
Newsgroups: comp.lang.python
Subject: Re: Sorting a list
Date: Sun, 3 Apr 2016 16:08:42 -0400
Organization: A noiseless patient Spider
Lines: 77
Message-ID: <ndrt29$36l$1@dont-email.me>
References: <ndrn97$bco$1@dont-email.me> <mailman.408.1459711883.28225.python-list@python.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Date: Sun, 3 Apr 2016 20:05:29 -0000 (UTC)
Injection-Info: mx02.eternal-september.org; posting-host="5399b1c4ae39fc7dd2f40cbc5f70036e"; logging-data="3285"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1+2xwfDh23bNmlMlWzWdh5r"
User-Agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.7.0
In-Reply-To: <mailman.408.1459711883.28225.python-list@python.org>
Cancel-Lock: sha1:vFnVE0X1qE9ReNAEdYcMAVVIqdU=
Xref: csiph.com comp.lang.python:106402

On 4/3/2016 3:31 PM, Peter Otten wrote:
> DFS wrote:
>
>> cntText = 60
>> cntBool = 20
>> cntNbrs = 30
>> cntDate = 20
>> cntBins = 20
>>
>> strText = "  text:     "
>> strBool = "  boolean:  "
>> strNbrs = "  numeric:  "
>> strDate = "  date-time:"
>> strBins = "  binary:   "
>>
>> colCounts = [(cntText,strText) , (cntBool,strBool), (cntNbrs,strNbrs) ,
>> (cntDate,strDate) , (cntBins,strBins)]
>>
>> # sort by alpha, then by column type count descending
>> colCounts.sort(key=lambda x: x[1])
>> colCounts.sort(key=lambda x: x[0], reverse=True)
>> for key in colCounts: print key[1], key[0]]
>>
>> -------------------------------------------------
>>
>> Output (which is exactly what I want):
>>
>>     text:      60
>>     numeric:   30
>>     binary:    20
>>     boolean:   20
>>     date-time: 20
>>
>> -------------------------------------------------
>>
>>
>> But, is there a 1-line way to sort and print?
>
> Yes, but I would not recommend it. You can replace the sort() method
> invocations with nested calls of sorted() and instead of
>
> for item in items:
>      print convert_to_str(item)
>
> use
>
> print "\n".join(convert_to_str(item) for item in items)
>
> Putting it together:
>
>>>> from operator import itemgetter as get
>>>> print "\n".join("{1} {0}".format(*p) for p in sorted(
> ... sorted(colCounts, key=get(1)), key=get(0), reverse=True))

Kind of clunky looking.  Is that why don't you recommend it?




>    text:      60
>    numeric:   30
>    binary:    20
>    boolean:   20
>    date-time: 20
>
> You could also cheat and use
>
> lambda v: (-v[0], v[1])
>
> and a single sorted().

That works well.  Why is it 'cheating'?


Thanks for the reply.