Functional way to compare things inside a list

[thor...@lavabit.com]

Hi,

list = [{'1': []}, {'2': []}, {'3': ['4', '5']}]

I want to check for a value (e.g. '4'), and get the key of the dictionary
that contains that value.
(Yep, this is bizarre.)

some_magic(list, '4')
=> '3'

What's the functional way to do it?
Is it possible to do it with a one-liner?

[Chris Angelico]

I'm thinking here of a list comprehension, filter(), and next() to
grab the first element. Let's see...

By the way, I wouldn't use 'list' as a variable name; you shadow the
built-in type.

lst = [{'1': []}, {'2': []}, {'3': ['4', '5']}]
def find_n(n,dic):
for key,searchme in dic.items():
if n in searchme: return key

next(filter(None,[find_n('4',x) for x in lst]))

That gets the result, but probably not in the cleanest way. I'm not
sure off-hand if Python has a convenient way to curry a function, but
if so, you could make the filter call rather simpler.

Note that this is written for Python 3, where filter() returns an
iterator, thus the algorithm is lazy and thus efficient (a very
Australian way to do things).

ChrisA
Proudly Australian, proudly lazy!

[Chris Rebert]

On Fri, Sep 21, 2012 at 1:23 AM, Chris Angelico <ros...@gmail.com> wrote:
> On Fri, Sep 21, 2012 at 8:58 AM, <thor...@lavabit.com> wrote:

<snip>

> That gets the result, but probably not in the cleanest way. I'm not
> sure off-hand if Python has a convenient way to curry a function,

http://docs.python.org/library/functools.html#functools.partial

Cheers,
Chris of the Northern Hemisphere

[Ivan@work]

Yes:

[key for d in list for key in d if '4' in d[key]]

[Chris Angelico]

On Fri, Sep 21, 2012 at 6:28 PM, Chris Rebert <cl...@rebertia.com> wrote:
> On Fri, Sep 21, 2012 at 1:23 AM, Chris Angelico <ros...@gmail.com> wrote:
>> On Fri, Sep 21, 2012 at 8:58 AM, <thor...@lavabit.com> wrote:
> <snip>
>> That gets the result, but probably not in the cleanest way. I'm not
>> sure off-hand if Python has a convenient way to curry a function,
>
> http://docs.python.org/library/functools.html#functools.partial

Thanks, that'd be it; didn't come up in a quick search for 'curry'
(for obvious reason). So it would probably be possible to do it as a
one-liner with a honking big lambda, but if code clarity is what you
want, I'd go with the externally-defined function.

> Cheers,
> Chris of the Northern Hemisphere

:)

ChrisA

[Alexander Blinne]

On 21.09.2012 00:58, thor...@lavabit.com wrote:

simple, but possibly slow solution:

import itertools

def some_magic(list, search):
return (key for key, val in itertools.chain(*(d.iteritems() for d in
list)) if search in val).next()

one-liner, yeah...

[Chris Rebert]

On Thu, Sep 20, 2012 at 3:58 PM, <thor...@lavabit.com> wrote:
> Hi,
>
> list = [{'1': []}, {'2': []}, {'3': ['4', '5']}]

Are the dictionaries each guaranteed to only contain a single
key-value pair? (Or is your example just simplistic?)

> I want to check for a value (e.g. '4'), and get the key of the dictionary
> that contains that value.

And what if there is no such dictionary? Or what if there are multiple
such dictionaries?

> (Yep, this is bizarre.)
>
> some_magic(list, '4')
> => '3'
>
> What's the functional way to do it?

Why do you care about the paradigm used?

> Is it possible to do it with a one-liner?

Who cares? It's possible to implement more complicated things in one
line of APL, but most people probably wouldn't recommend it.

Regards,
Chris R.

[Ulrich Eckhardt]

Am 21.09.2012 00:58, schrieb thor...@lavabit.com:
> list = [{'1': []}, {'2': []}, {'3': ['4', '5']}]
>
> I want to check for a value (e.g. '4'), and get the key of the dictionary
> that contains that value.

Note:
1. list is a built-in type, who's name is rebound above
2. The list above contains dictionaries that all only contain a single key?
3. You have strings containing decimal representations of numbers?

> (Yep, this is bizarre.)

The data are really stored in a strange way and you might be able to
make things clearer by reorganizing them a bit.

> some_magic(list, '4')
> => '3'
>
> What's the functional way to do it?

Functional as in functional programming and an emphasis on lazy
evaluation? In that case I'd write a generator that emits the keys where
the values contain the requested string.

> Is it possible to do it with a one-liner?

Yep, filter(), lambda and the 'in' operator. Question remains if this is
readable. Note that you can use a local function, too, if you just want
to reduce the scope/visibility.


Good luck!


Uli

[88888 Dihedral]

Ulrich Eckhardt於 2012年9月21日星期五UTC+8下午5時15分03秒寫道：

I don't think functional aspects are only marked as lazy
programming.

It just means when one is experimenting something
the efficient execution in speed is not on focus
yet.

[Ian Kelly]

On Fri, Sep 21, 2012 at 1:54 PM, 88888 Dihedral
<dihedr...@googlemail.com> wrote:
> I don't think functional aspects are only marked as lazy
> programming.

He wrote "lazy evaluation", not "lazy programming". Two entirely
different things.

> It just means when one is experimenting something
> the efficient execution in speed is not on focus
> yet.

No, what you're describing is a "prototype". It has nothing to do
with functional programming at all.

[88888 Dihedral]

A

Ian於 2012年9月22日星期六UTC+8上午4時50分49秒寫道：

A function with varaible arguments can be stored as a variable
to functions called decorators in python to return enhanced functions.

A function mapps a decorator to another decorator can be called
a decorator map or a decorator maker in python.

The closure level is guaranteed for decorators to be mapped by
multi-levels of decorator mappers trivially in python.

What do you want else for functional prgramming in python?

[88888 Dihedral]

A

Ian於 2012年9月22日星期六UTC+8上午4時50分49秒寫道：

[Steven D'Aprano]

On Fri, 21 Sep 2012 14:49:55 -0600, Ian Kelly wrote:

> On Fri, Sep 21, 2012 at 1:54 PM, 88888 Dihedral
> <dihedr...@googlemail.com> wrote:
>> I don't think functional aspects are only marked as lazy programming.
>
> He wrote "lazy evaluation", not "lazy programming". Two entirely
> different things.

For the record, the consensus here is that 88888 Dihedral is probably a
bot. It appears to be a pretty good bot, I haven't spotted it making any
egregious or obvious grammatical mistakes, but the semantics of its posts
don't seem quite human.

88888 Dihedral, if you're not a bot, you can go a long way towards
proving that by telling us what colour a purple elephant is.


--
Steven

[Ian Kelly]

On Fri, Sep 21, 2012 at 7:25 PM, Steven D'Aprano
<steve+comp....@pearwood.info> wrote:
> On Fri, 21 Sep 2012 14:49:55 -0600, Ian Kelly wrote:
>
>> On Fri, Sep 21, 2012 at 1:54 PM, 88888 Dihedral
>> <dihedr...@googlemail.com> wrote:
>>> I don't think functional aspects are only marked as lazy programming.
>>
>> He wrote "lazy evaluation", not "lazy programming". Two entirely
>> different things.
>
>
> For the record, the consensus here is that 88888 Dihedral is probably a
> bot. It appears to be a pretty good bot, I haven't spotted it making any
> egregious or obvious grammatical mistakes, but the semantics of its posts
> don't seem quite human.

I'm aware of that, although sometimes the posts seem coherent enough
that I think maybe it's not. Especially the ones where it posts
almost-working code snippets, complete with obvious typos.

Then it posts a complete non sequitur like the reply to my reply in
this thread, and the illusion is shattered.

[88888 Dihedral]

Ian於 2012年9月22日星期六UTC+8下午2時23分43秒寫道：

Do you want to use the lisp way for implementing functional programming?

[88888 Dihedral]

Ian於 2012年9月22日星期六UTC+8下午2時23分43秒寫道：

[Jamie Paul Griffin]

[ Ian Kelly wrote on Sat 22.Sep'12 at 0:22:43 -0600 ]

> On Fri, Sep 21, 2012 at 7:25 PM, Steven D'Aprano
> <steve+comp....@pearwood.info> wrote:

> > On Fri, 21 Sep 2012 14:49:55 -0600, Ian Kelly wrote:
> >
> >> On Fri, Sep 21, 2012 at 1:54 PM, 88888 Dihedral
> >> <dihedr...@googlemail.com> wrote:
> >>> I don't think functional aspects are only marked as lazy programming.
> >>
> >> He wrote "lazy evaluation", not "lazy programming". Two entirely
> >> different things.
> >
> >
> > For the record, the consensus here is that 88888 Dihedral is probably a
> > bot. It appears to be a pretty good bot, I haven't spotted it making any
> > egregious or obvious grammatical mistakes, but the semantics of its posts
> > don't seem quite human.
>

> I'm aware of that, although sometimes the posts seem coherent enough
> that I think maybe it's not. Especially the ones where it posts
> almost-working code snippets, complete with obvious typos.
>
> Then it posts a complete non sequitur like the reply to my reply in
> this thread, and the illusion is shattered.

I find this intriguing, I had no idea bots existed to post to mailing
lists in this way. What's the point of them?

[Andrew Berg]

On 2012.09.22 02:08, Jamie Paul Griffin wrote:
> I find this intriguing, I had no idea bots existed to post to mailing
> lists in this way. What's the point of them?

To amuse their owners is my guess.
--
CPython 3.3.0rc2 | Windows NT 6.1.7601.17835

[Ramchandra Apte]

The bot could be used for automatic replies for duplicate posts (common questions are quite often repeated)

[Ramchandra Apte]

On Saturday, 22 September 2012 12:50:08 UTC+5:30, Andrew Berg wrote: