fmap(), "inverse" of Python map() function

[vasudevram]

http://jugad2.blogspot.in/2012/10/fmap-inverse-of-python-map-function.html

- Vasudev Ram
www.dancingbison.com
jugad2.blogspot.com
twitter.com/vasudevram

[Ian Kelly]

On Fri, Oct 5, 2012 at 2:19 PM, vasudevram <vasud...@gmail.com> wrote:
>
> http://jugad2.blogspot.in/2012/10/fmap-inverse-of-python-map-function.html

Your fmap is a special case of reduce.

def fmap(functions, argument):
return reduce(lambda result, func: func(result), functions, argument)

[Ian Kelly]

In a more functional style, you could also use reduce to compose the
functions before applying them:

def compose(f, g):
return lambda x: f(g(x))

def fmap(functions):
return reduce(compose, reversed(functions))

# Allowing you to then do:
result = fmap(functions)(argument)

Cheers,
Ian

[Devin Jeanpierre]

On Fri, Oct 5, 2012 at 5:31 PM, Ian Kelly <ian.g...@gmail.com> wrote:
> On Fri, Oct 5, 2012 at 2:19 PM, vasudevram <vasud...@gmail.com> wrote:
>>
>> http://jugad2.blogspot.in/2012/10/fmap-inverse-of-python-map-function.html
>
> Your fmap is a special case of reduce.

So is map.

def map(f, seq):
return reduce(
lambda rseq, newpre:
rseq.append(f(newpre)) or rseq,
seq,
[])

"X is a special case of reduce" is basically the same as saying "X can
be implemented using a for loop". If it's meant as a complaint, it's a
poor one.

-- Devin

[Ian Kelly]

On Fri, Oct 5, 2012 at 4:52 PM, Devin Jeanpierre <jeanpi...@gmail.com> wrote:
> On Fri, Oct 5, 2012 at 5:31 PM, Ian Kelly <ian.g...@gmail.com> wrote:
>> On Fri, Oct 5, 2012 at 2:19 PM, vasudevram <vasud...@gmail.com> wrote:
>>>
>>> http://jugad2.blogspot.in/2012/10/fmap-inverse-of-python-map-function.html
>>
>> Your fmap is a special case of reduce.
>
> So is map.

I realize that. My point is that the function *feels* more like a
variant of reduce than of map.

> If it's meant as a complaint, it's a poor one.

It's not.

[Devin Jeanpierre]

On Fri, Oct 5, 2012 at 7:24 PM, Ian Kelly <ian.g...@gmail.com> wrote:
> I realize that. My point is that the function *feels* more like a
> variant of reduce than of map.
>
>> If it's meant as a complaint, it's a poor one.
>
> It's not.

Fair enough all around. Sorry for misunderstanding.

-- Devin

[vasudevram]

On Saturday, October 6, 2012 5:01:40 AM UTC+5:30, Devin Jeanpierre wrote:

> On Fri, Oct 5, 2012 at 7:24 PM, Ian Kelly wrote:
>
> > I realize that. My point is that the function *feels* more like a
>
> > variant of reduce than of map.
>
> >
>
> >> If it's meant as a complaint, it's a poor one.
>
> >
>
> > It's not.
>
>
>
> Fair enough all around. Sorry for misunderstanding.
>
>
>
> -- Devin

Thanks to all who replied. Always good to learn something new.

- Vasudev

[vasudevram]

On Saturday, October 6, 2012 5:01:40 AM UTC+5:30, Devin Jeanpierre wrote:

> On Fri, Oct 5, 2012 at 7:24 PM, Ian Kelly wrote:
>
> > I realize that. My point is that the function *feels* more like a
>
> > variant of reduce than of map.
>
> >
>
> >> If it's meant as a complaint, it's a poor one.
>
> >
>
> > It's not.
>
>
>
> Fair enough all around. Sorry for misunderstanding.
>
>
>
> -- Devin

[vasudevram]

> Thanks to all who replied. Always good to learn something new.

P.S. A reader posted a good comment with Scala as well as Python code for a compose function (basically same functionality as fmap, or more - the compose once, run many times thing). It's the 4th comment on my blog post.

- Vasudev

[vasudevram]

> Thanks to all who replied. Always good to learn something new.