Logical operators and named arguments

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

Logical operators and named arguments

Ryan-4
Hi,

I'm wondering why calling ">" with named arguments doesn't work as expected:

> args(">")
function (e1, e2)
NULL

> sapply(c(1,2,3), `>`, e2=0)
[1] TRUE TRUE TRUE

> sapply(c(1,2,3), `>`, e1=0)
[1] TRUE TRUE TRUE

Shouldn't the latter be FALSE?

Thanks for any help,
Ryan


The information in this e-mail is intended only for the ...{{dropped:11}}

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Logical operators and named arguments

Joshua Wiley-2
Hi Ryan,

It does work, but the *apply family of functions always pass to the first
argument, so you can specify e2 = , but not e1 =.  For example:

> sapply(1:3, `>`, e2 = 2)
[1] FALSE FALSE  TRUE

>From ?sapply

     'lapply' returns a list of the same length as 'X', each element of
     which is the result of applying 'FUN' to the corresponding element
     of 'X'.

so `>` is applied to each element of 1:3

`>`(1, ...)
`>`(2, ...)
`>`(3, ...)

and if e2 is specified than that is passed

`>`(1, 2)
`>`(2, 2)
`>`(3, 2)

Further, see ?Ops

   If the members of this group are called as functions, any
          argument names are removed to ensure that positional matching
          is always used.

and you can see this at work:

> `>`(e1 = 1, e2 = 2)
[1] FALSE
> `>`(e2 = 1, e1 = 2)
[1] FALSE

If you want to the flexibility to specify which argument the elements of X
should be *applied to, use a wrapper:

> sapply(1:3, function(x) `>`(x, 2))
[1] FALSE FALSE  TRUE
> sapply(1:3, function(x) `>`(2, x))
[1]  TRUE FALSE FALSE


HTH,

Josh



On Thu, Aug 7, 2014 at 2:20 PM, Ryan <[hidden email]> wrote:

> Hi,
>
> I'm wondering why calling ">" with named arguments doesn't work as
> expected:
>
> > args(">")
> function (e1, e2)
> NULL
>
> > sapply(c(1,2,3), `>`, e2=0)
> [1] TRUE TRUE TRUE
>
> > sapply(c(1,2,3), `>`, e1=0)
> [1] TRUE TRUE TRUE
>
> Shouldn't the latter be FALSE?
>
> Thanks for any help,
> Ryan
>
>
> The information in this e-mail is intended only for th...{{dropped:23}}

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Logical operators and named arguments

Ryan-4
Josh,

Thank you for your detailed answer.

Best,
Ryan

On 7 Aug 2014, at 16:21, Joshua Wiley wrote:

> Hi Ryan,
>
> It does work, but the *apply family of functions always pass to the first
> argument, so you can specify e2 = , but not e1 =.  For example:
>
>> sapply(1:3, `>`, e2 = 2)
> [1] FALSE FALSE  TRUE
>
> From ?sapply
>
>    'lapply' returns a list of the same length as 'X', each element of
>    which is the result of applying 'FUN' to the corresponding element
>    of 'X'.
>
> so `>` is applied to each element of 1:3
>
> `>`(1, ...)
> `>`(2, ...)
> `>`(3, ...)
>
> and if e2 is specified than that is passed
>
> `>`(1, 2)
> `>`(2, 2)
> `>`(3, 2)
>
> Further, see ?Ops
>
>  If the members of this group are called as functions, any
>         argument names are removed to ensure that positional matching
>         is always used.
>
> and you can see this at work:
>
>> `>`(e1 = 1, e2 = 2)
> [1] FALSE
>> `>`(e2 = 1, e1 = 2)
> [1] FALSE
>
> If you want to the flexibility to specify which argument the elements of X
> should be *applied to, use a wrapper:
>
>> sapply(1:3, function(x) `>`(x, 2))
> [1] FALSE FALSE  TRUE
>> sapply(1:3, function(x) `>`(2, x))
> [1]  TRUE FALSE FALSE
>
>
> HTH,
>
> Josh
>
>
>
> On Thu, Aug 7, 2014 at 2:20 PM, Ryan <[hidden email]> wrote:
>
>> Hi,
>>
>> I'm wondering why calling ">" with named arguments doesn't work as
>> expected:
>>
>>> args(">")
>> function (e1, e2)
>> NULL
>>
>>> sapply(c(1,2,3), `>`, e2=0)
>> [1] TRUE TRUE TRUE
>>
>>> sapply(c(1,2,3), `>`, e1=0)
>> [1] TRUE TRUE TRUE
>>
>> Shouldn't the latter be FALSE?
>>
>> Thanks for any help,
>> Ryan
>>
>>
>> The information in this e-mail is intended only for t...{{dropped:28}}

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Logical operators and named arguments

Patrick Burns
In reply to this post by Joshua Wiley-2
On 07/08/2014 07:21, Joshua Wiley wrote:
> Hi Ryan,
>
> It does work, but the *apply family of functions always pass to the first
> argument, so you can specify e2 = , but not e1 =.  For example:
>
>> sapply(1:3, `>`, e2 = 2)
> [1] FALSE FALSE  TRUE

That is not true:

gt <- function(x, y) x > y

 > sapply(1:3, gt, y=2)
[1] FALSE FALSE  TRUE
 > sapply(1:3, gt, x=2)
[1]  TRUE FALSE FALSE

Specifying the first argument(s) in an apply
call is a standard way of getting flexibility.

I'd hazard to guess that the reason the original
version doesn't work is because `>` is Primitive.
There's speed at the expense of not behaving quite
the same as typical functions.

Pat

>
>>From ?sapply
>
>       'lapply' returns a list of the same length as 'X', each element of
>       which is the result of applying 'FUN' to the corresponding element
>       of 'X'.
>
> so `>` is applied to each element of 1:3
>
> `>`(1, ...)
> `>`(2, ...)
> `>`(3, ...)
>
> and if e2 is specified than that is passed
>
> `>`(1, 2)
> `>`(2, 2)
> `>`(3, 2)
>
> Further, see ?Ops
>
>     If the members of this group are called as functions, any
>            argument names are removed to ensure that positional matching
>            is always used.
>
> and you can see this at work:
>
>> `>`(e1 = 1, e2 = 2)
> [1] FALSE
>> `>`(e2 = 1, e1 = 2)
> [1] FALSE
>
> If you want to the flexibility to specify which argument the elements of X
> should be *applied to, use a wrapper:
>
>> sapply(1:3, function(x) `>`(x, 2))
> [1] FALSE FALSE  TRUE
>> sapply(1:3, function(x) `>`(2, x))
> [1]  TRUE FALSE FALSE
>
>
> HTH,
>
> Josh
>
>
>
> On Thu, Aug 7, 2014 at 2:20 PM, Ryan <[hidden email]> wrote:
>
>> Hi,
>>
>> I'm wondering why calling ">" with named arguments doesn't work as
>> expected:
>>
>>> args(">")
>> function (e1, e2)
>> NULL
>>
>>> sapply(c(1,2,3), `>`, e2=0)
>> [1] TRUE TRUE TRUE
>>
>>> sapply(c(1,2,3), `>`, e1=0)
>> [1] TRUE TRUE TRUE
>>
>> Shouldn't the latter be FALSE?
>>
>> Thanks for any help,
>> Ryan
>>
>>
>> The information in this e-mail is intended only for th...{{dropped:23}}
>
> ______________________________________________
> [hidden email] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

--
Patrick Burns
[hidden email]
twitter: @burnsstat @portfolioprobe
http://www.portfolioprobe.com/blog
http://www.burns-stat.com
(home of:
  'Impatient R'
  'The R Inferno'
  'Tao Te Programming')

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Logical operators and named arguments

Joshua Wiley-2
On Sat, Aug 9, 2014 at 9:56 AM, Patrick Burns <[hidden email]>
wrote:

> On 07/08/2014 07:21, Joshua Wiley wrote:
>
>> Hi Ryan,
>>
>> It does work, but the *apply family of functions always pass to the first
>> argument, so you can specify e2 = , but not e1 =.  For example:
>>
>>  sapply(1:3, `>`, e2 = 2)
>>>
>> [1] FALSE FALSE  TRUE
>>
>
> That is not true:
>

But it is passed as the first argument, not by name, but positionally.  The
reason it works with your gt() is because R with regular functions is
flexible:

> f <- function(x, y) x > y
> f(1:3, x = 2)
[1]  TRUE FALSE FALSE

but primitives ARE positionally matched

> `>`(1:3, 2)
[1] FALSE FALSE  TRUE
> `>`(1:3, e1 = 2)
[1] FALSE FALSE  TRUE



>
> gt <- function(x, y) x > y
>
> > sapply(1:3, gt, y=2)
> [1] FALSE FALSE  TRUE
> > sapply(1:3, gt, x=2)
> [1]  TRUE FALSE FALSE
>
> Specifying the first argument(s) in an apply
> call is a standard way of getting flexibility.
>
> I'd hazard to guess that the reason the original
> version doesn't work is because `>` is Primitive.
> There's speed at the expense of not behaving quite
> the same as typical functions.
>
> Pat
>
>
>>  From ?sapply
>>>
>>
>>       'lapply' returns a list of the same length as 'X', each element of
>>       which is the result of applying 'FUN' to the corresponding element
>>       of 'X'.
>>
>> so `>` is applied to each element of 1:3
>>
>> `>`(1, ...)
>> `>`(2, ...)
>> `>`(3, ...)
>>
>> and if e2 is specified than that is passed
>>
>> `>`(1, 2)
>> `>`(2, 2)
>> `>`(3, 2)
>>
>> Further, see ?Ops
>>
>>     If the members of this group are called as functions, any
>>            argument names are removed to ensure that positional matching
>>            is always used.
>>
>> and you can see this at work:
>>
>>  `>`(e1 = 1, e2 = 2)
>>>
>> [1] FALSE
>>
>>> `>`(e2 = 1, e1 = 2)
>>>
>> [1] FALSE
>>
>> If you want to the flexibility to specify which argument the elements of X
>> should be *applied to, use a wrapper:
>>
>>  sapply(1:3, function(x) `>`(x, 2))
>>>
>> [1] FALSE FALSE  TRUE
>>
>>> sapply(1:3, function(x) `>`(2, x))
>>>
>> [1]  TRUE FALSE FALSE
>>
>>
>> HTH,
>>
>> Josh
>>
>>
>>
>> On Thu, Aug 7, 2014 at 2:20 PM, Ryan <[hidden email]> wrote:
>>
>>  Hi,
>>>
>>> I'm wondering why calling ">" with named arguments doesn't work as
>>> expected:
>>>
>>>  args(">")
>>>>
>>> function (e1, e2)
>>> NULL
>>>
>>>  sapply(c(1,2,3), `>`, e2=0)
>>>>
>>> [1] TRUE TRUE TRUE
>>>
>>>  sapply(c(1,2,3), `>`, e1=0)
>>>>
>>> [1] TRUE TRUE TRUE
>>>
>>> Shouldn't the latter be FALSE?
>>>
>>> Thanks for any help,
>>> Ryan
>>>
>>>
>>> The information in this e-mail is intended only for th...{{dropped:23}}
>>>
>>
>> ______________________________________________
>> [hidden email] mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/
>> posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>>
> --
> Patrick Burns
> [hidden email]
> twitter: @burnsstat @portfolioprobe
> http://www.portfolioprobe.com/blog
> http://www.burns-stat.com
> (home of:
>  'Impatient R'
>  'The R Inferno'
>  'Tao Te Programming')
>



--
Joshua F. Wiley
Ph.D. Student, UCLA Department of Psychology
http://joshuawiley.com/
Senior Analyst, Elkhart Group Ltd.
http://elkhartgroup.com
Office: 260.673.5518

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
Reply | Threaded
Open this post in threaded view
|

Re: Logical operators and named arguments

Prof Brian Ripley
On 09/08/2014 01:10, Joshua Wiley wrote:

> On Sat, Aug 9, 2014 at 9:56 AM, Patrick Burns <[hidden email]>
> wrote:
>
>> On 07/08/2014 07:21, Joshua Wiley wrote:
>>
>>> Hi Ryan,
>>>
>>> It does work, but the *apply family of functions always pass to the first
>>> argument, so you can specify e2 = , but not e1 =.  For example:
>>>
>>>   sapply(1:3, `>`, e2 = 2)
>>>>
>>> [1] FALSE FALSE  TRUE
>>>
>>
>> That is not true:
>>
>
> But it is passed as the first argument, not by name, but positionally.  The
> reason it works with your gt() is because R with regular functions is
> flexible:
>
>> f <- function(x, y) x > y
>> f(1:3, x = 2)
> [1]  TRUE FALSE FALSE
>
> but primitives ARE positionally matched

That's not true either.  Almost all primitives intended to be called as
functions do have standard argument-matching semantics.  (Once upon a
time they did not, but I added the requisite code years ago.)  There are
six exceptions plus binary operators and other language elements.

See
http://cran.r-project.org/doc/manuals/r-release/R-ints.html#g_t_002eInternal-vs-_002ePrimitive 
  and the comments about primitive functions in ?lapply.

>
>> `>`(1:3, 2)
> [1] FALSE FALSE  TRUE
>> `>`(1:3, e1 = 2)
> [1] FALSE FALSE  TRUE
>
>
>
>>
>> gt <- function(x, y) x > y
>>
>>> sapply(1:3, gt, y=2)
>> [1] FALSE FALSE  TRUE
>>> sapply(1:3, gt, x=2)
>> [1]  TRUE FALSE FALSE
>>
>> Specifying the first argument(s) in an apply
>> call is a standard way of getting flexibility.
>>
>> I'd hazard to guess that the reason the original
>> version doesn't work is because `>` is Primitive.
>> There's speed at the expense of not behaving quite
>> the same as typical functions.
>>
>> Pat
>>
>>
>>>   From ?sapply
>>>>
>>>
>>>        'lapply' returns a list of the same length as 'X', each element of
>>>        which is the result of applying 'FUN' to the corresponding element
>>>        of 'X'.
>>>
>>> so `>` is applied to each element of 1:3
>>>
>>> `>`(1, ...)
>>> `>`(2, ...)
>>> `>`(3, ...)
>>>
>>> and if e2 is specified than that is passed
>>>
>>> `>`(1, 2)
>>> `>`(2, 2)
>>> `>`(3, 2)
>>>
>>> Further, see ?Ops
>>>
>>>      If the members of this group are called as functions, any
>>>             argument names are removed to ensure that positional matching
>>>             is always used.
>>>
>>> and you can see this at work:
>>>
>>>   `>`(e1 = 1, e2 = 2)
>>>>
>>> [1] FALSE
>>>
>>>> `>`(e2 = 1, e1 = 2)
>>>>
>>> [1] FALSE
>>>
>>> If you want to the flexibility to specify which argument the elements of X
>>> should be *applied to, use a wrapper:
>>>
>>>   sapply(1:3, function(x) `>`(x, 2))
>>>>
>>> [1] FALSE FALSE  TRUE
>>>
>>>> sapply(1:3, function(x) `>`(2, x))
>>>>
>>> [1]  TRUE FALSE FALSE
>>>
>>>
>>> HTH,
>>>
>>> Josh
>>>
>>>
>>>
>>> On Thu, Aug 7, 2014 at 2:20 PM, Ryan <[hidden email]> wrote:
>>>
>>>   Hi,
>>>>
>>>> I'm wondering why calling ">" with named arguments doesn't work as
>>>> expected:
>>>>
>>>>   args(">")
>>>>>
>>>> function (e1, e2)
>>>> NULL
>>>>
>>>>   sapply(c(1,2,3), `>`, e2=0)
>>>>>
>>>> [1] TRUE TRUE TRUE
>>>>
>>>>   sapply(c(1,2,3), `>`, e1=0)
>>>>>
>>>> [1] TRUE TRUE TRUE
>>>>
>>>> Shouldn't the latter be FALSE?
>>>>
>>>> Thanks for any help,
>>>> Ryan
>>>>
>>>>
>>>> The information in this e-mail is intended only for th...{{dropped:23}}
>>>>
>>>
>>> ______________________________________________
>>> [hidden email] mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide http://www.R-project.org/
>>> posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>>
>>>
>> --
>> Patrick Burns
>> [hidden email]
>> twitter: @burnsstat @portfolioprobe
>> http://www.portfolioprobe.com/blog
>> http://www.burns-stat.com
>> (home of:
>>   'Impatient R'
>>   'The R Inferno'
>>   'Tao Te Programming')
>>
>
>
>


--
Brian D. Ripley,                  [hidden email]
Emeritus Professor of Applied Statistics, University of Oxford
1 South Parks Road, Oxford OX1 3TG, UK

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.