[R] Mystified - comparing chron times

[MacQueen, Don]

Not that anything more needs to be said ... it doesn't, not really ...

But I think that in cases like this it is helpful to use the digits arg to
print().
It probably would have shown in a simple way that the two numeric versions
aren't really equal
-- and demonstrated a little about R's default printing convention.

> print( pi, digits=3)
[1] 3.14

> print( pi, digits=12)
[1] 3.14159265359

-Don

-- 
Don MacQueen

Lawrence Livermore National Laboratory
7000 East Ave., L-627
Livermore, CA 94550
925-423-1062





On 9/16/11 9:30 AM, "Sarah Goslee" <sarah.goslee at gmail.com> wrote:

>I'll freely admit that I know little about chron and didn't have a
>computer with R in
>front of me at the time.
>
>But the original querent converted the chron times to numeric and then
>used ==,
>and thus it became a 7.31 issue. Converting to character is a different
>way to
>approach the problem; which is "better" depends on more context than we
>have available, as David points out.
>
>Sarah
>
>On Fri, Sep 16, 2011 at 10:58 AM, David Winsemius
><dwinsemius at comcast.net> wrote:
>>
>> On Sep 16, 2011, at 10:08 AM, B77S wrote:
>>
>>> I'm sure Sarah's solution works (and she knows more about R than
>>>myself),
>>> but
>>> I ran into a similar problem and used:
>>> as.character(start.time)==as.character(expected_start.time)
>>
>> What you did was an implicit rounding of the values to the digits level
>> specified by some process, possibly a default formatting spec, guessing
>>to
>> round(num, 7) . I tried to see whether package chron might specify i,t
>>but
>> this effort only produced this
>>
>>> cur.opts <- names(options())
>>> require(chron)
>> Loading required package: chron
>>>  names(options())[!names(options()) %in% cur.opts]
>> [1] "chron.year.abb"    "chron.year.expand"
>>
>> What Sarah suggested was essentially rounding to 16 digits to the right
>>of
>> the decimal.
>>
>> If the OP had offered dput() on these objects, there would  have been a
>>lot
>> less guessing.
>>
>> --
>> David
>>
>>>
>>> good luck regardless.
>>>
>>> -BS
>>>
>>>
>>> Sarah Goslee wrote:
>>>>
>>>> Sounds like a case for FAQ 7.31, or, yet another machine precision
>>>>issue.
>>>> Try all.equal() instead of ==
>>>>
>>>> Sarah
>>>>
>>>> On Fri, Sep 16, 2011 at 7:36 AM, mebstyne <mebstyne at me.com>
>>>>wrote:
>>>>>
>>>>> I have two local variables: startTime and expectedStartTime.  Both
>>>>>are
>>>>> chron
>>>>> related objects.
>>>>> When I look at the class for the objects I can see they are of class
>>>>> "times".
>>>>> When I print them to the console, they both read: "09:30:00"
>>>>> When I print them as.numeric(), they both read: 0.3958333
>>>>> When I try and compare them: (as.numeric(startTime) ==
>>>>> as.numeric(expectedStartTime)) it returns FALSE.
>>>>>
>>>>> I'm mystified.  I would expect them to be true.
>>>>> Perhaps a key to the riddle is how the two objects were created.
>>>>>
>>>>> "startTime" was created by reading a text field from a socket,
>>>>> converting
>>>>> it
>>>>> into a chron object using chron(x, "%m/$d/%Y %H:%M:%S"), then
>>>>>finally I
>>>>> created a time out of the chron by doing a quick butchering of the
>>>>> "integer"
>>>>> portion of the numeric: (times(as.numeric(x) - as.integer(x))
>>>>>
>>>>> "expectedStartTime" was created by the command times('09:30:00')
>>>>>
>>>>> Any suggestions?  Tips?  Alternative approaches?  I've pulled too
>>>>>many
>>>>> hairs
>>>>> triaging this.
>>>>> All hands welcomed on this little challenge.
>>>>>
>>>>> Big picture goal of what I'm doing: I have a list of chron objects
>>>>>with
>>>>> both
>>>>> dates and times portion filled out and I'm trying to determine if the
>>>>> time
>>>>> is a specific time (specific to the minute).
>>>>>
>>>>> Thanks!
>>>>>
>>>>> -Michael
>>>>>
>>>>
>
>
>-- 
>Sarah Goslee
>http://www.functionaldiversity.org
>
>______________________________________________
>R-help at r-project.org 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.