[R] Find tibble row with maximum recorded value

[Avi Gross]

The right answer obviously depends on the REQUIREMENTS and they may not have been fully stated.

This is a bit like finding the mode of a set of numbers. The most frequent value may not be as representative of the data as the mean or even the median for some purposes, as well as other measures of central tendency.

What does Rich want? Choices can include getting the entire row where the value happens first, or maybe even the last or ALL of them as a group.

Be that as it may, if you return all of them, as a mini-data.frame, it is easy enough to then choose to keep them all, or just the first by subsetting  it to get row 1, or even to get the last row by applying nrow() to the result.

I can think of many ways you want an answer including beyond this, some kind of blending of the other columns you get back to make some kind of composite row from all the rows that matched.

Without clear requirements, there often is no right answer.

-----Original Message-----
From: R-help <r-help-bounces using r-project.org> On Behalf Of Rui Barradas
Sent: Saturday, December 4, 2021 2:34 AM
To: Bert Gunter <bgunter.4567 using gmail.com>
Cc: R-help <r-help using r-project.org>
Subject: Re: [R] Find tibble row with maximum recorded value

Hello,

You're right, I carelessly coded this.

which.max returns the index to the first maximum of a vector, while the comparison of a vector with its max() returns an index to all vector elements.

Às 23:27 de 03/12/21, Bert Gunter escreveu:
> Perhaps you meant to point this out, but the cfs[which.max(cfs)] and 
> cfs == ... are not the same:
> 
>> x <- rep(1:2,3)
>> x
> [1] 1 2 1 2 1 2
>> x[which.max(x)]
> [1] 2
>> x[x==max(x)]
> [1] 2 2 2
> 
> So maybe your point is: which does the OP want (in case there are 
> repeated maxes)? I suspect the == forms, but ...?
> 
> Bert Gunter
> 
> On Fri, Dec 3, 2021 at 2:56 PM Rui Barradas <ruipbarradas using sapo.pt> wrote:
>>
>> Hello,
>>
>> Inline.
>>
>> Às 22:08 de 03/12/21, Rich Shepard escreveu:
>>> On Fri, 3 Dec 2021, Rich Shepard wrote:
>>>
>>>> I find solutions when the data_frame is grouped, but none when it's not.
>>>
>>> Thanks, Bert. ?which.max confirmed that's all I need to find the 
>>> maximum value.
>>>
>>> Now I need to read more than ?filter to learn why I'm not getting 
>>> the relevant row with:
>>>> which.max(pdx_disc$cfs)
>>> [1] 8054
>>
>> This is the *index* for which cfs is the first maximum, not the 
>> maximum value itself.
>>
>>>
>>>> filter(pdx_disc, cfs == 8054)
>>
>> Therefore, you probably want any of

Should be "one of", not "any of"

Rui Barradas

>>
>>
>> filter(pdx_disc, cfs == cfs[8054])
>>
>> filter(pdx_disc, cfs == cfs[which.max(cfs)])
>>
>> filter(pdx_disc, cfs == max(cfs))    # I find this one better, simpler
>>
>>
>> Hope this helps,
>>
>> Rui Barradas
>>
>>
>>> # A tibble: 0 × 9
>>> # … with 9 variables: site_nbr <chr>, year <int>, mon <int>, day <int>,
>>> #   hr <dbl>, min <dbl>, tz <chr>, cfs <dbl>, sampdt <dttm>
>>>
>>> Regards,
>>>
>>> Rich
>>>
>>> ______________________________________________
>>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see 
>>> 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.
>>
>> ______________________________________________
>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see 
>> 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.

______________________________________________
R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.