Is there in R a function equivalent to the mround, as found in most spreadsheets?

[Dr. Ottorino-Luca Pantani]

Dear R-users,

I have the following problem

In a lab experiment I have to mix three solutions to get different
concentrations of various molecules in a cuvette

I've used R to calculate the necessary µliters for each of the level of
the experiment and I must confess that it is more useful and easier to
achieve the results than using spreadsheets.

But there's a problem.

Imagine that for a particular cuvette (I have 112 different cuvettes !!)
you have to mix the following volumes of solution A, B, and C respectively.

c(1803.02, 193.51, 3.47)

Each solution is to be taken with 3 different pipettes (5000, 250 and 10
µL Volume max) and each of those delivers volumes in steps of 50 µL,  5
µL  or 1µL, respectively
Since the above values  would eventually become

c(1800, 195, 3)

it is then necessary to recalculate all the final concentrations
of A, B and C, because the volumes are changed.


I know that in most spreadsheets (Calc, Gnumeric, Excel
and so on) there's a function such as

mround(num; num)

that give the results I need, but I want to learn more on R functions.


I played a little with R functions such as

round, signif, ceiling, trunc, and floor

but without success.
Any hint to solve this problem ?

Thanks a lot

http://www.openofficetips.com/blog/archives/2005/04/rounding_to_the.html
http://www.gnome.org/projects/gnumeric/doc/gnumeric-MROUND.shtml

--
Ottorino-Luca Pantani, Università di Firenze
Dip. Scienza del Suolo e Nutrizione della Pianta
P.zle Cascine 28 50144 Firenze Italia
Tel [phone] (348 lab) Fax [phone]
OLPantani@...  http://www4.unifi.it/dssnp/

______________________________________________
R-help@... 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.

[Dieter Menne]

Dr. Ottorino-Luca Pantani <ottorino-luca.pantani <at> unifi.it> writes:

>
> In a lab experiment I have to mix three solutions to get different
> concentrations of various molecules in a cuvette
>
> I've used R to calculate the necessary µliters for each of the level of
> the experiment and I must confess that it is more useful and easier to
> achieve the results than using spreadsheets.
>
> But there's a problem.
>
> Imagine that for a particular cuvette (I have 112 different cuvettes !!)
> you have to mix the following volumes of solution A, B, and C respectively.
>
> c(1803.02, 193.51, 3.47)
>
> Each solution is to be taken with 3 different pipettes (5000, 250 and 10
> µL Volume max) and each of those delivers volumes in steps of 50 µL,  5
> µL  or 1µL, respectively
> Since the above values  would eventually become
>
> c(1800, 195, 3)
>
> it is then necessary to recalculate all the final concentrations
> of A, B and C, because the volumes are changed.

A first guess would be

a = c(1803.02, 193.51, 3.47)
round(a / 5)*5

which gives

1805  195    5

This is not exactly what you want, but it shows that the problem is a bit
ill-defined. In the example you gave, why do you want 1800, and not 1805, which
is possible with the pipettes? I assume that you laboratory experience is
working in the background, telling you to stop pipetmanning when you are close
to the result in some "percentage" feeling.

Dieter

______________________________________________
R-help@... 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.

[Charilaos Skiadas-3]

On May 9, 2008, at 5:39 AM, Dieter Menne wrote:

> Dr. Ottorino-Luca Pantani <ottorino-luca.pantani <at> unifi.it>  
> writes:
>>
>> Imagine that for a particular cuvette (I have 112 different  
>> cuvettes !!)
>> you have to mix the following volumes of solution A, B, and C  
>> respectively.
>>
>> c(1803.02, 193.51, 3.47)
>>
>> Each solution is to be taken with 3 different pipettes (5000, 250  
>> and 10
>> µL Volume max) and each of those delivers volumes in steps of 50  
>> µL,  5
>> µL  or 1µL, respectively

>> Since the above values  would eventually become
>>
>> c(1800, 195, 3)
>>
>> it is then necessary to recalculate all the final concentrations
>> of A, B and C, because the volumes are changed.
>
> A first guess would be
>
> a = c(1803.02, 193.51, 3.47)
> round(a / 5)*5
>
> which gives
>
> 1805  195    5

If I understand the OP's question properly, the first value is to be  
a multiple of 50, the second a multiple of 5, and the third a  
multiple of 1. This can be done with this slight variation on the  
above theme:

a <- c(1803.02, 193.51, 3.47)
b <- c(50,5,1)
round(a/b) *b

> This is not exactly what you want, but it shows that the problem is  
> a bit
> ill-defined. In the example you gave, why do you want 1800, and not  
> 1805, which
> is possible with the pipettes? I assume that you laboratory  
> experience is
> working in the background, telling you to stop pipetmanning when  
> you are close
> to the result in some "percentage" feeling.
>
> Dieter

Haris Skiadas
Department of Mathematics and Computer Science
Hanover College

______________________________________________
R-help@... 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.

[Dr. Ottorino-Luca Pantani]

Charilaos Skiadas ha scritto:
> On May 9, 2008, at 5:39 AM, Dieter Menne wrote:
>
> If I understand the OP's question properly, the first value is to be a
> multiple of 50, the second a multiple of 5, and the third a multiple
> of 1. This can be done with this slight variation on the above theme:
>
> a <- c(1803.02, 193.51, 3.47)
> b <- c(50,5,1)
> round(a/b) *b

This is exactly was I was looking for, and you understand it properly.
Sorry for the double post, but I'm experiencing problems with my account
in this list

>
>> This is not exactly what you want, but it shows that the problem is a
>> bit
>> ill-defined. In the example you gave, why do you want 1800, and not
>> 1805, which
>> is possible with the pipettes? I assume that you laboratory
>> experience is
>> working in the background, telling you to stop pipetmanning when you
>> are close
>> to the result in some "percentage" feeling.
>>
>> Dieter

No, even if there's a lab background working behind, I'm only
transferring some experience with spreadsheets to R.

I want (actually "the pipette wants") 1800 instead that 1805 since the
larger volumes are to be delivered with a digital pipette whose step is 50.

Thanks to all the people who contributed. This list is always a great
source.

8rino

______________________________________________
R-help@... 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.

[Alberto Santini]

Hello Luca.

Dr. Ottorino-Luca Pantani wrote:

...

c(1803.02, 193.51, 3.47)

Each solution is to be taken with 3 different pipettes (5000, 250 and 10
µL Volume max) and each of those delivers volumes in steps of 50 µL,  5
µL  or 1µL, respectively
Since the above values  would eventually become

c(1800, 195, 3)

...

You find a complete solution at
  http://albertosantini.blogspot.com/2008/05/mround.html

It checks the sign of the number and the multiple and it hacks the issue of rounding off a 5 respect IEC 60559 standard.

The test cases are:

mround(10, 3) # 9
mround(-10, -3) # -9
mround(1.3, 0.2) # 1.4
mround(5, -2) # error
mround(1.7, 0.2) # 1.8
mround(321.123, 0.12) # 321.12
mround(1803.02, 50) # 1800
mround(193.51, 5) # 195
mround(3.47, 1) # 3


Regards,
Alberto