On Tue, May 25, 2010 at 7:57 PM, Phil Spector <spec...@stat.berkeley.edu>wrote:
> The scale function can use whatever vector you choose for
> subtraction and division. (It's basically a wrapper for
> the sweep function.)
Yeah, I know, I said it a bit strong. In this case it becomes
scale(data,center=F,scale=apply(data,2,function(x){median(abs(x))}))
for dividing through the median absolute value, or
scale(data,center=apply(data,2,function(x){median(abs(x))}),scale=F)
for substracting the median absolute value. But that's neither easier to
read nor faster than just using an apply. I personally believe that it's not
because you can do something someway that it is meant to be done that way.
That's basically what I wanted to point out.
> For example, to subtract the median and divide by the median absolute
> deviation, use
>
> scale(x,center=apply(x,2,median),scale=apply(x,2,mad))
>
> Either the center= or scale= arguments can be omitted if
> you only want to divide or subtract.
>
Not true: both arguments have a default value of TRUE, so omiting them means
you either substract the mean or divide by the sd. (try it out and check the
attributes of the result)
Cheers
Joris
>
> - Phil Spector
> Statistical Computing Facility
> Department of Statistics
> UC Berkeley
> spec...@stat.berkeley.edu
>
>
>
>
>
>
> On Tue, 25 May 2010, Joris Meys wrote:
>
> Scale is written to do that IF you want to normalize according to the mean
>> and the sd. For any other form of normalization, apply or sweep constructs
>> will have to be used.
>>
>> I couldn't really see a way of using the absolute median value in a
>> sweep-statement.
>>
>> On Tue, May 25, 2010 at 7:11 PM, Bert Gunter <gunter.ber...@gene.com>
>> wrote:
>>
>> ?scale
>>>
>>> is specifically written for this. See also ?sweep
>>>
>>> Bert Gunter
>>> Genentech Nonclinical Biostatistics
>>>
>>>
>>>
>>> -----Original Message-----
>>> From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org]
>>> On
>>> Behalf Of Joris Meys
>>> Sent: Tuesday, May 25, 2010 9:54 AM
>>> To: cobbler_squad
>>> Cc: r-help@r-project.org
>>> Subject: Re: [R] File normalization
>>>
>>> My code substracts the median absolute value. If you want to divide by
>>> it,
>>> the code must be :
>>> apply(some_dataset,2,function(
>>>
>>>>
>>>> x){
>>>> x/median(abs(x))
>>>> })
>>>>
>>>
>>>
>>> Thanks to Peter Langfelder for pointing out my mistake.
>>>
>>> On Tue, May 25, 2010 at 6:24 PM, Joris Meys <jorism...@gmail.com> wrote:
>>>
>>> What kind of normalization do you want to do?
>>>> If you want to divide all columns by the median absolute value, try :
>>>>
>>>> apply(some_dataset,2,function(x){
>>>> x-median(abs(x))
>>>> })
>>>>
>>>> also look at ?scale for normalization using the average and the sd.
>>>> Cheers
>>>> Joris
>>>>
>>>>
>>>> On Tue, May 25, 2010 at 6:01 PM, cobbler_squad <la.f...@gmail.com>
>>>>
>>> wrote:
>>>
>>>>
>>>>
>>>>> Dear all,
>>>>>
>>>>> I have a file with 57 columns (671 time points in each column)
>>>>>
>>>>> File looks like this:
>>>>> 1 0.279191 -1.203200e-02 -0.166772 6.12080e-02 0.196379
>>>>> 4.591900e-02 0.293689
>>>>> 2 0.267017 -1.150700e-02 -0.159463 5.85400e-02 0.187775
>>>>> 4.392200e-02 0.280854
>>>>> 3 0.053778 -2.322000e-03 -0.032103 1.18490e-02 0.037921
>>>>> 8.867000e-03 0.056571
>>>>> 4 0.035469 -1.531000e-03 -0.021166 7.79200e-03 0.024937
>>>>> 5.843000e-03 0.037273
>>>>> 5 0.040774 -1.761000e-03 -0.024342 8.96000e-03 0.028674
>>>>> 6.726000e-03 0.042910
>>>>> 6 -0.359709 1.547400e-02 0.214844 -7.87320e-02 -0.253034
>>>>> -5.905100e-02 -0.378322
>>>>>
>>>>> I need to normalize it -- is it possible?
>>>>>
>>>>> I looked into normalize columns of a matrix to have the median absolute
>>>>> value in R, but I am not sure how to apply it in this case. Would very
>>>>> much
>>>>> appreciate any input you could give me..
>>>>>
>>>>> Thank you all in advance,
>>>>>
>>>>> Cobbler
>>>>> --
>>>>> View this message in context:
>>>>> http://r.789695.n4.nabble.com/File-normalization-tp2230251p2230251.html
>>>>> Sent from the R help mailing list archive at Nabble.com.
>>>>>
>>>>> ______________________________________________
>>>>> R-help@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.
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Joris Meys
>>>> Statistical Consultant
>>>>
>>>> Ghent University
>>>> Faculty of Bioscience Engineering
>>>> Department of Applied mathematics, biometrics and process control
>>>>
>>>> Coupure Links 653
>>>> B-9000 Gent
>>>>
>>>> tel : +32 9 264 59 87
>>>> joris.m...@ugent.be
>>>> -------------------------------
>>>> Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php
>>>>
>>>>
>>>
>>>
>>> --
>>> Joris Meys
>>> Statistical Consultant
>>>
>>> Ghent University
>>> Faculty of Bioscience Engineering
>>> Department of Applied mathematics, biometrics and process control
>>>
>>> Coupure Links 653
>>> B-9000 Gent
>>>
>>> tel : +32 9 264 59 87
>>> joris.m...@ugent.be
>>> -------------------------------
>>> Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php
>>>
>>> [[alternative HTML version deleted]]
>>>
>>> ______________________________________________
>>> R-help@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.
>>>
>>>
>>>
>>
>> --
>> Joris Meys
>> Statistical Consultant
>>
>> Ghent University
>> Faculty of Bioscience Engineering
>> Department of Applied mathematics, biometrics and process control
>>
>> Coupure Links 653
>> B-9000 Gent
>>
>> tel : +32 9 264 59 87
>> joris.m...@ugent.be
>> -------------------------------
>> Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php
>>
>> [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help@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.
>>
>>
--
Joris Meys
Statistical Consultant
Ghent University
Faculty of Bioscience Engineering
Department of Applied mathematics, biometrics and process control
Coupure Links 653
B-9000 Gent
tel : +32 9 264 59 87
joris.m...@ugent.be
-------------------------------
Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php
[[alternative HTML version deleted]]
______________________________________________
R-help@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.
