| Identifying outliers in data [message #91297] |
Thu, 25 June 2015 19:03  |
siumtesfai
Messages: 62
Registered: April 2013
|
Member |
|
|
Hi All,
I am using cgboxplot.pro to identify outliers in my data. It is nice program that I see I have outliers in my data
Next step I would like to store my good data to an array and continue processing them.
My data is two dimension wind data
wind = Array(number of days, pressure levels)
e.g wind= Array( 31, 17)
Once I am able to exclude the outliers from my daily dataset, I am interested to make monthly mean data set
Can anyone suggest me how I would solve my problem
Thank you in advance
Best regards
|
|
|
|
|
|
| Re: Identifying outliers in data [message #91302 is a reply to message #91297] |
Fri, 26 June 2015 11:57  |
siumtesfai
Messages: 62
Registered: April 2013
|
Member |
|
|
On Thursday, June 25, 2015 at 10:03:34 PM UTC-4, siumt...@gmail.com wrote:
> Hi All,
>
> I am using cgboxplot.pro to identify outliers in my data. It is nice program that I see I have outliers in my data
>
> Next step I would like to store my good data to an array and continue processing them.
>
>
> My data is two dimension wind data
>
> wind = Array(number of days, pressure levels)
>
> e.g wind= Array( 31, 17)
>
> Once I am able to exclude the outliers from my daily dataset, I am interested to make monthly mean data set
>
>
>
> Can anyone suggest me how I would solve my problem
>
> Thank you in advance
>
> Best regards
I would think i can do this
; Draw outliners if there are any.
IF maxcount GT 0 THEN BEGIN
outliermax=fltarr(maxcount)
FOR k = 0,maxcount-1 do outliermax(k)=imax(k)
print,'outliermax'
print,outliermax
FOR j=0,maxcount-1 DO PLOTS, xlocation, data[imax[j]], $
PSYM=cgSymCat(9), COLOR=cgColor(outliercolor), NOCLIP=0
ENDIF
IF mincount GT 0 THEN BEGIN
outliermin=fltarr(mincount)
FOR kk = 0,mincount-1 do outliermin(kk)=imin(kk)
print,'outliermin'
print,outliermin
FOR j=0,mincount-1 DO PLOTS, xlocation, data[imin[j]], $
PSYM=cgSymCat(9), COLOR=cgColor(outliercolor), NOCLIP=0
ENDIF
But the problem would be the original data have been sorted . I would have a problem locating the location or index of the outlier in the original data.
I found in the above step is the index or location from the already sorted data.
Best regards
|
|
|
|
| Re: Identifying outliers in data [message #91306 is a reply to message #91302] |
Sat, 27 June 2015 10:06 |
Jeremy Bailin
Messages: 618
Registered: April 2008
|
Senior Member |
|
|
On Friday, June 26, 2015 at 2:57:37 PM UTC-4, siumt...@gmail.com wrote:
> On Thursday, June 25, 2015 at 10:03:34 PM UTC-4, siumt...@gmail.com wrote:
>> Hi All,
>>
>> I am using cgboxplot.pro to identify outliers in my data. It is nice program that I see I have outliers in my data
>>
>> Next step I would like to store my good data to an array and continue processing them.
>>
>>
>> My data is two dimension wind data
>>
>> wind = Array(number of days, pressure levels)
>>
>> e.g wind= Array( 31, 17)
>>
>> Once I am able to exclude the outliers from my daily dataset, I am interested to make monthly mean data set
>>
>>
>>
>> Can anyone suggest me how I would solve my problem
>>
>> Thank you in advance
>>
>> Best regards
>
>
>
>
> I would think i can do this
>
> ; Draw outliners if there are any.
> IF maxcount GT 0 THEN BEGIN
> outliermax=fltarr(maxcount)
> FOR k = 0,maxcount-1 do outliermax(k)=imax(k)
> print,'outliermax'
> print,outliermax
>
> FOR j=0,maxcount-1 DO PLOTS, xlocation, data[imax[j]], $
> PSYM=cgSymCat(9), COLOR=cgColor(outliercolor), NOCLIP=0
> ENDIF
> IF mincount GT 0 THEN BEGIN
>
> outliermin=fltarr(mincount)
> FOR kk = 0,mincount-1 do outliermin(kk)=imin(kk)
> print,'outliermin'
> print,outliermin
> FOR j=0,mincount-1 DO PLOTS, xlocation, data[imin[j]], $
> PSYM=cgSymCat(9), COLOR=cgColor(outliercolor), NOCLIP=0
> ENDIF
>
>
>
> But the problem would be the original data have been sorted . I would have a problem locating the location or index of the outlier in the original data.
>
> I found in the above step is the index or location from the already sorted data.
>
> Best regards
What are imin and imax? How do they get defined? Once you understand that, then you will know how they relate to your original data.
-Jeremy.
|
|
|
|
| Re: Identifying outliers in data [message #91307 is a reply to message #91306] |
Sat, 27 June 2015 10:16 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Jeremy Bailin writes:
> What are imin and imax? How do they get defined? Once you understand that, then you will know how they relate to your original data.
Right. iMin and iMax *are* the locations of the outliers in the original
data, assuming your original data didn't contain "missing" values.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.idlcoyote.com/
Sepore ma de ni thue. ("Perhaps thou speakest truth.")
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
