comp.lang.idl-pvwave archive: archive

Home » Public Forums » archive » meanclip

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: meanclip [message #82367 is a reply to message #82319]

Sat, 08 December 2012 14:32

wlandsman
Messages: 743
Registered: June 2000

Senior Member

One problem with using meanclip.pro for your example, is that it determines outliers by comparing with the mean of *all* the data. But your data presumably has real trends, (which you are trying to determine with poly_fit).

Better would be to use "robust" polynomial fitting, in which outliers are determined by their deviation from a polynomial fit. This what the program robust_poly_fit.pro does.

http://idlastro.gsfc.nasa.gov/ftp/pro/robust/robust_poly_fit .pro

--Wayne

On Saturday, December 8, 2012 3:32:57 PM UTC-5, Charlie Paul D'auria wrote:

>
> The array contains data relating to the flux versus time of an astronomical event. The data shows strange characteristics such as dips, which I am hoping to remove by a method other that just 'cutting' them out by eye. I was hoping to use a program that discards data that is outside an allowed limit from the previous data point.
>
>
>
> I have used poly fit to fit a trend line to see if the data fits a certain gradient.
>
>
>
> Anyway, i'll give what i've got a go.
>
>
>
> Charlie

Report message to a moderator

[Message index]

		meanclip By: Charlie Paul D'auria on Thu, 06 December 2012 03:13
		Re: meanclip By: wlandsman on Sat, 08 December 2012 14:32
		Re: meanclip By: Charlie Paul D'auria on Sat, 08 December 2012 12:32

Previous Topic:	Re: How to correct errors an IDL
Next Topic:	shmmap questions

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Oct 10 04:00:12 PDT 2025

Total time taken to generate the page: 0.80016 seconds