correlation coefficient with uncertain data [message #87972] |
Fri, 07 March 2014 00:52  |
amin farhang
Messages: 39 Registered: November 2010
|
Member |
|
|
Dear all,
Is it possible to calculate correlation coefficient for uncertain data with errors in both x and y direction?
IDL> X = [4,5,6,7]
IDL> Xerr = [1,1.2,1.5,1.7]
IDL> Y = [3,4,5,6]
IDL> Yerr = [1.5,2,2.5,2.7]
Best regards,
|
|
|
Re: correlation coefficient with uncertain data [message #87973 is a reply to message #87972] |
Fri, 07 March 2014 07:56   |
Andy Sayer
Messages: 127 Registered: February 2009
|
Senior Member |
|
|
The calculation of a correlation coefficient does not depend on the uncertainty associated with the data. http://www.exelisvis.com/docs/CORRELATE.html
(Of course, dependent on the uncertainty of the data, the apparent correlation on the data may be stronger or weaker than the true correlation between the mechanisms which the data are derived from.)
On Friday, March 7, 2014 3:52:52 AM UTC-5, Amin Farhang wrote:
> Dear all,
>
>
>
> Is it possible to calculate correlation coefficient for uncertain data with errors in both x and y direction?
>
>
>
> IDL> X = [4,5,6,7]
>
> IDL> Xerr = [1,1.2,1.5,1.7]
>
> IDL> Y = [3,4,5,6]
>
> IDL> Yerr = [1.5,2,2.5,2.7]
>
>
>
> Best regards,
|
|
|
|