Author |
Topic |
|
MatthieuB
Canada
7 Posts |
Posted - 05/18/2009 : 3:51:30 PM
|
Hello,
I was wondering if origin provided a tool which could be used to bin data. I searched in the help files but could not find anything. When I speak of binning data I do not mean as in a histogram where a frequency is associated to the bin value. But the pre-processing technique which reduces the amount of data points by binning the data. I am not sure if I am clear but there is a video example of what I mean at the following link :
http://www.lolife.com/astronomy/starhouse/Starhouse%20Blog/F63FA055-9018-485F-A630-812AA06F9B9B.html
Any help would be greatly appreciated. Thank you, Matt |
|
easwar
USA
1964 Posts |
Posted - 05/18/2009 : 9:02:14 PM
|
Hi Matt,
Thank you for the link, we are looking into adding such data reduction capability in the next version, so this is useful/timely!
Wanted to ask you about the error bars on that movie...do you know if the error bars represent some statistical error, or there is indeed three columns x, y, yerr and the reduction takes into account the yerr as well?
Thanks,
Easwar OriginLab
|
|
|
MatthieuB
Canada
7 Posts |
Posted - 05/19/2009 : 11:15:32 AM
|
I believe the error is statistical error and not systematic but I am not 100% sure on that(I just found the link by googling so i believe there would be a x,y and yer column). The reason why the error is diminishing is because the points are being replaced by an average of points (taking an average obviously decreases error).
The particular purpose for which I require this data treatment does not actually have the error bars before the treatment. I am just analyzing noisy data which has many points. I would like to replace this data by fewer points which could have an error associated to them such as the standard deviation of the averaged points. I have tried replacing my data by polynomial fits but we are also interested in the derivatives at the edges of the curve. Unfortunately the poly fits tend to diverge a bit away from the curve at the edges.
I guess I could write some program in matlab to pretreat the data and then import it back into origin but I do not enjoy having to switch back and forth between programs. It would definitely be practical to incorporate this sort of data treatment in future versions of origin(at least for me).
Matt |
|
|
easwar
USA
1964 Posts |
Posted - 05/19/2009 : 12:38:20 PM
|
Hi Matt,
Thank you for providing more information. Our tool currently in development was not generating an error column, so we will look into that.
I have e-mailed you with some more information and am awaiting your reply. If you do not get the e-mail, please let us know.
Easwar OriginLab |
|
|
|
Topic |
|
|
|