| T O P I C R E V I E W |
| rdelvill |
Posted - 06/11/2014 : 03:48:03 AM
Hi,
I am doing a linear fitting using a set of data on a log-log graph. I am using "Apparent Fit" because the log(x) is expected to behave linearly with log(y). However, it seems that the confidence and prediction bands are not calculated using the "apparent" data (i.e. log(x) & log(y)) so they do not correspond to the apparent fit. Is this a known problem or am I forgeting to select an option to have also an "apparent" confidence/prediction interval?
Thank you for your help.
Rémi Delville |
| 1 L A T E S T R E P L I E S (Newest First) |
| jasonzhao |
Posted - 06/11/2014 : 08:03:17 AM
Hi, Rémi Delville
Apparent Fit is only useful when you fit from a graph and change the plot axis type (from Linear to Log10, for example). If you check this option, the fitter will first transform raw data into new data space as specified in the graph axis type, and then fit the curve with the new data (then, fitting results will transform back to x-y space). Apparent fit is equivalent to direct fit if you first transform raw data on the worksheet.
If I understand you correctly, you would like to have the confidence bands suitable for log(x) & log(y) linear plot.
the confidence and prediction bands are calculated using the data log(x) & log(y), after the calculation, the results will transform back to (x,y) space but not Log(x)Log(y) space. If you would like to have the confidence bands suitable for X=log(x)&Y=log(y) plot, please do a Log calculation for them after the Apparent Fit.
Don't hesitate to contact us for further questions.
Best regards,
Jason Zhao
OriginLab Tech Service |
|
|
