How can I resample data with errors linearly in log space?

[Veles]

I need to resample a set of data and its errors linearly in log space, with the same number of points. I was just going to interpolate between points to get the data - but how do I calculate the errors?

 

[grquanti]

what do you mean with "resample" if then you talk about interpolating them?

 

[plasmon_shmasmon]

If you're interpolating data points, why not interpolate their errors? The errors produce two additional sets of data points above/below the original data. Treat them the same way you're treating your original data.

I assume the "data" are from a numerical simulation? Otherwise it is probably not legit to interpolate experimental errors.

 

[Veles]

Its experimental errors in astronomy in a spectra. It seems to be quite a commonly done thing - rebinning from linear wavelength to logarithmic whilst maintaining the overall signal to noise. Given that it is "non legit" I should have probably asked in the physics rather than maths forum!

For example I have done it using this IRAF package. http://drforum.gemini.edu/wp-content/uploads/2014/05/README.txt