[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Looking at Data
Before I write a lot of code I try to do a project first with pencil and paper. It saves one from a big effort about nothing. Sometimes you have to write a lot of code before you see nothing. Sigh!
I am setting out to work on calibration of the tass data. For a test, I took one hour in RA worth and found the stars measured 20 times or more. I computed means and sigmas for the measurement sets.
Then I downloaded the Landolt list from the tass home page and put the two lists up side by side on the monitor.
This way I found 4 that were on both lists. I don't think I could do this easily with a Unix compare since the positions do not exactly match. I recall they all matched to a minute or less of arc.
Landolt tass tass sigma
11.424 11.379 0.042
11.452 11.382 0.076
9.088 9.077 0.071
9.905 9.866 0.065
The tass mags are all smaller (brighter) than the Landolt values. The mean of the differences is 0.041. Not so bad. I can call this a calibration error and subtract it from the tass values. Then with this large data set I can compute a sigma. I did. It is 0.024.
First, everyone, don't take this seriously. I am not claiming a "reference to the standard system" error of 0.024. This is just for fun. But it is sort of encouraging and drives me to go write code to compare the whole data set to the Landolt list and to the nice big photometry of Brian Skiff. Then we shall see what the experts will allow me to say about the data.
Tom Droege
Thomas Droege
Why Wait? Move to EarthLink.