put the cas field star plots from last week on the same axes as the Zeljko plots above to compare them: u'-g' Vs u'-u, g'-r' Vs g'-g, r'-i' Vs r'-r, r'-i' Vs i'-i, and i'-z' Vs z'-z.


7/19:
Got actual ugriz' data to compare to my converted stuff, found them not to be quite the same. Plotted a histogram of the g' residuals, and also a g'-r' Vs g' residuals graph.


7/20:
Got the NumRec least-squares fitting program to work and ran cubic fits for all the median points with .03 error cuts from my converted Zeljko data. The fits are of the solid red dots in the plot. The numbers on the bottom are the points considered for each median-dot. The fit ignores any dots with fewer than 4 data points in it, and multiplies the error by 5 of any points that have between 3 and 10 points: u'-g' Vs u'-u, g'-r' Vs g'-g, r'-i' Vs r'-r, r'-i' Vs i'-i, and i'-z' Vs z'-z. I added some more fits to some of the plots that looked like they could be fit better. On the following plots green is the cubic fit, orange is a quadratic fit, gold is a linear fit, and purple is a quartic fit: g'-r' Vs g'-g, r'-i' Vs r'-r, and r'-i' Vs i'-i. Made a couple of zoomed in versions for r'-i' Vs r'-r and r'-i' Vs r'-r.


7/21:
Since Tucker's data is somehow slightly different from ours, I plotted up a transform of his data (and Zeljko, for the non-prime) to see if the differences were truly noticeable: r'-i' Vs r'-r.

Decided on the best fit curves for each transform, and plotted them with the corresponding equations.

Back to main