The collapse code had me thoroughly confused today when I realized that it doesn’t actually take the strip of data and rotate it based on the inputted roll angle, which is what I previously thought it did. I still haven’t fully figured out what it does. The roll angle does change the slope of the strip, but it also changes the strip thickness and significantly shifts the x-range and y-range. Negative roll angles do something truly weird and the data is plotted as a 2D pattern instead of a strip. I couldn’t quite figure it out.

I resorted to the method of trial and error. Starting with a roll angle of 0 (no rotation, the original configuration), I ran the program multiple times in increments of 5 degrees to observe the changes. Finally, I discovered that an input of 78.5 degrees for that particular event file rotated the strip to horizontal. This corresponded to a formula of 360 – 90 – PA_PNT = roll, or 270 – PA_PNT = roll, where PA_PNT is an angle obtained from the header and roll is the angle used in calculations for this particular code.

With that hurdle finally behind me, I can continue modifying a larger code to correctly pinpoint an x-range in which the source is above the background and filter out anything outside the range. This requires me to learn about HPD and percentages within a Gaussian and other mathematical basics I know nothing about.

Last time I checked earlier today, the GSFC FTP site was still down, leaving me still unable to download CALDB or barycorr. I would think a server of that importance would be fixed quickly. I wonder if there is a mirror site somewhere.