The Occultation of a 12.7 star by Sthenelos

Sept 7 at 4:04 am

 

This event has Santa Cruz just inside the limit. Normally I'd try it of course - but I bailed out in favor of getting badly needed sleep, ahead of my planned GraniteMan race starting 2 days afterward. Kirk was game, and got a nice positive from his driveway. 1.5 second predicted maximum. Strange that PyOTE would not see this obvious event when done at the 4x actual integration, but with FEWER points by manually integrating as if 8x, it did find the event. PyOTE, in my opinion, needs re-working as it is not properly identifying valid events. I've had this complaint for some time, but the people in IOTA in charge, do not put any importance, it seems on short or difficult events even though those dominate the predictions and useful data is being wasted. I'll bet that better more accurate timings from Kirk's event below could have been gotten with PyOTE if the test for an event were better coded and better reasoned.

 

 
Got an obvious event of Sthenelos from my sidewalk at predicted time 04:04:56.  
I recorded at 4x, ran PyMovie in field mode.
But PyOTE could not find an event with auto block integrate.
That's bogus, any person you could grab off the street could point to the event in the graph.
I tried PyOTE again with 8 point block manual integration, it found an event, passed false positive.
But duration is longer than predicted max 1.5: 

magDrop: 1.689 +/- 0.684 (0.95 ci) snr: 2.44

D time: [11:04:55.6528] D: 0.6800 containment intervals: {+/- 0.0628} seconds D: 0.9500 containment intervals: {+/- 0.1885} seconds D: 0.9973 containment intervals: {+/- 0.4711} seconds R time: [11:04:57.5213] R: 0.6800 containment intervals: {+/- 0.0628} seconds R: 0.9500 containment intervals: {+/- 0.1885} seconds R: 0.9973 containment intervals: {+/- 0.4711} seconds Duration (R - D): 1.8685 seconds Duration: 0.6800 containment intervals: {+/- 0.0960} seconds Duration: 0.9500 containment intervals: {+/- 0.2535} seconds Duration: 0.9973 containment intervals: {+/- 0.5615} seconds

I might make a couple more passes later in PyMovie to see if PyOTE finds it with auto block.
Odd.