[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
This is the TASS star I've been working on lately. I got some data last
night and that means more questions. The plot of all the data is here:
I thought, "What the heck is up with all the scatter at the beginning of
the night?" So I tried an experiment. I deleted all observations that were
more than 1 standard deviation from the average, using K-C as my guide.
This, of course reduced by standard deviation of what was left. Then I did
this again. After these two iterations, I was left with a standard
deviation of less than 1%. A plot of this is here:
Question one would be: am I doing a statistical no-no by playing with this?
I understand that I may be removing some actual variation, but it seems
like an interesting exercise nonetheless.
The scatter at the beginning was still there on the variable, even though
there is virtually no scatter on K-C. So I went into Mira and did an
ensemble solution (using only K and C as my ensemble) of just the
beginning of the night (using all the data, not the pruned data). That
plot is here:
You can see clearly that the variable is varying much more than C or K. I
suspect this is near the top hump of the curve, but to me it looks like
there is some substantial flickering going on here. Is this significant?