Bug in ellipse aperture?

[altuson]

tess-asteroids/src/tess_asteroids/movingtpf.py

Line 707 in 8608663

X, Y, X2, Y2, XY = compute_moments(self.flux, threshold_mask)

@jorgemarpa I think compute_moments() should be using self.corr_flux, not self.flux. I assume we want the aperture computed for BG corrected flux, not raw flux. Could you please fix this bug?

[altuson]

I see the problem here... corr_flux has negative values (due to BG correction) and hence moments can become negative (unphysical). Maybe we can discuss this.

[altuson]

For now I have masked out negative values from computation of moments. This means we can apply function to corrected flux even when BG correction results in negative values.

[altuson]

@jorgemarpa There is also a problem when there are no pixels above the threshold value. In this case, there are no pixels in the threshold mask and hence compute_moments() errors. Maybe we can add a catch that if there are no pixels in the threshold mask, ellipse mask is also none?