search

LSSTDESC / SLTimer

Time delay measurement from LSST light curves
BSD 3-Clause "New" or "Revised" License
5 stars 5 forks source link

Upgrade whitening function #39

Open drphilmarshall opened 8 years ago

drphilmarshall commented 8 years ago

Over in TDC2, Kai Liao and I experimented with the following alternative whitening scheme:

for each image
    for each filter
        for each season
            find mean flux and subtract from light curve
            find rms flux and divide light curve and error bars through by this number

The result is a whitened light curve that has zero mean and unit variance in each season. For us to implement this we'd to work on the light curves before converting them to magnitudes.

chto commented 8 years ago

I had a short discussion with @drphilmarshall today. Right now we are thinking to do this in two ways. The first one is that given a season, we shift the magnitude of each band to the mean value of the total data in that season. This corresponds to rescale the flux and the error in magnitude should leave the same. ( because sigma_magnitude = sigma_flux/flux) Another way to do this is to shift the flux directly and leave the sigma_flux the same.

drphilmarshall commented 7 years ago

Is this now implemented, with docstrings that explain the different options?