diaObject summary stats are only provided for the observed filter

Support Data Products
#1

Each time we receive an alert, we get summary stats with the diaObject relating to the current band only. All the other bands are NaN or zero (see r-band example below). Is there a reason for this? I assume this is a speed issue, but if not, are there plans to preserve the summary stats for each band as we receive the diaObject rather than only providing only the observed filter information? (This object has multiple observations in multiple filters.)

It’s true that the brokers could do their own calculations based on the received lightcurve, or even only write data their database tables when the populated columns are not NaN/zero, but I’m wondering if this can be done at source. In our broker (Lasair) the object table data is currently overwritten on receipt of each alert. Users have requested the ability to write Lasair alert filters based on multiple band summary stats.

  "alert": {
    "diaObject": {
      "diaObjectId": 314003014107006318,
      "validityStartMjdTai": 61172.00409415646,
      "ra": 149.92866618500474,
      "raErr": 3.2370393455494195e-05,
      "decErr": 5.428689837572165e-05,
      "ra_dec_Cov": 2.5497976993693605e-11,
      "u_psfFluxMean": NaN,
      "u_psfFluxMeanErr": NaN,
      "u_psfFluxSigma": NaN,
      "u_psfFluxNdata": 0,
      "u_fpFluxMean": NaN,
      "u_fpFluxMeanErr": NaN,
      "g_psfFluxMean": NaN,
      "g_psfFluxMeanErr": NaN,
      "g_psfFluxSigma": NaN,
      "g_psfFluxNdata": 0,
      "g_fpFluxMean": NaN,
      "g_fpFluxMeanErr": NaN,
      "r_psfFluxMean": 12221.5986328125,
      "r_psfFluxMeanErr": 48.470069885253906,
      "r_psfFluxSigma": 166975.859375,
      "r_psfFluxNdata": 25,
      "r_fpFluxMean": NaN,
      "r_fpFluxMeanErr": NaN,
      "i_psfFluxMean": NaN,
      "i_psfFluxMeanErr": NaN,
      "i_psfFluxSigma": NaN,
      "i_psfFluxNdata": 0,
      "i_fpFluxMean": NaN,
      "i_fpFluxMeanErr": NaN,
      "z_psfFluxMean": NaN,
      "z_psfFluxMeanErr": NaN,
      "z_psfFluxSigma": NaN,
      "z_psfFluxNdata": 0,
      "z_fpFluxMean": NaN,
      "z_fpFluxMeanErr": NaN,
      "y_psfFluxMean": NaN,
      "y_psfFluxMeanErr": NaN,
      "y_psfFluxSigma": NaN,
      "y_psfFluxNdata": 0,
      "y_fpFluxMean": NaN,
      "y_fpFluxMeanErr": NaN,
      "u_scienceFluxMean": NaN,
      "u_scienceFluxMeanErr": NaN,
      "g_scienceFluxMean": NaN,
      "g_scienceFluxMeanErr": NaN,
      "r_scienceFluxMean": 40552.015625,
      "r_scienceFluxMeanErr": 46.600669860839844,
      "i_scienceFluxMean": NaN,
      "i_scienceFluxMeanErr": NaN,
      "z_scienceFluxMean": NaN,
      "z_scienceFluxMeanErr": NaN,
      "y_scienceFluxMean": NaN,
      "y_scienceFluxMeanErr": NaN,
      "u_psfFluxMin": NaN,
      "u_psfFluxMax": NaN,
      "u_psfFluxMaxSlope": NaN,
      "u_psfFluxErrMean": NaN,
      "g_psfFluxMin": NaN,
      "g_psfFluxMax": NaN,
      "g_psfFluxMaxSlope": NaN,
      "g_psfFluxErrMean": NaN,
.
.
. etc
2 Likes
#2

Hi @genghisken, this is indeed just a bug–one I had noted earlier but hadn’t chased down yet. We’ll see if we can expedite a fix.

3 Likes
#3

This turned out to be an easy fix and should be show up in the next prompt processing release sometime next week. Note that the DIAObject statistics will only be updated when they receive new DIASource detections.

2 Likes
#4

Thanks @ebellm. Does the same bug account for the lack of information in the forced photometry fields. For example, I notice that in the above JSON the populated band is r, but there is no entry for the r_fpFluxMean or r_fpFluxMeanErr. Or is that a different problem?

1 Like