A1763: xband h2o maser search

mar05


Intro:

    Project a1763 is doing on off position switching at xband. Each on is 50 six second integrations followed by 50 six second off positions. There are multiple on/offs done per source.  To process a single pattern of 50 records:
  1. compute on/off -1 by record
  2. For each freq channel do a linear fit along the time direction throwing out any points that are outliers (fit residuals greater than 3 sigma). Iterate the fit till no more points are thrown out. Average the results keeping track of the number of seconds of integration for each channel.
The routine then combines all patterns from the same source and averages the polarizations. It uses the integration time per channel as the weight. A baseline is then removed from the final spectra (linear polynomial and second order harmonic). The baseline fit is also iterated throwing out outliers and refitting till no outliers are thrown out. The rms is then computed over the points used in the fit.
 


The plots

    Each file contains 5 sources. Each source has 3 plots:

Note on the various days:

processing:usr/a1763/mar05/do_rawdata.pro, procall.pro
 home_~phil