I notice that the raw data is now available at http://idc-corotn2-public.ias.u-psud.fr/in...sva=browseGraph at least for the first two sessions of COROT observations.
I collected the monochromatic data for the initial session, which is 5600 data series each of about ten thousand points; easy to parse, and my initial analysis was to find ones where the standard deviation of the measurements divided by the average of the measurement standard deviations was large, and then to look at (99th percentile brightness - 1st percentile brightness) / (1st percentile brightness): that gives the gallery at http://fivemack.livejournal.com/182633.html#cutid1
The pretty curves are mostly eclipsing binaries; could someone give me an idea of what the geometry implied by a light curve like http://www.chiark.greenend.org.uk/~twomack...C0102791304.png is? I can get periods to four decimal places by trying to minimise the sum of the standard deviations of the amplitudes in 64 buckets collected under the purported period, but I don't see how to get to the six decimal places that astronomers often seem to mention for orbital periods.
I assume that cataclysmic variables are much less common than cosmic-ray hits on the CCD, but I don't have a very clear idea how to go about cleaning up the sudden jumps that you see in time series like http://www.chiark.greenend.org.uk/~twomack...C0102741414.png - I don't think that 30% jumps in flux in less than one 512-second sample period are likely to be of astronomical origin.
