Bob has been spending this week working on the Informix database. He found a data corruption in a bad disc, so he has to re-create it.

Josh has spent some time investigating the reason why his chi-squared algorithm doesn’t behave like a chi-squared test. And as it turns out, it isn’t supposed to. So the algorithm actually does work for RFI rejection, but we don’t really understand the theory behind why it works. The figure of merit is still valid. He is starting to set up some programs to look at some individual pulses. There are several things we can start working on, such as making a map of the sky, starting to looking at a time series, and look at pulses in a row. Then we need to apply it to the data.

Jeff’s work for the week as been on blanking signal synchronization, and working on figuring out what is slowing NTPCkr down. It used to be fast but now, well, it is as slow as dirt.

Last week Dave visited MilkyWay@Home at RPI in Albany NY, a project that models the structure of the Milky Way. The problem that they are running into is that they are minimizing functions that have a lot of local minima. If we ever run into problems like this we can use some of the code they have developed.

Dave also shared a sad story from RPI: the frequency of the jackhammers from the construction outside of the one of the buildings caused 1/3 of the disc drives to fail. Ouch!

Matt is working on the software radar blanker, and continuing to feed things through the pipeline. We’ve already gone through 10% of data that had never been radar blanked, but we should be ok since we are getting some data from Arecibo here soon.

That’s all for this week from SSL!