ProgressReport 11 October 2024

The big thing on my mind has been the house. Looks like this is going to happen. Will stress the finances in the near term, but probably the right move. Should allow us to start earning revenue on 720. Not drinking has not been a factor. Was solo dad for a week, tiring but manageable.

Started systematics project with Chris and Jake. Not much progress on the stats work, nice presentation by Souhen at STAMPS and paper accepted for conference NeurIPS (ML4PS2024)

Finished synthetic datasets for now. A more detailed debugging uncovered some bugs that have been fixed, but dont really change the big picture. I tried hT reweighting, but again didnt affect the big picture much. On this front the next step is to get it running in Run3.

Lots of progress on run 3 skims in 2023. Not have a reasonable understanding of how to use condor. The skims have been produced for 2023. (There is still a bug in 2022). Using dask with condor has been easy, although its still not clear to me how I would debug problems. Ive added tools to ensure that we are seeing all the data (lumi blocks) expected. Will then proceed with synthetic dataset making.

AE has been chasing down a change in final sensitivity. This was noticed with coming back to the Boosted overlap study. He has added tools to better monitor if and when changes to final sensitivity happen (reAna now running weekly). Once this is understood, he will continue to work on Low Pt.

I think Chuyuan finally has the classifier code under control. This needs to be verified soon. Then he needs to move to kλ gains, SvB and tt vs no tt. Looks like we will also need a strategy to incorporate VBF

I think the double peak in m4j is understood. I'm waiting on Sindhu to write it up and move along to systematics. Progress here has been slow.

I've started working on mW and mH constraints. I think I understand whats happening in mW, but still need to push on mH. There has also not been as much progress here as I would want from Aniket.

Still not much. We have 150 Si sensors now stored in new dry air space. We have also made module with the new glue pattern. It looks like the high noise and strange IV behaviour are not present in the new module. The next steps here are to move on the Lab tech positions.

Peter is back and I think Wes has understood the comine fit problems. The issue of the empty bins might be bug in the binning. Peter will look into this.