XENON1T / pax

The XENON1T raw data processor [deprecated]
BSD 3-Clause "New" or "Revised" License
16 stars 15 forks source link

S1 patterns binning issue #602

Closed l-althueser closed 7 years ago

l-althueser commented 7 years ago

I found that the S1 per-PMT map has a binning issue around z=0. The current map is generated with only 10 bins in z which causes the AFT map to have a continuous LXe/GXe transition instead of having a hard cut at the gate. Therefore I improved the binning of the map (increased the number of bins to 60) before interpolating and extrapolating.

I tried to test this new map with the following command and the montecarlo scripts (which seems to not be successful):

# fax+pax run-dependent configuration
FAX_PAX_CONFIG="[WaveformSimulator]drift_field=82 * V / cm;truth_file_name=\"${FAX_FILENAME}\";diffusion_constant_liquid=${DIFFUSION_CONSTANT} * cm**2 / s;s1_patterns_file=\"${start_dir}/XENON1T_s1_xyz_patterns_Aug2017.json.gz\";s2_mean_area_fraction_top=0.6323;[DEFAULT]drift_velocity_liquid=${DRIFT_VELOCITY} * um / ns;electron_lifetime_liquid=${ELECTRON_LIFETIME} * us"

The S1 AFT still looks like on this wiki page. So I don't know if there is still an issue in FAX or if the new map wasn't loaded an thus the map look equal.

You can also see in the link how I generate the PAX map. I also miss an option in FAX which logs the used map or an indicator which tells me that a specific map was used ..

I will continue to try testing the map before making a PR.

pdeperio commented 7 years ago

can you try running locally or on Midway to confirm the map is getting loaded?

l-althueser commented 7 years ago

I found a way to load the map in the meantime. It is now in the dax catalogue and used by FAX/PAX. Have to wait for some jobs to finish until I can give an estimation if I solved this issue.

I have also added a drift field and S2 mean AFT value to the config string just to see what happens.

Update: The jobs still have errors. At the moment because of the new map which has empty bins according to pax.

pax.exceptions.CoordinateOutOfRangeException: Expected light pattern at coordinates (34.13744812011719, 33.26744384765625, -41.67567749023438) consists of only zeros!

Don't know how this could happen.

Update: Solved the map issues and the production is going well. Will be finished in some hours.