search

SND-LHC / sndsw

SND@LHC experiment framework based on FairShip and FairRoot
6 stars 16 forks source link

Updates start2024 data taking #212

Closed siilieva closed 6 months ago

siilieva commented 6 months ago

At the start of the emulsion run, the necessity of a few updates emerged - raw data paths etc. Then, there is the reverse board mapping that we use in the DAQ monitoring and which was missing the Veto3 support - I missed it in #210 since I only tested on MC and so with no DAQ monitoring. Also I had to fix the shipunit bug introduced when renaming snd specific parameters...

olantwin commented 6 months ago

Also I had to fix the shipunit bug introduced when renaming snd specific parameters...

Sorry, I should have spotted this in the original PR...

siilieva commented 6 months ago

Also I had to fix the shipunit bug introduced when renaming snd specific parameters...

Sorry, I should have spotted this in the original PR...

Well, I also missed it when renaming, good it is an quick and eazy fix.

olantwin commented 6 months ago

Btw I just noticed there is also shipdata/ShipUnit.h that we use in c++ classes. I'll add the snd paramters there as well before the next push.

I've been wondering for a while how to best handle this longer-term. It would be easy to load the C++ from python using PyROOT, so shipunit.py could be made a wrapper around the header. But there is also a scikit-hep package for units, so it might be nice to use that for our python, with minimal additions for SND@LHC specific things...

siilieva commented 6 months ago

Btw I just noticed there is also shipdata/ShipUnit.h that we use in c++ classes. I'll add the snd paramters there as well before the next push.

I've been wondering for a while how to best handle this longer-term. It would be easy to load the C++ from python using PyROOT, so shipunit.py could be made a wrapper around the header. But there is also a scikit-hep package for units, so it might be nice to use that for our python, with minimal additions for SND@LHC specific things...

For sure double definitions are nonsense. The wrapper solution seems nice. Would you mind creating an issue regarding this with the proposed solutions? Idk about the hepunits package. Units are quite important and what if there is a bug in that external package. And we generally need a few tens of units, that we already have.

olantwin commented 6 months ago

For sure double definitions are nonsense. The wrapper solution seems nice. Would you mind creating an issue regarding this with the proposed solutions?

Sure!

Idk about the hepunits package. Units are quite important and what if there is a bug in that external package. And we generally need a few tens of units, that we already have.

How about the following: we use our own, very limited set of units in the core software, but recommend using hepunits and particle in analysis, as they define many further things analysts might find useful, and we don't want to reimplement?

siilieva commented 6 months ago

For sure double definitions are nonsense. The wrapper solution seems nice. Would you mind creating an issue regarding this with the proposed solutions?

Sure!

Idk about the hepunits package. Units are quite important and what if there is a bug in that external package. And we generally need a few tens of units, that we already have.

How about the following: we use our own, very limited set of units in the core software, but recommend using hepunits and particle in analysis, as they define many further things analysts might find useful, and we don't want to reimplement?

I like this approach, yes.

As usual, grazie della collaborazione per PR :D

olantwin commented 6 months ago

Figurati!