Run-3: Copied Regrouped Unc and Added L2Relative for PNet and UParT

[ravindkv]

This MR is intended to copy the "Regrouped" uncertainty set [*] AND include the currently available L2Relative JECs for PNET/UParT for most of the years of Run-3

• [Summer22_22Sep2023_V4]: Regrouped unc

• [Summer22EE_22Sep2023_V4]: Regrouped unc

• [Summer23Prompt23_V4]: Regrouped unc

• [Summer23BPixPrompt23_V4]: Regrouped unc

• [Summer24Prompt24_V3_V4]: Regrouped unc, L2Relative for PNET and UParT, also added AK4PFChs for consistency. The Residual JECs for PNET and UParT are clonned from regular jets.

Except these, all other JECs are exactly the same (symlinked to the previous versions)

cc: @izisopou

[*] originally copied from from Summer19UL18, we just change the year string name

[izisopou]

Thanks @ravindkv , looks good.

I have one comment:

• We are currently missing UParT MC-truth JECs for 2022/2023. I recall that Jessy has derived them but he never opened an MR in the JERCProtolab so I pinged him. Once he does and we review them, you could add them in this PR.

I also have two questions:

• I see that in 2024 you added AK4PFchs JECs for consistency, cloned from AK4PFPuppi. However, these JECs would have a very bad performance on CHS jets because CHS jets have a lot of pileup and PUPPI jets do not have L1FastJet JECS. I believe it would be better to clone the CHS JECs from Summer23BPix instead. Or if 2024 is closer to 2025, then clone from the Winter25 CHS JECs. What do you think?
• I see you uploaded db and tar files for DATA tags in 2022 and 2023. However, the corresponding text have not changed in this PR. Are these new files the same as before?

[izisopou]

Hi @ravindkv , could you also consider removing the L2Residual files, as we had agreed to stop including them from now on? Thanks.