Fix: None padding order, adding otherjets in Wc WF #60
Conversation
There was a problem hiding this comment.
Hi @mondalspandan thank you very much for the changes and also find out the type bug!
Generally I have one comment and two question
💬 I would suggest all the boolean should do ak.fill_none(boolean,False) to remove None values
❓ I had the impression you want to store the flag (isMuJet) in the tree instead of the skimmed object(MuonJet, OtherJets), is this still the case?
❓ Did you plan to include the LHE NJet cut for the jet split sample? I have an idea the Njet split can be triggered when checking through all the dataset in runner step.
| jet_sel = jet_id(events, self._campaign) & ( | ||
| ak.all(events.Jet.metric_table(iso_muon) > 0.5, axis=2) | ||
| ) |
There was a problem hiding this comment.
I think the additional ak.fill_none for the selection to avoid None in the jet selection. (just to be sure None filtered out 😉 )
| ) | ||
| ) | ||
| mujetsel = ( | ||
| (ak.all(event_jet.metric_table(soft_muon) <= 0.4, axis=2)) |
There was a problem hiding this comment.
The mask_identity=True should be there to avoid None events pass True criteria.
The ak.fill_none(mujetsel,False,axis=-1) then remove None, this cut can be proceed anywhere. (Then you don't need mujetsel2)
| @@ -518,7 +513,7 @@ def process_shift(self, events, shift_name): | |||
| out_branch, | |||
| np.where( | |||
| (out_branch == "SoftMuon") | |||
| | (out_branch == "MuonJet") | |||
| # | (out_branch == "MuonJet") | |||
There was a problem hiding this comment.
Did I understand correctly you want to store PFcands in the MuonJet collection so you commented out?(also OtherJets also not included)
| @@ -528,7 +523,7 @@ def process_shift(self, events, shift_name): | |||
| "Muon", | |||
| "Jet", | |||
| "SoftMuon", | |||
| "MuonJet", | |||
| # "MuonJet", | |||
There was a problem hiding this comment.
The kinematics variables will not save, is this done intentionally?
| @@ -544,6 +539,7 @@ def process_shift(self, events, shift_name): | |||
| out_branch, ["Jet_btagDeep*", "Jet_DeepJet*", "PFCands_*", "SV_*"] | |||
| ) | |||
| # write to root files | |||
| print("Branches to write:", out_branch) | |||
There was a problem hiding this comment.
Is this the check you want to keep when running the code?
Indeed the out_branch also contains some regexp which not really returns full variable. Maybe would be interesting to check with fout["Events"].fields for check all the entries
| pruned_ev["MuonJet"] = smuon_jet | ||
| pruned_ev["SoftMuon"] = ssmu | ||
| pruned_ev["OtherJets"] = sotherjets |
There was a problem hiding this comment.
Do you still want to store the jet flag instead of collections on MuonJet & OtherJet? The mujetsel can still considered as a flag in the Jet variables
| @@ -523,7 +514,7 @@ def process_shift(self, events, shift_name): | |||
| "Muon", | |||
| "Jet", | |||
| "SoftMuon", | |||
| "MuonJet", | |||
| # "MuonJet", | |||
There was a problem hiding this comment.
The kinematics variables will not save, is this done intentionally?
| @@ -539,6 +530,7 @@ def process_shift(self, events, shift_name): | |||
| out_branch, ["Jet_btagDeep*", "Jet_DeepJet*", "PFCands_*", "SV_*"] | |||
| ) | |||
| # write to root files | |||
| print("Branches to write:", out_branch) | |||
There was a problem hiding this comment.
Is this the check you want to keep when running the code?
Indeed the out_branch also contains some regexp which not really returns full variable. Maybe would be interesting to check with fout["Events"].fields for check all the entries
| | (out_branch == "MuonJet") | ||
| # | (out_branch == "MuonJet") |
There was a problem hiding this comment.
Did I understand correctly you want to store PFcands in the MuonJet collection so you commented out?(also OtherJets also not included)
| pruned_ev["Jet"] = sjets | ||
| pruned_ev["Muon"] = shmu | ||
| pruned_ev["MuonJet"] = smuon_jet | ||
| pruned_ev["OtherJets"] = sotherjets |
There was a problem hiding this comment.
Do you still want to store the jet flag instead of collections on MuonJet & OtherJet? The mujetsel can still considered as a flag in the Jet variables
Ming-Yan
changed the title
|
@mondalspandan and @Ming-Yan what is the current status of this PR? |
|
Hi, I stopped working on it because my base branch diverged. Ming-Yan made a PR after this. @Ming-Yan Did you also fix the None padding issue or is it still present in the master? |
Hi @mondalspandan , indeed I fix the padding issue in the master PR. |
|
Then I think this PR can be deleted :) |
No description provided.