forked from CMS-HGCAL/ntuple-tools
-
Notifications
You must be signed in to change notification settings - Fork 3
/
collections.py
1748 lines (1461 loc) · 66 KB
/
collections.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
"""
Manager of the ntuple data collections.
This module manages the code and the instances of objects actually
accessing the ntuple data.
Classes:
EventManager
DFCollection
TPSet
Objects:
all collections (DFCollection and TPSet instances) that
can be used by plotters.
"""
from __future__ import print_function
from __future__ import absolute_import
import pandas as pd
import numpy as np
import ROOT
import math
import sys
import root_numpy.tmva as rnptmva
from .utils import debugPrintOut
import python.clusterTools as clAlgo
from python.mp_pool import POOL
import python.classifiers as classifiers
import python.calibrations as calib
import python.pf_regions as pf_regions
from scipy.spatial import cKDTree
class WeightFile(object):
def __init__(self, file_name):
self.file_ = ROOT.TFile(file_name)
self.cache_ = {}
def get_weight_1d(self, histo_name, variable):
histo = None
if histo_name not in self.cache_.keys():
self.cache_[histo_name] = self.file_.Get(histo_name)
histo = self.cache_[histo_name]
bin_n = histo.FindBin(variable)
return histo.GetBinContent(bin_n)
class EventManager(object):
"""
EventManager.
Manages the registration of collections and reads them when needed,
i.e. only if they have been activated (typically by a plotter).
It is a singleton.
"""
class __TheManager:
def __init__(self):
self.collections = []
self.active_collections = []
self.weight_file = None
def read_weight_file(self, weight_file_name):
self.weight_file = WeightFile(weight_file_name)
def registerCollection(self, collection):
# print '[EventManager] registering collection: {}'.format(collection.name)
self.collections.append(collection)
def registerActiveCollection(self, collection):
print('[EventManager] registering collection as active: {}'.format(collection.name))
self.active_collections.append(collection)
def read(self, event, debug):
for collection in self.active_collections:
if debug >= 3:
print('[EventManager] filling collection: {}'.format(collection.name))
collection.fill(event, self.weight_file, debug)
def get_labels(self):
label_dict = {}
for col in self.collections:
label_dict[col.name] = col.label
return label_dict
instance = None
def __new__(cls):
if not EventManager.instance:
EventManager.instance = EventManager.__TheManager()
return EventManager.instance
def __getattr__(self, name):
return getattr(self.instance, name)
def __setattr__(self, name):
return setattr(self.instance, name)
class DFCollection(object):
"""
[DFCollection]: collection of objects consumed by plotters.
This class represents the DataFrame of the objects which need to be plotted.
The objects are registered with the EventManager at creation time but they
are actually created/read event by event only if one plotter object activates
them (or anotrher DFCollection depending on them) at booking time.
As a result you can instantiate as many objects as needed and only those
actually consumed by a plotter will be read.
Args:
name (string): name which enters the histo name
label (string): drawn on plots or legends
filler_function (callable): function accepting event
as parameter and returning the DataFrame filled
fixture_function (callable): adds columns to the data-frame
depends_on (list): specify the dependencies (typically called as argument by the filler_function at filling time)
debug (int): specify the debug level object by object (also the global one is used)
"""
def __init__(self, name, label,
filler_function,
fixture_function=None,
read_entry_block=1000,
depends_on=[],
debug=0,
print_function=lambda df: df,
max_print_lines=-1,
weight_function=None):
self.df = None
self.name = name
self.label = label
self.is_active = False
self.filler_function = filler_function
self.fixture_function = fixture_function
self.depends_on = depends_on
self.debug = debug
self.print_function = print_function
self.max_print_lines = max_print_lines
self.weight_function = weight_function
self.n_queries = 0
self.cached_queries = dict()
self.cached_entries = dict()
self.entries = None
self.empty_df = None
self.next_entry_read = 0
self.read_entry_block = read_entry_block
# print (f'Create collection: {self.name} with read_entry_block: {read_entry_block}')
self.new_read = False
self.new_read_nentries = 0
self.register()
def register(self):
event_manager = EventManager()
event_manager.registerCollection(self)
def activate(self):
if len(self.depends_on) > 0:
common_block_size = -1
for coll in self.depends_on:
if common_block_size == -1:
common_block_size = coll.read_entry_block
else:
if coll.read_entry_block != common_block_size:
raise ValueError(f'Collection {self.name} depends on collections with different common_block_size!')
if common_block_size != self.read_entry_block:
print(f'Collection {self.name}: common_block_size set to dependent value: {common_block_size}')
self.read_entry_block = common_block_size
if not self.is_active:
for dep in self.depends_on:
dep.activate()
self.is_active = True
event_manager = EventManager()
event_manager.registerActiveCollection(self)
return self.is_active
def fill(self, event, weight_file=None, debug=0):
stride = self.read_entry_block
# print (f'Coll: {self.name} fill for entry: {event.file_entry}')
if event.file_entry == 0 or event.file_entry == self.next_entry_read or event.global_entry == event.entry_range[0]:
# print ([self.read_entry_block, (event.entry_range[1]-event.global_entry), (event.tree.num_entries - event.file_entry)])
if event.entry_range[1] != -1:
stride = min([self.read_entry_block, (1+event.entry_range[1]-event.global_entry), (event.tree.num_entries - event.file_entry)])
else:
stride = min([self.read_entry_block, (event.tree.num_entries - event.file_entry)])
# print(f'[fill] stride: {stride}')
if stride == 0:
print('ERROR Last event????')
self.new_read = False
else:
self.new_read = True
self.next_entry_read = event.file_entry + stride
self.new_read_nentries = stride
self.fill_real(event, stride, weight_file, debug)
else:
self.new_read = False
if self.debug > 0:
df_print = self.empty_df
if event.file_entry in self.df.index.get_level_values('entry'):
df_print = self.df.loc[event.file_entry]
debugPrintOut(max(debug, self.debug), self.label,
toCount=df_print,
toPrint=self.print_function(df_print),
max_lines=self.max_print_lines)
def fill_real(self, event, stride, weight_file=None, debug=0):
self.clear_query_cache(debug)
self.df = self.filler_function(event, stride)
if self.fixture_function is not None:
# FIXME: wouldn't this be more efficient
# self.fixture_function(self.df)
self.df = self.fixture_function(self.df)
if self.weight_function is not None:
self.df = self.weight_function(self.df, weight_file)
self.empty_df = pd.DataFrame(columns=self.df.columns)
self.entries = self.df.index.get_level_values('entry').unique()
if debug > 2:
print(f'read coll. {self.name} from entry: {event.file_entry} to entry: {event.file_entry+stride} (stride: {stride}), # rows: {self.df.shape[0]}, # entries: {len(self.entries)}')
def query(self, selection):
self.n_queries += 1
if selection.all or self.df.empty:
return self.df
if selection.selection not in self.cached_queries:
ret = self.df.query(selection.selection)
self.cached_queries[sys.intern(selection.selection)] = ret
entries = ret.index.get_level_values('entry').unique()
self.cached_entries[selection.hash] = entries
return ret
return self.cached_queries[selection.selection]
def query_event(self, selection, idx):
self.n_queries += 1
# print (f'coll: {self.name}, query selection: {selection.selection} for entry: {idx}')
if idx not in self.entries:
# print (' enrty not in frame!')
return self.empty_df
if selection.all or self.df.empty:
# print (' frame is empty')
return self.df.loc[idx]
if selection.selection not in self.cached_queries:
# print (' query already cached')
ret = self.df.query(selection.selection)
self.cached_queries[sys.intern(selection.selection)] = ret
entries = ret.index.get_level_values('entry').unique()
self.cached_entries[selection.hash] = entries
if idx not in entries:
return self.empty_df
return ret.loc[idx]
# print (' query not cached')
# print (f' {self.cached_queries.keys()}')
entries = self.cached_entries[selection.hash]
if idx not in entries:
return self.empty_df
return self.cached_queries[selection.selection].loc[idx]
def clear_query_cache(self, debug=0):
if (debug > 5):
print('Coll: {} # queries: {} # unique queries: {}'.format(
self.name, self.n_queries, len(self.cached_queries.keys())))
self.n_queries = 0
self.cached_entries.clear()
self.cached_queries.clear()
def tkeg_fromcluster_fixture(tkegs):
# print tkegs
tkegs.loc[tkegs.hwQual == 1, 'hwQual'] = 3
return tkegs
# NOTE: scorporate the part wich computes the layer_weights
# (needed only by rthe calib plotters) from the rest (creating ad-hoc collections)
# this should also allow for removing the tc dependency -> huge speedup in filling
# FIXME: this needs to be ported to the new interface reading several entries at once
def cl3d_layerEnergy_hoe(clusters, tcs):
""" """
if clusters.empty:
return clusters
do_compute_hoe = False
do_compute_layer_energy = False
if 'hoe' not in clusters.columns:
do_compute_hoe = True
if 'layer_energy' not in clusters.columns:
do_compute_layer_energy = True
def compute_layer_energy(cluster, do_layer_energy=True, do_hoe=False):
components = tcs[tcs.id.isin(cluster.clusters)]
hist, bins = np.histogram(components.layer.values,
bins=range(0, 29, 2),
weights=components.energy.values)
results = []
if do_layer_energy:
results.append(hist)
if do_hoe:
em_energy = np.sum(hist)
hoe = -1
if em_energy != 0:
hoe = max(0, cluster.energy - em_energy)/em_energy
results.append(hoe)
return results
if do_compute_hoe or do_compute_layer_energy:
new_columns = []
if do_compute_layer_energy:
new_columns.append('layer_energy')
if do_compute_hoe:
new_columns.append('hoe')
clusters[new_columns] = clusters.apply(
lambda cl: compute_layer_energy(
cl,
do_compute_layer_energy,
do_compute_hoe),
result_type='expand',
axis=1)
return clusters
def cl3d_fixtures(clusters):
# print(clusters.columns)
# for backward compatibility
if clusters.empty:
# return
return clusters
clusters.rename(columns={'clusters_id': 'clusters',
'clusters_n': 'nclu'},
inplace=True)
# clusters['ptem'] = clusters.pt/(1+clusters.hoe)
# clusters['eem'] = clusters.energy/(1+clusters.hoe)
clusters['emax'] = clusters.emaxe*clusters.energy
clusters.loc[clusters.hoe == -1, ['hoe']] = 999
# print(clusters[clusters.hoe == -1][['hoe']])
clusters['bdt_pu'] = rnptmva.evaluate_reader(
classifiers.mva_pu_classifier_builder(),
'BDT',
clusters[['emax', 'emaxe', 'spptot', 'srrtot', 'ntc90']])
if False:
clusters['bdt_pi'] = rnptmva.evaluate_reader(
classifiers.mva_pi_classifier_builder(), 'BDT', clusters[['pt', 'eta', 'maxlayer', 'hoe', 'emaxe', 'szz']])
clusters['pt_em'] = clusters.apply(lambda x: x.ipt[x.name[1]][1], axis=1)
clusters.drop('ipt', axis=1, inplace=True)
clusters.drop('ienergy', axis=1, inplace=True)
clusters['meanz_scaled'] = np.abs(clusters.meanz) - 320
# return
return clusters
def gen_fixtures(particles, mc_particles):
if particles.empty:
return particles
# print particles.columns
particles['pdgid'] = particles.pid
particles['abseta'] = np.abs(particles.eta)
def get_mother_pdgid(particle, mc_particles):
if particle.gen == -1:
return -1
return mc_particles.df.loc[(particle.name[0], particle.gen-1)].firstmother_pdgid
particles['firstmother_pdgid'] = particles.apply(func=lambda x: get_mother_pdgid(x, mc_particles), axis=1)
return particles
def mc_fixtures(particles):
particles['firstmother'] = particles.index.to_numpy()
particles['firstmother_pdgid'] = particles.pdgid
return particles
# FIXME: this is broken
# print(particles)
for particle in particles.itertuples():
print(particle.daughters)
if particle.daughters == [[], []]:
continue
particles.loc[particle.daughters, 'firstmother'] = particle.Index
particles.loc[particle.daughters, 'firstmother_pdgid'] = particle.pdgid
return particles
def tc_fixtures(tcs):
# print tcs.columns
tcs['ncells'] = 1
if not tcs.empty:
tcs['cells'] = tcs.apply(func=lambda x: [int(x.id)], axis=1)
tcs['abseta'] = np.abs(tcs.eta)
# tcs['xproj'] = tcs.x/tcs.z
# tcs['yproj'] = tcs.y/tcs.z
return tcs
def cl2d_fixtures(clusters):
clusters['ncells'] = 1
if not clusters.empty:
clusters['ncells'] = [len(x) for x in clusters.cells]
return clusters
def tower_fixtures(towers):
if towers.empty:
# print '***[compute_tower_data]:WARNING input data-frame is empty'
return towers
towers.eval('HoE = etHad/etEm', inplace=True)
def fill_momentum(tower):
vector = ROOT.TLorentzVector()
vector.SetPtEtaPhiE(tower.pt, tower.eta, tower.phi, tower.energy)
tower.momentum = vector
# print tower.pt, tower.momentum.Pt()
return tower
towers['momentum'] = ROOT.TLorentzVector()
towers = towers.apply(fill_momentum, axis=1)
return towers
def recluster_mp(cl3ds, tcs, cluster_size, cluster_function, pool):
# FIXME: need to be ported to uproot multiindexing
# cluster_sides = [x for x in [cl3ds[cl3ds.eta > 0],
# cl3ds[cl3ds.eta < 0]]]
# tc_sides = [x for x in [tcs[tcs.eta > 0],
# tcs[tcs.eta < 0]]]
#
# cluster_sizes = [cluster_size, cluster_size]
#
# cluster_and_tc_sides = zip(cluster_sides, tc_sides, cluster_sizes)
#
# result_3dcl = pool.map(cluster_function, cluster_and_tc_sides)
#
# # result_3dcl = []
# # result_3dcl.append(cluster_function(cluster_and_tc_sides[0]))
# # result_3dcl.append(cluster_function(cluster_and_tc_sides[1]))
#
# # print result_3dcl[0]
# # print result_3dcl[1]
#
merged_clusters = pd.DataFrame(columns=cl3ds.columns)
# for res3D in result_3dcl:
# merged_clusters = merged_clusters.append(res3D, ignore_index=True, sort=False)
return merged_clusters
def get_cylind_clusters_mp(cl3ds, tcs, cylind_size, pool):
return recluster_mp(cl3ds, tcs,
cluster_size=cylind_size,
cluster_function=clAlgo.get_cylind_clusters_unpack,
pool=pool)
def get_dr_clusters_mp(cl3ds, tcs, dr_size, pool):
return recluster_mp(cl3ds, tcs,
cluster_size=dr_size,
cluster_function=clAlgo.get_dr_clusters_unpack,
pool=pool)
def get_dtdu_clusters_mp(cl3ds, tcs, dr_size, pool):
return recluster_mp(cl3ds, tcs,
cluster_size=dr_size,
cluster_function=clAlgo.get_dtdu_clusters_unpack,
pool=pool)
def get_emint_clusters(triggerClusters):
clusters_emint = triggerClusters.copy(deep=True)
def interpret(cluster):
cluster.energy = cluster.ienergy[-1]
cluster.pt = cluster.ipt[-1]
return cluster
clusters_emint = clusters_emint.apply(interpret, axis=1)
return clusters_emint
def get_merged_cl3d(triggerClusters, pool, debug=0):
# FIXME: need to be ported to uproot multiindexing
merged_clusters = pd.DataFrame(columns=triggerClusters.columns)
# if triggerClusters.empty:
# return merged_clusters
# # FIXME: filter only interesting clusters
# clusterSides = [x for x in [triggerClusters[triggerClusters.eta > 0],
# triggerClusters[triggerClusters.eta < 0]] if not x.empty]
#
# results3Dcl = pool.map(clAlgo.merge3DClustersEtaPhi, clusterSides)
# for res3D in results3Dcl:
# merged_clusters = merged_clusters.append(res3D, ignore_index=True, sort=False)
return merged_clusters
def get_trackmatched_egs(egs, tracks, debug=0):
entries = tracks.df.index.get_level_values('entry').union(egs.df.index.get_level_values('entry')).unique()
data = []
index = []
for entry in entries:
subentry = 0
kd_tree1 = cKDTree(egs.df.loc[entry][['eta', 'phi']])
kd_tree2 = cKDTree(tracks.df.loc[entry][['caloeta', 'calophi']])
sdm = kd_tree1.sparse_distance_matrix(kd_tree2, 0.2)
for key,dr in sdm.items():
eg_match = egs.df.loc[(entry, key[0])]
tk_match = tracks.df.loc[(entry, key[1])]
index.append((entry, subentry))
data_entry=[
eg_match.pt_em,
eg_match.eta,
eg_match.phi,
eg_match.quality,
eg_match.bdteg,
eg_match.bdt_pu,
tk_match.pt,
tk_match.eta,
tk_match.phi,
tk_match.caloeta,
tk_match.calophi,
tk_match.z0,
tk_match.chi2,
tk_match.chi2Red,
dr,
key[0],
key[1]]
data.append(data_entry)
subentry += 1
newcolumns = ['pt', 'eta', 'phi', 'quality', 'bdteg', 'bdt_pu']
newcolumns.extend(['tkpt', 'tketa', 'tkphi', 'tkcaloeta', 'tkcalophi', 'tkz0', 'tkchi2', 'tkchi2red', 'dr', 'clidx', 'tkidx'])
newindex = pd.MultiIndex.from_tuples(index, names=egs.df.index.names)
matched_egs = pd.DataFrame(data=data, columns=newcolumns, index=newindex)
return matched_egs
# print(entries)
# self.entries = self.df.index.get_level_values('entry').unique()
# if debug > 2:
# print(f'read coll. {self.name} from entry: {event.file_entry} to entry: {event.file_entry+stride} (stride: {stride}), # rows: {self.df.shape[0]}, # entries: {len(self.entries)}')
# def query(self, selection):
# self.n_queries += 1
# if selection.all or self.df.empty:
# return self.df
# if selection.selection not in self.cached_queries:
# ret = self.df.query(selection.selection)
# self.cached_queries[sys.intern(selection.selection)] = ret
# entries = ret.index.get_level_values('entry').unique()
# self.cached_entries[selection.hash] = entries
# return ret
# return self.cached_queries[selection.selection]
# def query_event(self, selection, idx):
# self.n_queries += 1
# FIXME: need to be ported to uproot multiindexing
#
#
# if egs.empty or tracks.empty:
# return matched_egs
# best_match_indexes, allmatches = match_etaphi(egs[['eta', 'phi']],
# tracks[['caloeta', 'calophi']],
# tracks['pt'],
# deltaR=0.1)
# for bestmatch_idxes in best_match_indexes.iteritems():
# # bestmatch_eg = egs.loc[bestmatch_idxes[0]]
# bestmatch_tk = tracks.loc[bestmatch_idxes[1]]
# matched_egs = matched_egs.append({'pt': bestmatch_eg.pt,
# 'energy': bestmatch_eg.energy,
# 'eta': bestmatch_eg.eta,
# 'phi': bestmatch_eg.phi,
# 'hwQual: bestmatch_eg.hwQual,
# 'tkpt': bestmatch_tk.pt,
# 'tketa': bestmatch_tk.eta,
# 'tkphi': bestmatch_tk.phi,
# 'tkz0': bestmatch_tk.z0,
# 'tkchi2': bestmatch_tk.chi2,
# 'tkchi2Red': bestmatch_tk.chi2Red,
# 'tknstubs': bestmatch_tk.nStubs,
# 'deta': bestmatch_tk.eta - bestmatch_eg.eta,
# 'dphi': bestmatch_tk.phi - bestmatch_eg.phi,
# 'dr': math.sqrt((bestmatch_tk.phi-bestmatch_eg.phi)**2+(bestmatch_tk.eta-bestmatch_eg.eta)**2)},
# ignore_index=True, sort=False)
return matched_egs
def get_layer_calib_clusters(input_clusters,
layer_calib_factors,
eta_corr=(0., 0.),
debug=False):
if debug:
print(layer_calib_factors)
print(eta_corr)
calibrated_clusters = input_clusters.copy(deep=True)
# def apply_calibration(cluster):
# cluster['energy'] = np.sum(np.array(cluster['layer_energy'])*np.array(layer_calib_factors))+eta_corr[1]+np.abs(cluster['eta'])*eta_corr[0]
# cluster['pt'] = cluster.energy/np.cosh(cluster.eta)
# return cluster
# calibrated_clusters = calibrated_clusters.apply(apply_calibration, axis=1)
if not calibrated_clusters.empty:
energies_lcalib = calibrated_clusters.layer_energy.apply(lambda x: np.dot(x, layer_calib_factors))
calibrated_clusters['energy'] = energies_lcalib+eta_corr[1]+np.abs(calibrated_clusters['eta'])*eta_corr[0]
calibrated_clusters['pt'] = calibrated_clusters.energy/np.cosh(calibrated_clusters.eta)
# print calibrated_clusters[['energy', 'energy_1']]
return calibrated_clusters
def get_calibrated_clusters(calib_factors, input_3Dclusters):
calibrated_clusters = input_3Dclusters.copy(deep=True)
def apply_calibration(cluster):
calib_factor = 1
calib_factor_tmp = calib_factors[(calib_factors.eta_l < abs(cluster.eta)) &
(calib_factors.eta_h >= abs(cluster.eta)) &
(calib_factors.pt_l <= cluster.pt) &
(calib_factors.pt_h > cluster.pt)]
if not calib_factor_tmp.empty:
# print 'cluster pt: {}, eta: {}, calib_factor: {}'.format(cluster.pt, cluster.eta, calib_factor_tmp.calib.values[0])
# print calib_factor_tmp
calib_factor = 1./calib_factor_tmp.calib.values[0]
# print cluster
# else:
# if cluster.eta <= 2.8 and cluster.eta > 1.52 and cluster.pt > 4 and cluster.pt <= 100:
# print cluster[['pt', 'eta']]
cluster.pt = cluster.pt*calib_factor
# cluster['pt1'] = cluster.pt*calib_factor
# cluster['cf'] = 1./calib_factor
return cluster
# input_3Dclusters[(input_3Dclusters.eta_l > abs(cluster.eta)) & ()]
calibrated_clusters = calibrated_clusters.apply(apply_calibration, axis=1)
return calibrated_clusters
def build3DClusters(name, algorithm, triggerClusters, pool, debug):
trigger3DClusters = pd.DataFrame()
# FIXME: need to be ported to uproot multiindexing
#
# if triggerClusters.empty:
# return trigger3DClusters
# clusterSides = [x for x in [triggerClusters[triggerClusters.eta > 0],
# triggerClusters[triggerClusters.eta < 0]] if not x.empty]
# results3Dcl = pool.map(algorithm, clusterSides)
# for res3D in results3Dcl:
# trigger3DClusters = trigger3DClusters.append(res3D, ignore_index=True, sort=False)
#
# debugPrintOut(debug, name='{} 3D clusters'.format(name),
# toCount=trigger3DClusters,
# toPrint=trigger3DClusters.iloc[:3])
return trigger3DClusters
# trigger3DClustersDBS = build3DClusters(
# 'DBS', clAlgo.build3DClustersEtaPhi, triggerClustersDBS, pool, debug)
# trigger3DClustersDBSp = build3DClusters(
# 'DBSp', clAlgo.build3DClustersProjTowers, triggerClustersDBS, pool, debug)
# trigger3DClustersP = build3DClusters(
# 'DEFp', clAlgo.build3DClustersProjTowers, triggerClusters, pool, debug)
def get_calibrated_clusters2(calib_factors, input_3Dclusters):
calibrated_clusters = input_3Dclusters.copy(deep=True)
def apply_calibration(cluster):
calib_factor = 1.
calib_factor_tmp = calib_factors[(calib_factors.eta_l < abs(cluster.eta)) &
(calib_factors.eta_h >= abs(cluster.eta)) &
(calib_factors.pt_l < cluster.pt) &
(calib_factors.pt_h >= cluster.pt)]
if not calib_factor_tmp.empty:
# print 'cluster pt: {}, eta: {}, calib_factor: {}'.format(cluster.pt, cluster.eta, calib_factor_tmp.calib.values[0])
# print calib_factor_tmp
calib_factor = 1./calib_factor_tmp.calib.values[0]
# print cluster
else:
if cluster.eta <= 2.8 and cluster.eta > 1.52 and cluster.pt > 4 and cluster.pt <= 100:
print(cluster[['pt', 'eta']])
cluster['pt2'] = cluster.pt*calib_factor
return cluster
# input_3Dclusters[(input_3Dclusters.eta_l > abs(cluster.eta)) & ()]
calibrated_clusters = calibrated_clusters.apply(apply_calibration, axis=1)
return calibrated_clusters
def select_and_merge_collections(barrel, endcap):
barrel_final = barrel
endcap_final = endcap
if not endcap.empty:
endcap_final = endcap[endcap.hwQual == 5]
return merge_collections(
barrel=barrel_final,
endcap=endcap_final)
def merge_collections(barrel, endcap):
if barrel.empty:
return endcap
return barrel.append(endcap, ignore_index=True)
def barrel_quality(electrons):
if electrons.empty:
# electrons.columns = ['pt', 'energy', 'eta', 'phi', 'hwQual']
return electrons
hwqual = pd.to_numeric(electrons['hwQual'], downcast='integer')
electrons['looseTkID'] = ((hwqual.values >> 1) & 1) > 0
electrons['photonID'] = ((hwqual.values >> 2) & 1) > 0
return electrons
def fake_endcap_quality(electrons):
# just added for compatibility with barrel
electrons['looseTkID'] = True
electrons['photonID'] = True
return electrons
def tkele_fixture_ee(electrons):
electrons['looseTkID'] = True
electrons['photonID'] = True
electrons['dpt'] = electrons.tkPt - electrons.pt
return electrons
def tkele_fixture_eb(electrons):
# if electrons.empty:
# # electrons.columns = ['pt', 'energy', 'eta', 'phi', 'hwQual']
# return electrons
hwqual = pd.to_numeric(electrons['hwQual'], downcast='integer')
electrons['looseTkID'] = ((hwqual.values >> 1) & 1) > 0
electrons['photonID'] = ((hwqual.values >> 2) & 1) > 0
electrons['dpt'] = electrons.tkPt - electrons.pt
return electrons
def quality_flags(objs):
hwqual = pd.to_numeric(objs['hwQual'], downcast='integer')
mask_tight_sta = 0b0001
mask_tight_ele = 0b0010
mask_tight_pho = 0b0100
mask_no_brem = 0b1000
objs['IDTightSTA'] = np.bitwise_and(hwqual.values, mask_tight_sta) > 0
objs['IDTightEle'] = np.bitwise_and(hwqual.values, mask_tight_ele) > 0
objs['IDTightPho'] = np.bitwise_and(hwqual.values, mask_tight_pho) > 0
objs['IDNoBrem'] = np.bitwise_and(hwqual.values, mask_no_brem) > 0
objs['IDBrem'] = np.bitwise_and(hwqual.values, mask_no_brem) == 0
return objs
def quality_ele_fixtures(objs):
objs['dpt'] = objs.tkPt - objs.pt
return quality_flags(objs)
def print_columns(df):
print(df.columns)
return df
def gen_part_pt_weights(gen_parts, weight_file):
def compute_weight(gen_part):
return weight_file.get_weight_1d('h_weights', gen_part.pt)
if weight_file is None:
gen_parts['weight'] = 1
else:
gen_parts['weight'] = gen_parts.apply(compute_weight, axis=1)
return gen_parts
def map2pfregions(objects, eta_var, phi_var, fiducial=False):
for ieta in range(0, pf_regions.regionizer.n_eta_regions()):
objects['eta_reg_{}'.format(ieta)] = False
for iphi in range(0, pf_regions.regionizer.n_phi_regions()):
objects['phi_reg_{}'.format(iphi)] = False
for ieta, eta_range in enumerate(pf_regions.regionizer.get_eta_boundaries(fiducial)):
query = '({} > {}) & ({} <= {})'.format(
eta_var,
eta_range[0],
eta_var,
eta_range[1]
)
region_objects = objects.query(query).index
objects.loc[region_objects, ['eta_reg_{}'.format(ieta)]] = True
for iphi, phi_range in enumerate(pf_regions.regionizer.get_phi_boundaries(fiducial)):
query = '({} > {}) & ({} <= {})'.format(
phi_var,
phi_range[0],
phi_var,
phi_range[1])
region_objects = objects.query(query).index
objects.loc[region_objects, ['phi_reg_{}'.format(iphi)]] = True
return objects
def maptk2pfregions_in(objects):
return map2pfregions(objects, 'caloeta', 'calophi', fiducial=False)
def mapcalo2pfregions_in(objects):
return map2pfregions(objects, 'eta', 'phi', fiducial=False)
def mapcalo2pfregions_out(objects):
return map2pfregions(objects, 'eta', 'phi', fiducial=True)
def decodedTk_fixtures(objects):
objects['deltaZ0'] = objects.z0 - objects.simz0
objects['deltaPt'] = objects.pt - objects.simpt
objects['deltaEta'] = objects.eta - objects.simeta
objects['deltaCaloEta'] = objects.caloeta - objects.simcaloeta
# have dphi between -pi and pi
comp_remainder = np.vectorize(math.remainder)
objects['deltaCaloPhi'] = comp_remainder(objects.calophi - objects.simcalophi, 2*np.pi)
objects['abseta'] = np.abs(objects.eta)
objects['simabseta'] = np.abs(objects.simeta)
return objects
def build_double_obj(obj):
ret = pd.concat([obj,obj], names=['leg'], keys=[0, 1])
ret = ret.swaplevel(0,1)
ret = ret.swaplevel(1,2)
return ret
calib_mgr = calib.CalibManager()
gen = DFCollection(
name='GEN', label='GEN particles',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='gen', entry_block=entry_block),
fixture_function=mc_fixtures,
print_function=lambda df: df[['pdgid', 'pt', 'eta', 'phi', 'daughters']],
max_print_lines=None,
debug=0)
sim_parts = DFCollection(
name='SIM', label='SIM particles',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='simpart', entry_block=entry_block),
fixture_function=lambda gen_parts: gen_fixtures(gen_parts, gen),
# read_entry_block=10,
depends_on=[gen],
debug=0,
# print_function=lambda df: df[['eta', 'phi', 'pt', 'energy', 'mother', 'fbrem', 'ovz', 'pid', 'gen', 'reachedEE', 'firstmother_pdgid']],
print_function=lambda df: df[['gen', 'pid', 'pt', 'eta', 'phi', 'mother', 'reachedEE', 'ovz', 'dvz', ]],
max_print_lines=None,
# print_function=lambda df: df.columns,
weight_function=gen_part_pt_weights)
gen_parts = DFCollection(
name='GEN', label='GEN particles',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='genpart', entry_block=entry_block),
fixture_function=lambda gen_parts: gen_fixtures(gen_parts, gen),
# read_entry_block=10,
depends_on=[gen],
debug=0,
# print_function=lambda df: df[['eta', 'phi', 'pt', 'energy', 'mother', 'fbrem', 'ovz', 'pid', 'gen', 'reachedEE', 'firstmother_pdgid']],
print_function=lambda df: df[['gen', 'pid', 'pt', 'eta', 'phi', 'mother', 'ovz', 'dvz', 'reachedEE']].sort_values(by='mother', ascending=False),
max_print_lines=None,
# print_function=lambda df: df.columns,
weight_function=gen_part_pt_weights)
tcs = DFCollection(
name='TC', label='Trigger Cells',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='tc', entry_block=entry_block),
read_entry_block=200,
fixture_function=tc_fixtures, debug=0)
tcs_truth = DFCollection(
name='TCTrue', label='Trigger Cells True',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='tctruth', entry_block=entry_block),
fixture_function=tc_fixtures)
cl2d_def = DFCollection(
name='DEF2D', label='dRC2d',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='cl', entry_block=entry_block),
fixture_function=cl2d_fixtures)
cl2d_truth = DFCollection(
name='DEF2DTrue', label='dRC2d True',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='cltruth', entry_block=entry_block),
fixture_function=cl2d_fixtures)
cl3d_truth = DFCollection(
name='HMvDRTrue', label='HM+dR(layer) True Cl3d',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='cl3dtruth', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters),
debug=0)
cl3d_def = DFCollection(
name='DEF', label='dRC3d',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='cl3d', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters))
cl3d_def_nc = DFCollection(
name='DEFNC', label='dRC3d NewTh',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='cl3dNC', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters))
cl3d_hm = DFCollection(
name='HMvDR', label='HM+dR(layer) Cl3d',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='HMvDR', entry_block=entry_block, fallback='hmVRcl3d'),
fixture_function=lambda clusters: cl3d_fixtures(clusters),
read_entry_block=200,
debug=0,
print_function=lambda df: df[['id', 'energy', 'pt', 'eta', 'phi', 'quality', 'ienergy', 'ipt']].sort_values(by='pt', ascending=False))
# cl3d_hm.activate()
cl3d_hm_emint = DFCollection(
name='HMvDREmInt', label='HM+dR(layer) Cl3d EM Int',
filler_function=lambda event, entry_block: get_emint_clusters(cl3d_hm.df[cl3d_hm.df.quality > 0]),
# fixture_function=lambda clusters: cl3d_fixtures(clusters, tcs.df),
depends_on=[cl3d_hm],
debug=0,
print_function=lambda df: df[['id', 'energy', 'pt', 'eta', 'quality', 'ienergy', 'ipt']].sort_values(by='pt', ascending=False)[:10])
cl3d_hm_emint_merged = DFCollection(
name='HMvDREmIntMerged', label='HM+dR(layer) Cl3d EM Int Merged',
filler_function=lambda event, entry_block: get_merged_cl3d(cl3d_hm_emint.df[cl3d_hm.df.quality >= 0], POOL),
# fixture_function=lambda clusters: cl3d_fixtures(clusters, tcs.df),
depends_on=[cl3d_hm_emint],
debug=0,
print_function=lambda df: df[['id', 'energy', 'pt', 'eta', 'quality', 'ienergy', 'ipt']])
# cl3d_hm_emint.activate()
cl3d_hm_rebin = DFCollection(
name='HMvDRRebin', label='HM+dR(layer) rebin Cl3d ',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='hmVRcl3dRebin', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters))
cl3d_hm_stc = DFCollection(
name='HMvDRsTC', label='HM+dR(layer) SuperTC Cl3d ',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='hmVRcl3dSTC', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters))
cl3d_hm_nc0 = DFCollection(
name='HMvDRNC0', label='HM+dR(layer) Cl3d + NewTh0',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='hmVRcl3dNC0', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters))
cl3d_hm_nc1 = DFCollection(
name='HMvDRNC1', label='HM+dR(layer) Cl3d + NewTh1',
filler_function=lambda event, entry_block: event.getDataFrame(
prefix='hmVRcl3dNC1', entry_block=entry_block),
fixture_function=lambda clusters: cl3d_fixtures(clusters))
cl3d_def_merged = DFCollection(
name='DEFMerged', label='dRC3d merged',
filler_function=lambda event, entry_block: get_merged_cl3d(cl3d_def.df[cl3d_def.df.quality > 0], POOL),
depends_on=[cl3d_def])
cl3d_def_calib = DFCollection(
name='DEFCalib', label='dRC3d calib.',
filler_function=lambda event, entry_block: get_calibrated_clusters(calib.get_calib_factors(), cl3d_def.df),
depends_on=[cl3d_def])
cl3d_hm_merged = DFCollection(
name='HMvDRMerged', label='HM+dR(layer) merged',
filler_function=lambda event, entry_block: get_merged_cl3d(cl3d_hm.df[cl3d_hm.df.quality >= 0], POOL),
depends_on=[cl3d_hm])
cl3d_hm_fixed = DFCollection(
name='HMvDRfixed', label='HM fixed',
filler_function=lambda event, entry_block: get_cylind_clusters_mp(cl3d_hm.df, tcs.df, [999.]*52, POOL),
depends_on=[cl3d_hm, tcs], debug=0)
cl3d_hm_cylind10 = DFCollection(
name='HMvDRcylind10', label='HM Cylinder 10cm',