forked from mbandrews/MLAnalyzer
/
convert_Tree2Dask_EBcropsv3.py
277 lines (249 loc) · 10.2 KB
/
convert_Tree2Dask_EBcropsv3.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
import numpy as np
import ROOT
from root_numpy import tree2array, root2array
from dask.delayed import delayed
import dask.array as da
import glob
eosDir='/eos/uscms/store/user/mba2012/IMGs/DoublePi0Pt50To60'
#decays = ['DoublePi0Pt30To90_pythia8_m000_2016_25ns_Moriond17MC_PoissonOOTPU']
decays = [
'DoublePhotonPt50To60_pythia8_2016_25ns_Moriond17MC_PoissonOOTPU',
'DoublePi0Pt50To60_m000_pythia8_2016_25ns_Moriond17MC_PoissonOOTPU',
'DoublePi0Pt50To60_m0To1600_pythia8_2016_25ns_Moriond17MC_PoissonOOTPU'
]
chunk_size_ = 500
scale = 1.
@delayed
def load_X(tree, start_, stop_, branches_, readouts, scale):
#X = tree2array(tree, start=start_, stop=stop_, branches=branches_)
X = root2array(tree, treename='fevt/RHTree', start=start_, stop=stop_, branches=branches_)
# Convert the object array X to a multidim array:
# 1: for each event x in X, concatenate the object columns (branches) into a flat array of shape (readouts*branches)
# 2: reshape the flat array into a stacked array: (branches, readouts)
# 3: embed each stacked array as a single row entry in a list via list comprehension
# 4: convert this list into an array with shape (events, branches, readouts)
X = np.array([np.concatenate(x).reshape(len(branches_),readouts[0]*readouts[1]) for x in X])
#print "X.shape:",X.shape
X = X.reshape((-1,len(branches_),readouts[0],readouts[1]))
X = np.transpose(X, [0,2,3,1])
# Rescale
X /= scale
return X
@delayed
def load_single(tree, start_, stop_, branches_):
#X = tree2array(tree, start=start_, stop=stop_, branches=branches_)
X = root2array(tree, treename='fevt/RHTree', start=start_, stop=stop_, branches=branches_)
if len(branches_) > 1:
X = np.array([np.concatenate(x).reshape(len(branches_),1) for x in X])
X = X.reshape((-1,len(branches_)))
else:
X = np.array([x[0] for x in X])
return X
for j,decay in enumerate(decays):
if j != 1:
pass
continue
#tfiles = glob.glob('%s/%s_AODSIM_IMGcrop*.root'%(eosDir,decay))
tfiles = glob.glob('%s/%s_IMGcrop*.root'%(eosDir,decay))
print " >> %d files found."%len(tfiles)
tree = ROOT.TChain("fevt/RHTree")
for f in tfiles:
tree.Add(f)
nevts = tree.GetEntries()
tree = tfiles
neff = (nevts//1000)*1000
#neff = int(nevts)
#neff = 112000
chunk_size = chunk_size_
#chunk_size = int(nevts)
if neff > nevts:
neff = int(nevts)
chunk_size = int(nevts)
#neff = 1000
#neff = 233000
print " >> Doing decay:", decay
print " >> Input file[0]:", tfiles[0]
print " >> Total events:", nevts
print " >> Effective events:", neff
# EB
readouts = [170,360]
branches = ["EB_energy"]
X = da.concatenate([\
da.from_delayed(\
load_X(tree,i,i+chunk_size, branches, readouts, scale),\
shape=(chunk_size, readouts[0], readouts[1], len(branches)),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", X.shape
# SC0
readouts = [32,32]
branches = ["SC_energy0"]
X_crop0 = da.concatenate([\
da.from_delayed(\
load_X(tree,i,i+chunk_size, branches, readouts, scale),\
shape=(chunk_size, readouts[0], readouts[1], len(branches)),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", X_crop0.shape
# SC1
readouts = [32,32]
branches = ["SC_energy1"]
X_crop1 = da.concatenate([\
da.from_delayed(\
load_X(tree,i,i+chunk_size, branches, readouts, scale),\
shape=(chunk_size, readouts[0], readouts[1], len(branches)),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", X_crop1.shape
X_crop0 = da.concatenate([X_crop0, X_crop1], axis=0)
# SC0
readouts = [32,32]
branches = ["SC_energyT0", "SC_energyZ0"]
X_crop_stack0 = da.concatenate([\
da.from_delayed(\
load_X(tree,i,i+chunk_size, branches, readouts, scale),\
shape=(chunk_size, readouts[0], readouts[1], len(branches)),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", X_crop_stack0.shape
# SC1
readouts = [32,32]
branches = ["SC_energyT1", "SC_energyZ1"]
X_crop_stack1 = da.concatenate([\
da.from_delayed(\
load_X(tree,i,i+chunk_size, branches, readouts, scale),\
shape=(chunk_size, readouts[0], readouts[1], len(branches)),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", X_crop_stack1.shape
X_crop_stack0 = da.concatenate([X_crop_stack0, X_crop_stack1], axis=0)
# SC_mass0
branches = ["SC_mass0"]
y_mass0 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", y_mass0.shape
# SC_pT0
branches = ["SC_pT0"]
y_pT0 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", y_pT0.shape
# SC_mass1
branches = ["SC_mass1"]
y_mass1 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", y_mass1.shape
y_mass0 = da.concatenate([y_mass0, y_mass1], axis=0)
# SC_pT1
branches = ["SC_pT1"]
y_pT1 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", y_pT1.shape
y_pT0 = da.concatenate([y_pT0, y_pT1], axis=0)
# SC_DR0
branches = ["SC_DR0"]
y_DR0 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", y_DR0.shape
## SC1
#readouts = [32,32]
#branches = ["SC_energy1"]
#X_crop1 = da.concatenate([\
# da.from_delayed(\
# load_X(tree,i,i+chunk_size, branches, readouts, scale),\
# shape=(chunk_size, readouts[0], readouts[1], len(branches)),\
# dtype=np.float32)\
# for i in range(0,neff,chunk_size)])
#print " >> Expected shape:", X_crop1.shape
# pho_pT0
branches = ["pho_pT0"]
pho_pT0 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", pho_pT0.shape
# pho_E0
branches = ["pho_E0"]
pho_E0 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", pho_E0.shape
# pho_eta0
branches = ["pho_eta0"]
pho_eta0 = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.float32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", pho_eta0.shape
# eventId
branches = ["eventId"]
eventId = da.concatenate([\
da.from_delayed(\
load_single(tree,i,i+chunk_size, branches),\
shape=(chunk_size,),\
dtype=np.int32)\
for i in range(0,neff,chunk_size)])
print " >> Expected shape:", eventId.shape
## Kinematics
#branches = ["pho_pT", "pho_E", "pho_eta", "pho_phi"]
#X_p4 = da.concatenate([\
# da.from_delayed(\
# load_single(tree,i,i+chunk_size, branches),\
# shape=(chunk_size,len(branches)),\
# dtype=np.float32)\
# for i in range(0,neff,chunk_size)])
#print " >> Expected shape:", X_p4.shape
# Class label
label = j
label = 0
print " >> Class label:",label
y = da.from_array(\
np.full(X.shape[0], label, dtype=np.float32),\
chunks=(chunk_size,))
#file_out_str = "%s/%s_IMG_RH%d_n%dk_label%d.hdf5"%(eosDir,decay,int(scale),neff//1000.,label)
file_out_str = "%s/%s_IMGcrop_RH%d_n%dkx2.hdf5"%(eosDir,decay,int(scale),neff//1000.)
#file_out_str = "test.hdf5"
print " >> Writing to:", file_out_str
#da.to_hdf5(file_out_str, {'/X': X, '/y': y, 'eventId': eventId, 'X_crop0': X_crop0, 'X_crop1': X_crop1}, compression='lzf')
da.to_hdf5(file_out_str, {
#'/X': X,
'/y': y,
#'eventId': eventId,
'X_crop0': X_crop0,
'X_crop_stack0': X_crop_stack0,
#'X_crop1': X_crop1
#'X_p4': X_p4
'y_mass': y_mass0,
'y_pT': y_pT0,
#'y_DR': y_DR0,
#'pho_pT0': pho_pT0,
#'pho_E0': pho_E0,
#'pho_eta0': pho_eta0
}, compression='lzf')
print " >> Done.\n"