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
|
from ROOT import TGraph, TGraphErrors
#import sys
import csv
from datetime import date
class Datapoint:
def __init__(self,x,y,dy):
self._x = x
self._y = y
self._dy = dy
class Dataset:
""" Reads performance data from a table: "version algo source timing uncertainty"
"""
def addPoint(self,x,y,dy):
self._data.append( Datapoint(x,y,dy) )
if x < self._minkey : self._minkey = x
if x > self._maxkey : self._maxkey = x
if y < self._minvalue : self._minvalue = y
if y > self._maxvalue : self._maxvalue = y
def __init__(self, shape, algo, impl):
""" constructor creates the data series"""
self._shape = shape
self._algo = algo
self._impl = impl
self._data = []
self._minkey = 1.0e16
self._maxkey = -1.0e16
self._minvalue = 1.0e16
self._maxvalue = -1.0e16
def tgraph(self):
""" returns data points in a TGraphErrors object."""
if(len(self._data)==0):
print "***** Dataset.tgraph(): ERROR: no data available for", self._algo+'_'+self._impl
return TGraphErrors()
result = TGraphErrors( len(self._data) )
i = 0
#for k in sorted(self._data.iterkeys()):
for k in self._data:
x = k._x
y = k._y
#if y < self._minvalue : self._minvalue = y
#if y > self._maxvalue : self._maxvalue = y
#print "raw %s %s %s: i=%d value=%f minVal=%f maxVal=%f" % (self._shape, self._algo, self._impl, i, k, self._minvalue, self._maxvalue)
result.SetPoint( i, x, y )
result.SetPointError( i, 0, k._dy )
i = i + 1;
return result
def tgraphNorm(self):
""" produces a timing series normalized to value=1.0 at the first data point"""
if(len(self._data)==0):
print "***** Dataset.tgraphNorm(): ERROR: no data available for", self._algo+'_'+self._impl
return TGraph()
result = TGraph( len(self._data) )
normFactor = 1. / self._data[0]._y
i = 0
for k in self._data:
normValue = k._y * normFactor
if normValue < self._minvalue : self._minvalue = normValue
if normValue > self._maxvalue : self._maxvalue = normValue
#print "Normalizing %s: i=%d days=%d val=%f normFactor=%f stored=%f minVal=%f maxVal=%f" % (self._name, i, k, self.serie[k], normFactor, normValue, self._minvalue, self._maxvalue)
result.SetPoint( i, k._x, normValue )
i = i + 1;
return result
def tgraphNormError(self):
""" returns a TGraphErrors object normalized to value=1.0 at the first data point"""
if(len(self._data)==0):
print "***** Dataset.tgraphNormError(): ERROR: no data available for", self._algo+'_'+self._impl
return TGraphErrors()
result = TGraphErrors( len(self._data) )
normFactor = 1. / self._data[0]._y
i = 0
for k in self._data:
normValue = k._y * normFactor
if normValue < self._minvalue : self._minvalue = normValue
if normValue > self._maxvalue : self._maxvalue = normValue
#print "Normalizing %s: i=%d days=%d val=%f normFactor=%f stored=%f minVal=%f maxVal=%f" % (self._name, i, k, self.serie[k], normFactor, normValue, self._minvalue, self._maxvalue)
result.SetPoint( i, k._x, normValue )
result.SetPointError( i, 0, k._dy*normFactor )
i = i + 1;
return result
def tgraphRatio(self, refDataset):
""" produces timing ratio graphs"""
if(len(self._data)==0):
print "***** Dataset.tgraphNorm(): ERROR: no data available for", self._algo+'_'+self._impl
return TGraphErrors()
result = TGraph( len(self._data) )
for i in range(len(self._data)):
ratioValue = self._data[i]._y / refDataset._data[i]._y
result.SetPoint( i, self._data[i]._x, ratioValue )
if ratioValue < self._minvalue : self._minvalue = ratioValue
if ratioValue > self._maxvalue : self._maxvalue = ratioValue
#print "Normalizing %s: i=%d days=%d val=%f normFactor=%f stored=%f minVal=%f maxVal=%f" % (self._name, i, k, self.serie[k], normFactor, normValue, self._minvalue, self._maxvalue)
return result
def tgraphSpeedup(self, refDataset):
""" produces speed-up graphs"""
if(len(self._data)==0):
print "***** Dataset.tgraphNorm(): ERROR: no data available for", self._algo+'_'+self._impl
return TGraphErrors()
result = TGraph( len(self._data) )
#print "Speed-up: sizes: refDataset=%i self=%i" % (len(refDataset._data), len(self._data))
for i in range(len(self._data)):
speedup = refDataset._data[i]._y / self._data[i]._y
result.SetPoint( i, self._data[i]._x, speedup )
if speedup < self._minvalue : self._minvalue = speedup
if speedup > self._maxvalue : self._maxvalue = speedup
return result
def getEarliestVersion(self):
""" return earliest data point in a time series"""
return self._minkey
def getLatestVersion(self):
""" return latest data point in a time series"""
return self._maxkey
def getLowestValue(self):
""" return lowest data point in a time series"""
if self._maxvalue<=0 and self._minvalue>=99999:
print "ERROR in getLowestValue:",self._name,": make sure to call tgraphNorm() first!!!"
return self._minvalue
def getHighestValue(self):
""" return highest value in a time series"""
if self._maxvalue<=0 and self._minvalue>=99999:
print "ERROR in getHighestValue:",self._name,": make sure to call tgraphNorm() first!!!"
return self._maxvalue
def getLastValue(self):
return ( self.serie[self._maxkey] / self.serie[self._minkey] )
def getFirstValue(self):
return ( self.serie[self._minkey] / self.serie[self._minkey] )
### Testing
if __name__ == "__main__":
### these are the full series (without normalization)
algoNames = ["inside", "distToIn", "safetyToIn", "contains", "distToOut", "safetyToOut"]
implNames = ["root", "usolids", "unspec", "vector", "spec"]
### read in the data points from external file
datasets = {}
for algo in algoNames:
for impl in implNames:
datasets[ algo+impl ] = Dataset( algo, impl )
ok = True
try:
file = open("trap-perf-hist.dat", "rt")
except:
print "probs reading file? file=", filename
ok=False
pass
if ok:
try:
reader = csv.reader(file, delimiter=' ')
first=False
for row in reader :
if not first :
#print 'read: <%s>' % row
#print "row[0] = <%s>" % row[0]
#print "row[0][6:]=",row[0][6:]
#print "row[0][3:5]=",row[0][3:5]
#print "row[0][0:2]=",row[0][0:2]
vers = int(row[0])
algo = row[1]
impl = row[2]
perf = float(row[3])
error = float(row[4])
datasets[algo+impl].addPoint(vers,perf,error)
#print vers,algo,impl,perf,error
else :
#.. just skip reading header line in file
#print "len(row)=",len(row),":",row
if len(row)>0 and row[0]=='Data': first=False
pass
finally:
file.close()
print len(datasets),"datasets read."
# #.. show how to loop over data
# for key,val in datasets.iteritems():
# print val._algo, val._impl, len(val._data),val._minkey,val._maxkey
# for i in range(len(val._data)):
# print " ",val._data[i]._x,val._data[i]._y,val._data[i]._dy
#.. Make plots
from ROOT import TCanvas, gROOT, TGraphErrors, TPaveText, TLatex, TH1F, TProfile, TFile, gStyle
### Plot 1
c1 = TCanvas('c1','c1', 900, 600 )
c1.Divide(3,2)
### graphs contain the normalized datasets
graphs = []
for i in range(6): graphs.append([]) # 6 empty lists, one for each algo
#.. separate the data into the 6 different algorithms
for key,value in datasets.iteritems() :
#print "Dataset %s: %d elements between %s and %s" % (key, len(value._data), value._minkey, value._maxkey )
for ialgo in range(6):
ilen = len(algoNames[ialgo])
if key[0:ilen] == algoNames[ialgo]:
normGraph = value.tgraphNorm()
for impl in range(6):
icolor = impl+2
if icolor>=5: icolor+=1
if key[ilen:] == implNames[impl]:
normGraph.SetLineColor(icolor)
normGraph.SetMarkerStyle(20+icolor)
normGraph.SetMarkerColor(icolor)
break
#graphs[ialgo].append( normGraph )
graphs[ialgo].append( )
#.. now plot the 6 separate groups of datasets
for ialgo in range(len(graphs)):
ipad = ialgo + 1
ymin = 1.1
ymax = 0.9
for igraph in range(len(graphs[ialgo])):
tmp = graphs[ialgo][igraph]
npts = tmp.GetN()
x = tmp.GetX()
y = tmp.GetY()
for i in range(npts):
#print "point %i: (%f; %f)" % (i,x[i],y[i])
if y[i]<ymin: ymin = y[i]
if y[i]>ymax: ymax = y[i]
#print "ialgo=%i, igraph=%i, ymin=%f, ymax=%f" % (ialgo,igraph,ymin,ymax)
thePad = c1.cd(ipad)
thePad.SetGrid(1,1)
frame1 = thePad.DrawFrame(0, 0.95*ymin, 10, 1.05*ymax)
frame1.SetTitle(algoNames[ialgo])
frame1.GetXaxis().SetTitle("Version")
frame1.GetYaxis().SetTitle("Time [sec]")
for igra in range(len(graphs[ialgo])):
tmp = graphs[ialgo][igra]
#tmp.SetLineColor(ialgo+1)
if(tmp.GetN()>0): tmp.Draw("pl")
c1.Update()
c1.SaveAs("last.png")
|
