-
Notifications
You must be signed in to change notification settings - Fork 1
/
Filterbank.py
380 lines (327 loc) · 15.7 KB
/
Filterbank.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
import os,numpy,pylab,sys,cPickle,time
import ctypes as C
from Header import Header,MultiHeader
from SigPyProcUtils import File,Buffer,arrayToPointer
class Filterbank(Header):
"""Class to handle filterbank files in all their glory.
Parameters
----------
\tfilename -- name of filterbank file
Methods
-------
\tcollapse -- Collapse filterbank in frequency and/or time
\tdedisperse -- Simple dedispersion algorithm (fixed some sigproc bugs)
\tdownsample -- Downsample filterbank in frequency and/or time
"""
def __init__(self,filename):
"""Create Filterbank instance.
Args:
filename -- string containing name of file to process
"""
Header.__init__(self,filename)
self.f = File(filename,"r",nbits=self.info["nbits"])
infofile = "%s.info"%(self.basename)
if os.path.isfile(infofile):
try:
self.stats = cPickle.load(open(infofile,"r"))
except:
self.stats = None
else:
self.stats = None
if self.ctype == C.c_float:
self.lib = C.CDLL("libSigPyProc32.so")
else:
self.lib = C.CDLL("libSigPyProc8.so")
def getDMdelays(self,dm):
"""For a given dispersion measure return delay for each channel
"""
chanFreqs = (numpy.arange(self.info["nchans"])
*self.info['foff'])+self.info['fch1']
chanDelays = (dm * 4.148808e3 *
((chanFreqs**-2)-(self.info['fch1']**-2)
)/self.info["tsamp"]).round().astype("int32")
return chanDelays
def collapse(self,blocksize=512):
"""Collapse a filterbank in frequency.
Args:
blocksize -- number of samples to read in each gulp
Return: TimeSeries instance
"""
readBuffer = Buffer(blocksize*self.info["nchans"],self.ctype)
timBuffer = Buffer(self.info["nsamples"],C.c_float)
passPlan = ReadPlan(self,readBuffer)
passPlan.readOnce(blocksize)
for nsamps,ii in passPlan.makePass():
self.lib.getTim(readBuffer.Cbuffer,timBuffer.Cbuffer,
self.info["nchans"],nsamps,ii*blocksize)
return TimeSeries(self.info.copy(),timBuffer)
def bandpass(self,blocksize=512):
"""Collapse a filterbank in time.
Args:
blocksize -- number of samples to read in each gulp
Return: BandpassFromBuffer instance
"""
readBuffer = Buffer(blocksize*self.info["nchans"],self.ctype)
bpassBuffer = Buffer(self.info["nchans"],C.c_float)
passPlan = ReadPlan(self,readBuffer)
passPlan.readOnce(blocksize)
for nsamps,ii in passPlan.makePass():
self.lib.getBpass(readBuffer.Cbuffer,bpassBuffer.Cbuffer,
self.info["nchans"],blocksize)
return BandpassFromBuffer(self.info.copy(),bpassBuffer)
def dedisperse(self,dm,gulp=10000):
"""Dedisperse filterbank to timeseries.
Args:
dm -- Dispersion measure to dedisperse to
gulp -- size of block to read at a time, if chosen gulp
is less than maximum dispersion delay gulp is taken as 2 * max delay.
Returns: TimeSeries instance
"""
chanDelays = self.getDMdelays(dm)
delayPointer = arrayToPointer(chanDelays)
maxDelay = int(chanDelays.max())
gulp = max(2*maxDelay,gulp)
timLen = self.info["nsamples"]-maxDelay
timBuffer = Buffer(timLen,C.c_float)
readBuffer = Buffer(self.info["nchans"]*gulp,self.ctype)
passPlan = ReadPlan(self,readBuffer)
passPlan.readSkipBack(gulp,maxDelay)
for nsamps,ii in passPlan.makePass():
self.lib.dedisperse(readBuffer.Cbuffer,timBuffer.Cbuffer,delayPointer,
maxDelay, self.info["nchans"], nsamps, ii*(gulp-maxDelay))
timInfo = self.info.copy()
timInfo["nsamples"] = timLen
timInfo["refdm"] = dm
return TimeSeries(timInfo,timBuffer)
def downsample(self,tfactor=1,ffactor=1,filename=None):
"""Downsample filterbank in frequency and/or time.
Args:
tfactor -- Factor to downsample in time
ffactor -- Factor to downsample in frequecy (must be factor of nchans)
filename -- File to write downsampled data to.
Returns: Filterbank instance (from output file)
"""
if filename is None:
filename = "%s_f%d_t%d.fil"%(self.basename,ffactor,tfactor)
self.f.seek(self.hdrlen)
if not self.info["nchans"]%ffactor == 0:
raise ValueError,"Bad frequency factor given"
self.downsample_header(tfactor=tfactor,ffactor=ffactor)
outFile = self.prepOutfile(filename,
(("tsamp",self.info["tsamp"]*tfactor),
("nchans",self.info["nchans"]/ffactor),
("foff",self.info["foff"]*ffactor)))
readBuffer = Buffer(tfactor*self.info["nchans"],self.ctype)
writeBuffer = Buffer(self.info["nchans"]/ffactor,self.ctype)
tempBuffer = Buffer(self.info["nchans"]/ffactor,C.c_int)
self.lib.downsampleFil(self.f.f, outFile.f, readBuffer.Cbuffer,
writeBuffer.Cbuffer,tempBuffer.Cbuffer,
tfactor, ffactor, self.info["nchans"],
self.info["nsamples"])
return Filterbank(filename)
def fold(self,period,dm,nbins=50,nints=32,nbands=32,gulp=10000):
"""Fold filterbank phase bins, subintegrations and subbands.
Args:
period -- period in seconds to fold with
nbins -- number of phase bins in output
nints -- number of subintegrations in output
Returns: FoldedData instance
"""
nbands = min(nbands,self.info["nchans"])
chanDelays = self.getDMdelays(dm)
delayPointer = arrayToPointer(chanDelays)
maxDelay = int(chanDelays.max())
gulp = max(2*maxDelay,gulp)
foldBuffer = Buffer(nbins*nints*nbands,C.c_float)
countBuffer = Buffer(nbins*nints*nbands,C.c_int)
readBuffer = Buffer(self.info["nchans"]*gulp,self.ctype)
passPlan = ReadPlan(self,readBuffer)
passPlan.readSkipBack(gulp,maxDelay)
for nsamps,ii in passPlan.makePass():
self.lib.foldFil(readBuffer.Cbuffer, foldBuffer.Cbuffer, countBuffer.Cbuffer,
delayPointer, maxDelay, C.c_double(self.info["tsamp"]),
C.c_double(period), gulp, self.info["nchans"], nbins,
nints, nbands, ii*(gulp-maxDelay))
parentInfo = self.info.copy()
return FoldedData(parentInfo,foldBuffer,countBuffer,
period,dm,nbins,nints,nbands)
def dropBits32to8(self,gulp=1024,flag=0,clip=0):
if self.stats is None:
raise AttributeError,"First run getStatistics()"
readBuffer = Buffer(self.info["nchans"]*gulp,self.ctype)
writeBuffer = Buffer(self.info["nchans"]*gulp,C.c_ubyte)
passPlan = ReadPlan(self,readBuffer)
passPlan.readOnce(gulp)
if flag is not 0:
flagBuffer = Buffer(self.info["nchans"]*gulp,C.c_ubyte)
flagMax = flag*self.stats["sigma"].astype("float32")
flagMin = -flag*self.stats["sigma"].astype("float32")
if clip is not 0:
chanMax = numpy.array([min(a,b) for a,b in zip(clip*self.stats["sigma"],self.stats["maxima"])])
chanMin = numpy.array([max(a,b) for a,b in zip(-clip*self.stats["sigma"],self.stats["minima"])])
else:
chanMax = self.stats["maxima"]
chanMin = self.stats["minima"]
chanFactor = ((chanMax-chanMin)/255.)
chanPlus = (chanMin/chanFactor)
outFile = self.prepOutfile("%s_8bit.fil"%(self.basename),(("nbits",8)))
if flag is not 0:
changes = (("nbits",8),("source_name","%s_mask"%(self.info["source_name"].split()[0])))
flagFile = self.prepOutfile("%s_8bit_mask.fil"%(self.basename),changes)
for nsamps,ii in passPlan.makePass():
self.lib.to8bit(readBuffer.Cbuffer,writeBuffer.Cbuffer,
arrayToPointer(chanFactor),arrayToPointer(chanPlus),
nsamps,self.info["nchans"])
outFile.cwrite(writeBuffer)
if flag is not 0:
lib.flagBlock(readBuffer.Cbuffer,flagBuffer.Cbuffer,
arrayToPointer(flagMax),arrayToPointer(flagMin),
nsamps,self.info["nchans"])
flagFile.cwrite(flagBuffer)
if flag is not 0:
return(Filterbank(outFile.name),Filterbank(flagFile.name))
else:
return Filterbank(outFile.name)
def getStatistics(self,window=10001,gulp=30003):
if gulp < window: raise ValueError,"gulp must be > window"
readBuffer = Buffer(self.info["nchans"]*gulp,self.ctype)
writeBuffer = Buffer(self.info["nchans"]*gulp,self.ctype)
maximaBuffer = Buffer(self.info["nchans"],C.c_float)
minimaBuffer = Buffer(self.info["nchans"],C.c_float)
meansBuffer = Buffer(self.info["nchans"],C.c_float)
bpassBuffer = Buffer(self.info["nchans"],C.c_float)
stdevBuffer = Buffer(self.info["nchans"],C.c_float)
outFile = self.prepOutfile("%s_RM.fil"%(self.basename))
passPlan = ReadPlan(self,readBuffer)
passPlan.readSkipBack(gulp,window)
for nsamps,ii in passPlan.makePass():
self.lib.getStats(readBuffer.Cbuffer, meansBuffer.Cbuffer,
bpassBuffer.Cbuffer, stdevBuffer.Cbuffer,
writeBuffer.Cbuffer,maximaBuffer.Cbuffer,
minimaBuffer.Cbuffer,self.info["nchans"],nsamps,window,ii)
if ii == 0:
outFile.cwrite(writeBuffer)
elif ii == passPlan.nreads-1:
outFile.cwrite(writeBuffer, nunits=nsamps*self.info["nchans"],
offset=self.info["nchans"]*window)
else:
outFile.cwrite(writeBuffer,
nunits = (gulp-window)*self.info["nchans"],
offset = self.info["nchans"]*window)
stdevBuffer.Ndarray = numpy.sqrt(stdevBuffer.Ndarray/self.info["nsamples"])
infoFile = open("%s_RM.info"%(self.basename),"w+")
info = {"sigma":stdevBuffer.Ndarray,
"bandpass":bpassBuffer.Ndarray,
"maxima":maximaBuffer.Ndarray,
"minima":minimaBuffer.Ndarray}
cPickle.dump(info,infoFile)
infoFile.close()
return Filterbank(outFile.name)
def prepOutfile(self,filename,headerChanges=None):
outFile = File(filename,"w+")
if headerChanges is not None:
for key,val in headerChanges:
self.modify_header(key,val)
outFile.write(self.write_header())
self.reset_header()
return outFile
class ReadPlan:
def __init__(self,ObjInst,readBuffer,multi=False):
self.I = ObjInst
self.multi = multi
self.readBuffer = readBuffer
self.nsamps = self.I.info['nsamples']
self.nchans = self.I.info['nchans']
self.dtypesize = C.sizeof(self.I.ctype)
if self.multi:
self.nfiles = ObjInst.nfiles
self.fils = [inst for inst in self.I.filterbanks]
for inst in self.fils: inst.f.seek(inst.hdrlen)
else:
self.f = self.I.f
self.f.seek(self.I.hdrlen)
def readOnce(self,readsamps=512):
nreads = self.nsamps//readsamps
lastread = self.nsamps%readsamps
self.blocks = [ (ii,readsamps*self.nchans,0) for ii in range(nreads) ]
self.blocks.append( (nreads,lastread*self.nchans,0) )
self.nreads = nreads+1
self.printInfo()
def readSkipBack(self,readsamps=512,skipback=0):
skipback = abs(skipback)
if skipback >= readsamps:
raise ValueError,"readsamps must be > skipback value"
nreads = self.nsamps//(readsamps-skipback)
lastread = self.nsamps-(nreads*(readsamps-skipback))
if lastread<skipback:
lastread += skipback
nreads -= 1
self.blocks = [(ii,readsamps*self.nchans,-skipback*self.nchans) for ii in range(nreads)]
self.blocks.append((nreads,lastread*self.nchans,0))
self.nreads = nreads+1
self.printInfo()
def printInfo(self):
print
print "Filterbank reading plan:"
print "------------------------"
print "Number of reads: ",self.nreads
print "Nsamps of first read: ",self.blocks[0][1]/self.nchans
print "Nsamps of last read: ",self.blocks[-1][1]/self.nchans
print "Nsamps to skip back: ",-1*self.blocks[0][2]/self.nchans
print
def makePass(self):
start = time.time()
for ii,block,skip in self.blocks:
sys.stdout.write("Percentage complete: %d%%\r"%(100*ii/self.nreads))
sys.stdout.flush()
if self.multi:
for jj in range(self.nfiles):
self.fils[jj].f.cread(self.readBuffer,nunits=block,n=jj)
self.f.seek(skip*self.dtypesize,os.SEEK_CUR)
else:
self.f.cread(self.readBuffer,nunits=block)
self.f.seek(skip*self.dtypesize,os.SEEK_CUR)
yield block/self.nchans,ii
print "Execution time: %f seconds \n"%(time.time()-start)
class MultiFilterbank(MultiHeader):
def __init__(self,files):
MultiHeader.__init__(self,files)
self.filterbanks = [Filterbank(filename) for filename in files]
self.nfiles = len(files)
self.lags = numpy.zeros(self.nfiles)
def findLags(self):
zeroDMs = MultiTimeSeries()
for fil in self.filterbanks:
zeroDMs.append(fil.collapse(freq=False))
zeroDMs.findLags()
self.lags = numpy.array([zeroDMs.lags[fil.info["filename"]] for fil in self.filterbanks])
self.lags*=self.info["nchans"]
return zeroDMs,self.lags
def sumPols(self):
outFile = self.filterbanks[0].prepOutfile("SummedPols.fil")
readBuffers = Buffer(gulp*self.info["nchans"],self.ctype,dim=2)
writeBuffer = Buffer(gulp*self.info["nchans"],self.ctype)
passPlan = ReadPlan(self,readBuffers,multi=True)
passPlan.readOnce()
for nsamps,ii in passPlan.makePass():
lib.sumPols(readBuffers.Cbuffer,writeBuffer.Cbuffer,nsamps,self.info["nchans"])
outFile.cwrite(writeBuffer)
return Filterbank(outFile)
def genMBmask(self,gulp=512,threshold=4):
outFile = self.filterbanks[0].prepOutfile("MBmask.fil")
clags = Buffer(self.lags.shape[0],C.c_int)
clags.Ndarray = self.lags
for fil,offset in zip(self.filterbanks,self.lags):
fil.f.seek(fil.hdrlen+offset)
minlen = min([ fil.info["nsamples"]-offset for fil,offset in zip(self.filterbanks,self.lags)])
readBuffers = Buffer(gulp*self.info["nchans"],C.c_ubyte,dim=self.nfiles)
writeBuffer = Buffer(gulp*self.info["nchans"],C.c_ubyte)
passPlan = ReadPlan(self,readBuffers,multi=True)
passPlan.readOnce()
for nsamps,ii in passPlan.makePass():
self.lib.genMBmask(readBuffers.Cbuffer,writeBuffer.Cbuffer,
threshold,self.nfiles,nsamps,self.info["nchans"])
outFile.cwrite(writeBuffer,nsamps*self.info["nchans"])
from TimeSeries import TimeSeries,MultiTimeSeries
from FoldedData import FoldedData
from Bandpass import BandpassFromBuffer