| @@ -1,2158 +1,2599 | |||
|
|
1 | 1 | import numpy |
|
|
2 | 2 | import math |
|
|
3 | from scipy import optimize | |
|
|
4 | from scipy import interpolate | |
|
|
5 | from scipy import signal | |
|
|
6 | from scipy import stats | |
|
|
3 | from scipy import optimize, interpolate, signal, stats, ndimage | |
|
|
7 | 4 | import re |
|
|
8 | 5 | import datetime |
|
|
9 | 6 | import copy |
|
|
10 | 7 | import sys |
|
|
11 | 8 | import importlib |
|
|
12 | 9 | import itertools |
|
|
13 | 10 | |
|
|
14 | 11 | from jroproc_base import ProcessingUnit, Operation |
|
|
15 | from schainpy.model.data.jrodata import Parameters | |
|
|
12 | from schainpy.model.data.jrodata import Parameters, hildebrand_sekhon | |
|
|
16 | 13 | |
|
|
17 | 14 | |
|
|
18 | 15 | class ParametersProc(ProcessingUnit): |
|
|
19 | 16 | |
|
|
20 | 17 | nSeconds = None |
|
|
21 | 18 | |
|
|
22 | 19 | def __init__(self): |
|
|
23 | 20 | ProcessingUnit.__init__(self) |
|
|
24 | 21 | |
|
|
25 | 22 | # self.objectDict = {} |
|
|
26 | 23 | self.buffer = None |
|
|
27 | 24 | self.firstdatatime = None |
|
|
28 | 25 | self.profIndex = 0 |
|
|
29 | 26 | self.dataOut = Parameters() |
|
|
30 | 27 | |
|
|
31 | 28 | def __updateObjFromInput(self): |
|
|
32 | 29 | |
|
|
33 | 30 | self.dataOut.inputUnit = self.dataIn.type |
|
|
34 | 31 | |
|
|
35 | 32 | self.dataOut.timeZone = self.dataIn.timeZone |
|
|
36 | 33 | self.dataOut.dstFlag = self.dataIn.dstFlag |
|
|
37 | 34 | self.dataOut.errorCount = self.dataIn.errorCount |
|
|
38 | 35 | self.dataOut.useLocalTime = self.dataIn.useLocalTime |
|
|
39 | 36 | |
|
|
40 | 37 | self.dataOut.radarControllerHeaderObj = self.dataIn.radarControllerHeaderObj.copy() |
|
|
41 | 38 | self.dataOut.systemHeaderObj = self.dataIn.systemHeaderObj.copy() |
|
|
42 | 39 | self.dataOut.channelList = self.dataIn.channelList |
|
|
43 | 40 | self.dataOut.heightList = self.dataIn.heightList |
|
|
44 | 41 | self.dataOut.dtype = numpy.dtype([('real','<f4'),('imag','<f4')]) |
|
|
45 | 42 | # self.dataOut.nHeights = self.dataIn.nHeights |
|
|
46 | 43 | # self.dataOut.nChannels = self.dataIn.nChannels |
|
|
47 | 44 | self.dataOut.nBaud = self.dataIn.nBaud |
|
|
48 | 45 | self.dataOut.nCode = self.dataIn.nCode |
|
|
49 | 46 | self.dataOut.code = self.dataIn.code |
|
|
50 | 47 | # self.dataOut.nProfiles = self.dataOut.nFFTPoints |
|
|
51 | 48 | self.dataOut.flagDiscontinuousBlock = self.dataIn.flagDiscontinuousBlock |
|
|
52 | 49 | # self.dataOut.utctime = self.firstdatatime |
|
|
53 | 50 | self.dataOut.utctime = self.dataIn.utctime |
|
|
54 | 51 | self.dataOut.flagDecodeData = self.dataIn.flagDecodeData #asumo q la data esta decodificada |
|
|
55 | 52 | self.dataOut.flagDeflipData = self.dataIn.flagDeflipData #asumo q la data esta sin flip |
|
|
56 |
|
|
|
|
53 | self.dataOut.nCohInt = self.dataIn.nCohInt | |
|
|
57 | 54 | # self.dataOut.nIncohInt = 1 |
|
|
58 | 55 | self.dataOut.ippSeconds = self.dataIn.ippSeconds |
|
|
59 | 56 | # self.dataOut.windowOfFilter = self.dataIn.windowOfFilter |
|
|
60 | 57 | self.dataOut.timeInterval = self.dataIn.timeInterval |
|
|
61 | 58 | self.dataOut.heightList = self.dataIn.getHeiRange() |
|
|
62 | 59 | self.dataOut.frequency = self.dataIn.frequency |
|
|
60 | self.dataOut.noise = self.dataIn.noise | |
|
|
63 | 61 | |
|
|
64 | 62 | def run(self): |
|
|
65 | 63 | |
|
|
66 | 64 | #---------------------- Voltage Data --------------------------- |
|
|
67 | 65 | |
|
|
68 | 66 | if self.dataIn.type == "Voltage": |
|
|
69 | 67 | |
|
|
70 | 68 | self.__updateObjFromInput() |
|
|
71 | 69 | self.dataOut.data_pre = self.dataIn.data.copy() |
|
|
72 | 70 | self.dataOut.flagNoData = False |
|
|
73 | 71 | self.dataOut.utctimeInit = self.dataIn.utctime |
|
|
74 | 72 | self.dataOut.paramInterval = self.dataIn.nProfiles*self.dataIn.nCohInt*self.dataIn.ippSeconds |
|
|
75 | 73 | return |
|
|
76 | 74 | |
|
|
77 | 75 | #---------------------- Spectra Data --------------------------- |
|
|
78 | 76 | |
|
|
79 | 77 | if self.dataIn.type == "Spectra": |
|
|
80 | self.dataOut.data_pre = self.dataIn.data_spc.copy() | |
|
|
78 | ||
|
|
79 | self.dataOut.data_pre = (self.dataIn.data_spc,self.dataIn.data_cspc) | |
|
|
81 | 80 | self.dataOut.abscissaList = self.dataIn.getVelRange(1) |
|
|
82 | 81 | self.dataOut.noise = self.dataIn.getNoise() |
|
|
83 | 82 | self.dataOut.normFactor = self.dataIn.normFactor |
|
|
84 | 83 | self.dataOut.groupList = self.dataIn.pairsList |
|
|
85 | 84 | self.dataOut.flagNoData = False |
|
|
86 | 85 | |
|
|
87 | 86 | #---------------------- Correlation Data --------------------------- |
|
|
88 | 87 | |
|
|
89 | 88 | if self.dataIn.type == "Correlation": |
|
|
90 | lagRRange = self.dataIn.lagR | |
|
|
91 | indR = numpy.where(lagRRange == 0)[0][0] | |
|
|
92 | 89 | |
|
|
93 | self.dataOut.data_pre = self.dataIn.data_corr.copy()[:,:,indR,:] | |
|
|
94 |
self.dataOut. |
|
|
|
90 | if self.dataIn.data_ccf is not None: | |
|
|
91 | self.dataOut.data_pre = (self.dataIn.data_acf,self.dataIn.data_ccf) | |
|
|
92 | else: | |
|
|
93 | self.dataOut.data_pre = self.dataIn.data_acf.copy() | |
|
|
94 | ||
|
|
95 | self.dataOut.abscissaList = self.dataIn.lagRange | |
|
|
95 | 96 | self.dataOut.noise = self.dataIn.noise |
|
|
96 | 97 | self.dataOut.normFactor = self.dataIn.normFactor |
|
|
97 | 98 | self.dataOut.data_SNR = self.dataIn.SNR |
|
|
98 | 99 | self.dataOut.groupList = self.dataIn.pairsList |
|
|
99 | 100 | self.dataOut.flagNoData = False |
|
|
101 | self.dataOut.nAvg = self.dataIn.nAvg | |
|
|
100 | 102 | |
|
|
101 |
#---------------------- |
|
|
|
103 | #---------------------- Parameters Data --------------------------- | |
|
|
102 | 104 | |
|
|
103 | 105 | if self.dataIn.type == "Parameters": |
|
|
104 | 106 | self.dataOut.copy(self.dataIn) |
|
|
105 | 107 | self.dataOut.flagNoData = False |
|
|
106 | 108 | |
|
|
107 | 109 | return True |
|
|
108 | 110 | |
|
|
109 | 111 | self.__updateObjFromInput() |
|
|
110 | 112 | self.dataOut.utctimeInit = self.dataIn.utctime |
|
|
111 | 113 | self.dataOut.paramInterval = self.dataIn.timeInterval |
|
|
112 | 114 | |
|
|
113 | 115 | #------------------- Get Moments ---------------------------------- |
|
|
114 | 116 | def GetMoments(self, channelList = None): |
|
|
115 | 117 | ''' |
|
|
116 | 118 | Function GetMoments() |
|
|
117 | 119 | |
|
|
118 | 120 | Input: |
|
|
119 | 121 | channelList : simple channel list to select e.g. [2,3,7] |
|
|
120 | 122 | self.dataOut.data_pre |
|
|
121 | 123 | self.dataOut.abscissaList |
|
|
122 | 124 | self.dataOut.noise |
|
|
123 | 125 | |
|
|
124 | 126 | Affected: |
|
|
125 | 127 | self.dataOut.data_param |
|
|
126 | 128 | self.dataOut.data_SNR |
|
|
127 | 129 | |
|
|
128 | 130 | ''' |
|
|
131 | self.dataOut.data_pre = self.dataOut.data_pre[0] | |
|
|
129 | 132 | data = self.dataOut.data_pre |
|
|
130 | 133 | absc = self.dataOut.abscissaList[:-1] |
|
|
131 | 134 | noise = self.dataOut.noise |
|
|
132 | 135 | |
|
|
133 | 136 | data_param = numpy.zeros((data.shape[0], 4, data.shape[2])) |
|
|
134 | 137 | |
|
|
135 | 138 | if channelList== None: |
|
|
136 | 139 | channelList = self.dataIn.channelList |
|
|
137 | 140 | self.dataOut.channelList = channelList |
|
|
138 | 141 | |
|
|
139 | 142 | for ind in channelList: |
|
|
140 | 143 | data_param[ind,:,:] = self.__calculateMoments(data[ind,:,:], absc, noise[ind]) |
|
|
141 | 144 | |
|
|
142 | 145 | self.dataOut.data_param = data_param[:,1:,:] |
|
|
143 | 146 | self.dataOut.data_SNR = data_param[:,0] |
|
|
144 | 147 | return |
|
|
145 | 148 | |
|
|
146 | 149 | def __calculateMoments(self, oldspec, oldfreq, n0, nicoh = None, graph = None, smooth = None, type1 = None, fwindow = None, snrth = None, dc = None, aliasing = None, oldfd = None, wwauto = None): |
|
|
147 | 150 | |
|
|
148 | 151 | if (nicoh == None): nicoh = 1 |
|
|
149 | 152 | if (graph == None): graph = 0 |
|
|
150 | 153 | if (smooth == None): smooth = 0 |
|
|
151 | 154 | elif (self.smooth < 3): smooth = 0 |
|
|
152 | 155 | |
|
|
153 | 156 | if (type1 == None): type1 = 0 |
|
|
154 | 157 | if (fwindow == None): fwindow = numpy.zeros(oldfreq.size) + 1 |
|
|
155 | 158 | if (snrth == None): snrth = -3 |
|
|
156 | 159 | if (dc == None): dc = 0 |
|
|
157 | 160 | if (aliasing == None): aliasing = 0 |
|
|
158 | 161 | if (oldfd == None): oldfd = 0 |
|
|
159 | 162 | if (wwauto == None): wwauto = 0 |
|
|
160 | 163 | |
|
|
161 | 164 | if (n0 < 1.e-20): n0 = 1.e-20 |
|
|
162 | 165 | |
|
|
163 | 166 | freq = oldfreq |
|
|
164 | 167 | vec_power = numpy.zeros(oldspec.shape[1]) |
|
|
165 | 168 | vec_fd = numpy.zeros(oldspec.shape[1]) |
|
|
166 | 169 | vec_w = numpy.zeros(oldspec.shape[1]) |
|
|
167 | 170 | vec_snr = numpy.zeros(oldspec.shape[1]) |
|
|
168 | 171 | |
|
|
169 | 172 | for ind in range(oldspec.shape[1]): |
|
|
170 | 173 | |
|
|
171 | 174 | spec = oldspec[:,ind] |
|
|
172 | 175 | aux = spec*fwindow |
|
|
173 | 176 | max_spec = aux.max() |
|
|
174 | 177 | m = list(aux).index(max_spec) |
|
|
175 | 178 | |
|
|
176 | 179 | #Smooth |
|
|
177 | 180 | if (smooth == 0): spec2 = spec |
|
|
178 | 181 | else: spec2 = scipy.ndimage.filters.uniform_filter1d(spec,size=smooth) |
|
|
179 | 182 | |
|
|
180 | 183 | # Calculo de Momentos |
|
|
181 | 184 | bb = spec2[range(m,spec2.size)] |
|
|
182 | 185 | bb = (bb<n0).nonzero() |
|
|
183 | 186 | bb = bb[0] |
|
|
184 | 187 | |
|
|
185 | 188 | ss = spec2[range(0,m + 1)] |
|
|
186 | 189 | ss = (ss<n0).nonzero() |
|
|
187 | 190 | ss = ss[0] |
|
|
188 | 191 | |
|
|
189 | 192 | if (bb.size == 0): |
|
|
190 | 193 | bb0 = spec.size - 1 - m |
|
|
191 | 194 | else: |
|
|
192 | 195 | bb0 = bb[0] - 1 |
|
|
193 | 196 | if (bb0 < 0): |
|
|
194 | 197 | bb0 = 0 |
|
|
195 | 198 | |
|
|
196 | 199 | if (ss.size == 0): ss1 = 1 |
|
|
197 | 200 | else: ss1 = max(ss) + 1 |
|
|
198 | 201 | |
|
|
199 | 202 | if (ss1 > m): ss1 = m |
|
|
200 | 203 | |
|
|
201 | 204 | valid = numpy.asarray(range(int(m + bb0 - ss1 + 1))) + ss1 |
|
|
202 | 205 | power = ((spec2[valid] - n0)*fwindow[valid]).sum() |
|
|
203 | 206 | fd = ((spec2[valid]- n0)*freq[valid]*fwindow[valid]).sum()/power |
|
|
204 | 207 | w = math.sqrt(((spec2[valid] - n0)*fwindow[valid]*(freq[valid]- fd)**2).sum()/power) |
|
|
205 | 208 | snr = (spec2.mean()-n0)/n0 |
|
|
206 | 209 | |
|
|
207 | 210 | if (snr < 1.e-20) : |
|
|
208 | 211 | snr = 1.e-20 |
|
|
209 | 212 | |
|
|
210 | 213 | vec_power[ind] = power |
|
|
211 | 214 | vec_fd[ind] = fd |
|
|
212 | 215 | vec_w[ind] = w |
|
|
213 | 216 | vec_snr[ind] = snr |
|
|
214 | 217 | |
|
|
215 | 218 | moments = numpy.vstack((vec_snr, vec_power, vec_fd, vec_w)) |
|
|
216 | 219 | return moments |
|
|
217 | 220 | |
|
|
218 | 221 | #------------------ Get SA Parameters -------------------------- |
|
|
219 | 222 | |
|
|
220 | 223 | def GetSAParameters(self): |
|
|
221 | 224 | pairslist = self.dataOut.groupList |
|
|
222 | 225 | num_pairs = len(pairslist) |
|
|
223 | 226 | |
|
|
224 | 227 | vel = self.dataOut.abscissaList |
|
|
225 | 228 | spectra = self.dataOut.data_pre |
|
|
226 | 229 | cspectra = self.dataIn.data_cspc |
|
|
227 | 230 | delta_v = vel[1] - vel[0] |
|
|
228 | 231 | |
|
|
229 | 232 | #Calculating the power spectrum |
|
|
230 | 233 | spc_pow = numpy.sum(spectra, 3)*delta_v |
|
|
231 | 234 | #Normalizing Spectra |
|
|
232 | 235 | norm_spectra = spectra/spc_pow |
|
|
233 | 236 | #Calculating the norm_spectra at peak |
|
|
234 | 237 | max_spectra = numpy.max(norm_spectra, 3) |
|
|
235 | 238 | |
|
|
236 | 239 | #Normalizing Cross Spectra |
|
|
237 | 240 | norm_cspectra = numpy.zeros(cspectra.shape) |
|
|
238 | 241 | |
|
|
239 | 242 | for i in range(num_chan): |
|
|
240 | 243 | norm_cspectra[i,:,:] = cspectra[i,:,:]/numpy.sqrt(spc_pow[pairslist[i][0],:]*spc_pow[pairslist[i][1],:]) |
|
|
241 | 244 | |
|
|
242 | 245 | max_cspectra = numpy.max(norm_cspectra,2) |
|
|
243 | 246 | max_cspectra_index = numpy.argmax(norm_cspectra, 2) |
|
|
244 | 247 | |
|
|
245 | 248 | for i in range(num_pairs): |
|
|
246 | 249 | cspc_par[i,:,:] = __calculateMoments(norm_cspectra) |
|
|
247 | 250 | #------------------- Get Lags ---------------------------------- |
|
|
248 | 251 | |
|
|
249 | 252 | def GetLags(self): |
|
|
250 | 253 | ''' |
|
|
251 | 254 | Function GetMoments() |
|
|
252 | 255 | |
|
|
253 | 256 | Input: |
|
|
254 | 257 | self.dataOut.data_pre |
|
|
255 | 258 | self.dataOut.abscissaList |
|
|
256 | 259 | self.dataOut.noise |
|
|
257 | 260 | self.dataOut.normFactor |
|
|
258 | 261 | self.dataOut.data_SNR |
|
|
259 | 262 | self.dataOut.groupList |
|
|
260 | 263 | self.dataOut.nChannels |
|
|
261 | 264 | |
|
|
262 | 265 | Affected: |
|
|
263 | 266 | self.dataOut.data_param |
|
|
264 | 267 | |
|
|
265 | 268 | ''' |
|
|
266 | 269 | |
|
|
267 | 270 | data = self.dataOut.data_pre |
|
|
268 | 271 | normFactor = self.dataOut.normFactor |
|
|
269 | 272 | nHeights = self.dataOut.nHeights |
|
|
270 | 273 | absc = self.dataOut.abscissaList[:-1] |
|
|
271 | 274 | noise = self.dataOut.noise |
|
|
272 | 275 | SNR = self.dataOut.data_SNR |
|
|
273 | 276 | pairsList = self.dataOut.groupList |
|
|
274 | 277 | nChannels = self.dataOut.nChannels |
|
|
275 | 278 | pairsAutoCorr, pairsCrossCorr = self.__getPairsAutoCorr(pairsList, nChannels) |
|
|
276 | 279 | self.dataOut.data_param = numpy.zeros((len(pairsCrossCorr)*2 + 1, nHeights)) |
|
|
277 | 280 | |
|
|
278 | 281 | dataNorm = numpy.abs(data) |
|
|
279 | 282 | for l in range(len(pairsList)): |
|
|
280 | 283 | dataNorm[l,:,:] = dataNorm[l,:,:]/normFactor[l,:] |
|
|
281 | 284 | |
|
|
282 | 285 | self.dataOut.data_param[:-1,:] = self.__calculateTaus(dataNorm, pairsCrossCorr, pairsAutoCorr, absc) |
|
|
283 | 286 | self.dataOut.data_param[-1,:] = self.__calculateLag1Phase(data, pairsAutoCorr, absc) |
|
|
284 | 287 | return |
|
|
285 | 288 | |
|
|
286 | 289 | def __getPairsAutoCorr(self, pairsList, nChannels): |
|
|
287 | 290 | |
|
|
288 | 291 | pairsAutoCorr = numpy.zeros(nChannels, dtype = 'int')*numpy.nan |
|
|
289 | 292 | |
|
|
290 | 293 | for l in range(len(pairsList)): |
|
|
291 | 294 | firstChannel = pairsList[l][0] |
|
|
292 | 295 | secondChannel = pairsList[l][1] |
|
|
293 | 296 | |
|
|
294 | 297 | #Obteniendo pares de Autocorrelacion |
|
|
295 | 298 | if firstChannel == secondChannel: |
|
|
296 | 299 | pairsAutoCorr[firstChannel] = int(l) |
|
|
297 | 300 | |
|
|
298 | 301 | pairsAutoCorr = pairsAutoCorr.astype(int) |
|
|
299 | 302 | |
|
|
300 | 303 | pairsCrossCorr = range(len(pairsList)) |
|
|
301 | 304 | pairsCrossCorr = numpy.delete(pairsCrossCorr,pairsAutoCorr) |
|
|
302 | 305 | |
|
|
303 | 306 | return pairsAutoCorr, pairsCrossCorr |
|
|
304 | 307 | |
|
|
305 | 308 | def __calculateTaus(self, data, pairsCrossCorr, pairsAutoCorr, lagTRange): |
|
|
306 | 309 | |
|
|
307 | 310 | Pt0 = data.shape[1]/2 |
|
|
308 | 311 | #Funcion de Autocorrelacion |
|
|
309 | 312 | dataAutoCorr = stats.nanmean(data[pairsAutoCorr,:,:], axis = 0) |
|
|
310 | 313 | |
|
|
311 | 314 | #Obtencion Indice de TauCross |
|
|
312 | 315 | indCross = data[pairsCrossCorr,:,:].argmax(axis = 1) |
|
|
313 | 316 | #Obtencion Indice de TauAuto |
|
|
314 | 317 | indAuto = numpy.zeros(indCross.shape,dtype = 'int') |
|
|
315 | 318 | CCValue = data[pairsCrossCorr,Pt0,:] |
|
|
316 | 319 | for i in range(pairsCrossCorr.size): |
|
|
317 | 320 | indAuto[i,:] = numpy.abs(dataAutoCorr - CCValue[i,:]).argmin(axis = 0) |
|
|
318 | 321 | |
|
|
319 | 322 | #Obtencion de TauCross y TauAuto |
|
|
320 | 323 | tauCross = lagTRange[indCross] |
|
|
321 | 324 | tauAuto = lagTRange[indAuto] |
|
|
322 | 325 | |
|
|
323 | 326 | Nan1, Nan2 = numpy.where(tauCross == lagTRange[0]) |
|
|
324 | 327 | |
|
|
325 | 328 | tauCross[Nan1,Nan2] = numpy.nan |
|
|
326 | 329 | tauAuto[Nan1,Nan2] = numpy.nan |
|
|
327 | 330 | tau = numpy.vstack((tauCross,tauAuto)) |
|
|
328 | 331 | |
|
|
329 | 332 | return tau |
|
|
330 | 333 | |
|
|
331 | 334 | def __calculateLag1Phase(self, data, pairs, lagTRange): |
|
|
332 | 335 | data1 = stats.nanmean(data[pairs,:,:], axis = 0) |
|
|
333 | 336 | lag1 = numpy.where(lagTRange == 0)[0][0] + 1 |
|
|
334 | 337 | |
|
|
335 | 338 | phase = numpy.angle(data1[lag1,:]) |
|
|
336 | 339 | |
|
|
337 | 340 | return phase |
|
|
338 | 341 | #------------------- Detect Meteors ------------------------------ |
|
|
339 | 342 | |
|
|
340 | 343 | def MeteorDetection(self, hei_ref = None, tauindex = 0, |
|
|
341 | 344 | phaseOffsets = None, |
|
|
342 | 345 | cohDetection = False, cohDet_timeStep = 1, cohDet_thresh = 25, |
|
|
343 | 346 | noise_timeStep = 4, noise_multiple = 4, |
|
|
344 | 347 | multDet_timeLimit = 1, multDet_rangeLimit = 3, |
|
|
345 | 348 | phaseThresh = 20, SNRThresh = 8, |
|
|
346 | 349 | hmin = 50, hmax=150, azimuth = 0, |
|
|
347 | 350 | channel25X = 0, channel20X = 4, channelCentX = 2, |
|
|
348 | 351 | channel25Y = 1, channel20Y = 3, channelCentY = 2) : |
|
|
349 | 352 | |
|
|
350 | 353 | ''' |
|
|
351 | 354 | Function DetectMeteors() |
|
|
352 | 355 | Project developed with paper: |
|
|
353 | 356 | HOLDSWORTH ET AL. 2004 |
|
|
354 | 357 | |
|
|
355 | 358 | Input: |
|
|
356 | 359 | self.dataOut.data_pre |
|
|
357 | 360 | |
|
|
358 | 361 | centerReceiverIndex: From the channels, which is the center receiver |
|
|
359 | 362 | |
|
|
360 | 363 | hei_ref: Height reference for the Beacon signal extraction |
|
|
361 | 364 | tauindex: |
|
|
362 | 365 | predefinedPhaseShifts: Predefined phase offset for the voltge signals |
|
|
363 | 366 | |
|
|
364 | 367 | cohDetection: Whether to user Coherent detection or not |
|
|
365 | 368 | cohDet_timeStep: Coherent Detection calculation time step |
|
|
366 | 369 | cohDet_thresh: Coherent Detection phase threshold to correct phases |
|
|
367 | 370 | |
|
|
368 | 371 | noise_timeStep: Noise calculation time step |
|
|
369 | 372 | noise_multiple: Noise multiple to define signal threshold |
|
|
370 | 373 | |
|
|
371 | 374 | multDet_timeLimit: Multiple Detection Removal time limit in seconds |
|
|
372 | 375 | multDet_rangeLimit: Multiple Detection Removal range limit in km |
|
|
373 | 376 | |
|
|
374 | 377 | phaseThresh: Maximum phase difference between receiver to be consider a meteor |
|
|
375 | 378 | SNRThresh: Minimum SNR threshold of the meteor signal to be consider a meteor |
|
|
376 | 379 | |
|
|
377 | 380 | hmin: Minimum Height of the meteor to use it in the further wind estimations |
|
|
378 | 381 | hmax: Maximum Height of the meteor to use it in the further wind estimations |
|
|
379 | 382 | azimuth: Azimuth angle correction |
|
|
380 | 383 | |
|
|
381 | 384 | Affected: |
|
|
382 | 385 | self.dataOut.data_param |
|
|
383 | 386 | |
|
|
384 | 387 | Rejection Criteria (Errors): |
|
|
385 | 388 | 0: No error; analysis OK |
|
|
386 | 389 | 1: SNR < SNR threshold |
|
|
387 | 390 | 2: angle of arrival (AOA) ambiguously determined |
|
|
388 | 391 | 3: AOA estimate not feasible |
|
|
389 | 392 | 4: Large difference in AOAs obtained from different antenna baselines |
|
|
390 | 393 | 5: echo at start or end of time series |
|
|
391 | 394 | 6: echo less than 5 examples long; too short for analysis |
|
|
392 | 395 | 7: echo rise exceeds 0.3s |
|
|
393 | 396 | 8: echo decay time less than twice rise time |
|
|
394 | 397 | 9: large power level before echo |
|
|
395 | 398 | 10: large power level after echo |
|
|
396 | 399 | 11: poor fit to amplitude for estimation of decay time |
|
|
397 | 400 | 12: poor fit to CCF phase variation for estimation of radial drift velocity |
|
|
398 | 401 | 13: height unresolvable echo: not valid height within 70 to 110 km |
|
|
399 | 402 | 14: height ambiguous echo: more then one possible height within 70 to 110 km |
|
|
400 | 403 | 15: radial drift velocity or projected horizontal velocity exceeds 200 m/s |
|
|
401 | 404 | 16: oscilatory echo, indicating event most likely not an underdense echo |
|
|
402 | 405 | |
|
|
403 | 406 | 17: phase difference in meteor Reestimation |
|
|
404 | 407 | |
|
|
405 | 408 | Data Storage: |
|
|
406 | 409 | Meteors for Wind Estimation (8): |
|
|
407 | 410 | Utc Time | Range Height |
|
|
408 | 411 | Azimuth Zenith errorCosDir |
|
|
409 | 412 | VelRad errorVelRad |
|
|
410 | 413 | Phase0 Phase1 Phase2 Phase3 |
|
|
411 | 414 | TypeError |
|
|
412 | 415 | |
|
|
413 | 416 | ''' |
|
|
414 | 417 | #Get Beacon signal |
|
|
415 | 418 | newheis = numpy.where(self.dataOut.heightList>self.dataOut.radarControllerHeaderObj.Taus[tauindex]) |
|
|
416 | 419 | |
|
|
417 | 420 | if hei_ref != None: |
|
|
418 | 421 | newheis = numpy.where(self.dataOut.heightList>hei_ref) |
|
|
419 | 422 | |
|
|
420 | 423 | heiRang = self.dataOut.getHeiRange() |
|
|
421 | 424 | #Pairs List |
|
|
422 | 425 | ''' |
|
|
423 | 426 | Cambiar este pairslist |
|
|
424 | 427 | ''' |
|
|
425 | 428 | pairslist = [(channelCentX, channel25X),(channelCentX, channel20X),(channelCentY,channel25Y),(channelCentY, channel20Y)] |
|
|
426 | 429 | |
|
|
427 | 430 | # nChannel = self.dataOut.nChannels |
|
|
428 | 431 | # for i in range(nChannel): |
|
|
429 | 432 | # if i != centerReceiverIndex: |
|
|
430 | 433 | # pairslist.append((centerReceiverIndex,i)) |
|
|
431 | 434 | |
|
|
432 | 435 | #****************REMOVING HARDWARE PHASE DIFFERENCES*************** |
|
|
433 | 436 | # see if the user put in pre defined phase shifts |
|
|
434 | 437 | voltsPShift = self.dataOut.data_pre.copy() |
|
|
435 | 438 | |
|
|
436 | 439 | # if predefinedPhaseShifts != None: |
|
|
437 | 440 | # hardwarePhaseShifts = numpy.array(predefinedPhaseShifts)*numpy.pi/180 |
|
|
438 | 441 | # |
|
|
439 | 442 | # # elif beaconPhaseShifts: |
|
|
440 | 443 | # # #get hardware phase shifts using beacon signal |
|
|
441 | 444 | # # hardwarePhaseShifts = self.__getHardwarePhaseDiff(self.dataOut.data_pre, pairslist, newheis, 10) |
|
|
442 | 445 | # # hardwarePhaseShifts = numpy.insert(hardwarePhaseShifts,centerReceiverIndex,0) |
|
|
443 | 446 | # |
|
|
444 | 447 | # else: |
|
|
445 | 448 | # hardwarePhaseShifts = numpy.zeros(5) |
|
|
446 | 449 | # |
|
|
447 | 450 | # voltsPShift = numpy.zeros((self.dataOut.data_pre.shape[0],self.dataOut.data_pre.shape[1],self.dataOut.data_pre.shape[2]), dtype = 'complex') |
|
|
448 | 451 | # for i in range(self.dataOut.data_pre.shape[0]): |
|
|
449 | 452 | # voltsPShift[i,:,:] = self.__shiftPhase(self.dataOut.data_pre[i,:,:], hardwarePhaseShifts[i]) |
|
|
450 | 453 | |
|
|
451 | 454 | #******************END OF REMOVING HARDWARE PHASE DIFFERENCES********* |
|
|
452 | 455 | |
|
|
453 | 456 | #Remove DC |
|
|
454 | 457 | voltsDC = numpy.mean(voltsPShift,1) |
|
|
455 | 458 | voltsDC = numpy.mean(voltsDC,1) |
|
|
456 | 459 | for i in range(voltsDC.shape[0]): |
|
|
457 | 460 | voltsPShift[i] = voltsPShift[i] - voltsDC[i] |
|
|
458 | 461 | |
|
|
459 | 462 | #Don't considerate last heights, theyre used to calculate Hardware Phase Shift |
|
|
460 | 463 | voltsPShift = voltsPShift[:,:,:newheis[0][0]] |
|
|
461 | 464 | |
|
|
462 | 465 | #************ FIND POWER OF DATA W/COH OR NON COH DETECTION (3.4) ********** |
|
|
463 | 466 | #Coherent Detection |
|
|
464 | 467 | if cohDetection: |
|
|
465 | 468 | #use coherent detection to get the net power |
|
|
466 | 469 | cohDet_thresh = cohDet_thresh*numpy.pi/180 |
|
|
467 | 470 | voltsPShift = self.__coherentDetection(voltsPShift, cohDet_timeStep, self.dataOut.timeInterval, pairslist, cohDet_thresh) |
|
|
468 | 471 | |
|
|
469 | 472 | #Non-coherent detection! |
|
|
470 | 473 | powerNet = numpy.nansum(numpy.abs(voltsPShift[:,:,:])**2,0) |
|
|
471 | 474 | #********** END OF COH/NON-COH POWER CALCULATION********************** |
|
|
472 | 475 | |
|
|
473 | 476 | #********** FIND THE NOISE LEVEL AND POSSIBLE METEORS **************** |
|
|
474 | 477 | #Get noise |
|
|
475 | 478 | noise, noise1 = self.__getNoise(powerNet, noise_timeStep, self.dataOut.timeInterval) |
|
|
476 | 479 | # noise = self.getNoise1(powerNet, noise_timeStep, self.dataOut.timeInterval) |
|
|
477 | 480 | #Get signal threshold |
|
|
478 | 481 | signalThresh = noise_multiple*noise |
|
|
479 | 482 | #Meteor echoes detection |
|
|
480 | 483 | listMeteors = self.__findMeteors(powerNet, signalThresh) |
|
|
481 | 484 | #******* END OF NOISE LEVEL AND POSSIBLE METEORS CACULATION ********** |
|
|
482 | 485 | |
|
|
483 | 486 | #************** REMOVE MULTIPLE DETECTIONS (3.5) *************************** |
|
|
484 | 487 | #Parameters |
|
|
485 | 488 | heiRange = self.dataOut.getHeiRange() |
|
|
486 | 489 | rangeInterval = heiRange[1] - heiRange[0] |
|
|
487 | 490 | rangeLimit = multDet_rangeLimit/rangeInterval |
|
|
488 | 491 | timeLimit = multDet_timeLimit/self.dataOut.timeInterval |
|
|
489 | 492 | #Multiple detection removals |
|
|
490 | 493 | listMeteors1 = self.__removeMultipleDetections(listMeteors, rangeLimit, timeLimit) |
|
|
491 | 494 | #************ END OF REMOVE MULTIPLE DETECTIONS ********************** |
|
|
492 | 495 | |
|
|
493 | 496 | #********************* METEOR REESTIMATION (3.7, 3.8, 3.9, 3.10) ******************** |
|
|
494 | 497 | #Parameters |
|
|
495 | 498 | phaseThresh = phaseThresh*numpy.pi/180 |
|
|
496 | 499 | thresh = [phaseThresh, noise_multiple, SNRThresh] |
|
|
497 | 500 | #Meteor reestimation (Errors N 1, 6, 12, 17) |
|
|
498 | 501 | listMeteors2, listMeteorsPower, listMeteorsVolts = self.__meteorReestimation(listMeteors1, voltsPShift, pairslist, thresh, noise, self.dataOut.timeInterval, self.dataOut.frequency) |
|
|
499 | 502 | # listMeteors2, listMeteorsPower, listMeteorsVolts = self.meteorReestimation3(listMeteors2, listMeteorsPower, listMeteorsVolts, voltsPShift, pairslist, thresh, noise) |
|
|
500 | 503 | #Estimation of decay times (Errors N 7, 8, 11) |
|
|
501 | 504 | listMeteors3 = self.__estimateDecayTime(listMeteors2, listMeteorsPower, self.dataOut.timeInterval, self.dataOut.frequency) |
|
|
502 | 505 | #******************* END OF METEOR REESTIMATION ******************* |
|
|
503 | 506 | |
|
|
504 | 507 | #********************* METEOR PARAMETERS CALCULATION (3.11, 3.12, 3.13) ************************** |
|
|
505 | 508 | #Calculating Radial Velocity (Error N 15) |
|
|
506 | 509 | radialStdThresh = 10 |
|
|
507 | 510 | listMeteors4 = self.__getRadialVelocity(listMeteors3, listMeteorsVolts, radialStdThresh, pairslist, self.dataOut.timeInterval) |
|
|
508 | 511 | |
|
|
509 | 512 | if len(listMeteors4) > 0: |
|
|
510 | 513 | #Setting New Array |
|
|
511 | 514 | date = self.dataOut.utctime |
|
|
512 | 515 | arrayParameters = self.__setNewArrays(listMeteors4, date, heiRang) |
|
|
513 | 516 | |
|
|
514 | 517 | #Correcting phase offset |
|
|
515 | 518 | if phaseOffsets != None: |
|
|
516 | 519 | phaseOffsets = numpy.array(phaseOffsets)*numpy.pi/180 |
|
|
517 | 520 | arrayParameters[:,8:12] = numpy.unwrap(arrayParameters[:,8:12] + phaseOffsets) |
|
|
518 | 521 | |
|
|
519 | 522 | #Second Pairslist |
|
|
520 | 523 | pairsList = [] |
|
|
521 | 524 | pairx = (0,1) |
|
|
522 | 525 | pairy = (2,3) |
|
|
523 | 526 | pairsList.append(pairx) |
|
|
524 | 527 | pairsList.append(pairy) |
|
|
525 | 528 | |
|
|
526 | 529 | meteorOps = MeteorOperations() |
|
|
527 | 530 | jph = numpy.array([0,0,0,0]) |
|
|
528 | 531 | h = (hmin,hmax) |
|
|
529 | 532 | arrayParameters = meteorOps.getMeteorParams(arrayParameters, azimuth, h, pairsList, jph) |
|
|
530 | 533 | |
|
|
531 | 534 | # #Calculate AOA (Error N 3, 4) |
|
|
532 | 535 | # #JONES ET AL. 1998 |
|
|
533 | 536 | # error = arrayParameters[:,-1] |
|
|
534 | 537 | # AOAthresh = numpy.pi/8 |
|
|
535 | 538 | # phases = -arrayParameters[:,9:13] |
|
|
536 | 539 | # arrayParameters[:,4:7], arrayParameters[:,-1] = meteorOps.getAOA(phases, pairsList, error, AOAthresh, azimuth) |
|
|
537 | 540 | # |
|
|
538 | 541 | # #Calculate Heights (Error N 13 and 14) |
|
|
539 | 542 | # error = arrayParameters[:,-1] |
|
|
540 | 543 | # Ranges = arrayParameters[:,2] |
|
|
541 | 544 | # zenith = arrayParameters[:,5] |
|
|
542 | 545 | # arrayParameters[:,3], arrayParameters[:,-1] = meteorOps.getHeights(Ranges, zenith, error, hmin, hmax) |
|
|
543 | 546 | # error = arrayParameters[:,-1] |
|
|
544 | 547 | #********************* END OF PARAMETERS CALCULATION ************************** |
|
|
545 | 548 | |
|
|
546 | 549 | #***************************+ PASS DATA TO NEXT STEP ********************** |
|
|
547 | 550 | # arrayFinal = arrayParameters.reshape((1,arrayParameters.shape[0],arrayParameters.shape[1])) |
|
|
548 | 551 | self.dataOut.data_param = arrayParameters |
|
|
549 | 552 | |
|
|
550 | 553 | if arrayParameters == None: |
|
|
551 | 554 | self.dataOut.flagNoData = True |
|
|
552 | 555 | |
|
|
553 | 556 | return |
|
|
554 | 557 | |
|
|
555 | 558 | def CorrectMeteorPhases(self, phaseOffsets, hmin = 50, hmax = 150, azimuth = 45): |
|
|
556 | 559 | |
|
|
557 | 560 | arrayParameters = self.dataOut.data_param |
|
|
558 | 561 | pairsList = [] |
|
|
559 | 562 | pairx = (0,1) |
|
|
560 | 563 | pairy = (2,3) |
|
|
561 | 564 | pairsList.append(pairx) |
|
|
562 | 565 | pairsList.append(pairy) |
|
|
563 | 566 | jph = numpy.zeros(4) |
|
|
564 | 567 | |
|
|
565 | 568 | phaseOffsets = numpy.array(phaseOffsets)*numpy.pi/180 |
|
|
566 | 569 | arrayParameters[:,8:12] = numpy.unwrap(arrayParameters[:,8:12] + phaseOffsets) |
|
|
567 | 570 | |
|
|
568 | 571 | meteorOps = MeteorOperations() |
|
|
569 | 572 | h = (hmin,hmax) |
|
|
570 | 573 | |
|
|
571 | 574 | arrayParameters = meteorOps.getMeteorParams(arrayParameters, azimuth, h, pairsList, jph) |
|
|
572 | 575 | self.dataOut.data_param = arrayParameters |
|
|
573 | 576 | return |
|
|
574 | 577 | |
|
|
575 | 578 | def __getHardwarePhaseDiff(self, voltage0, pairslist, newheis, n): |
|
|
576 | 579 | |
|
|
577 | 580 | minIndex = min(newheis[0]) |
|
|
578 | 581 | maxIndex = max(newheis[0]) |
|
|
579 | 582 | |
|
|
580 | 583 | voltage = voltage0[:,:,minIndex:maxIndex+1] |
|
|
581 | 584 | nLength = voltage.shape[1]/n |
|
|
582 | 585 | nMin = 0 |
|
|
583 | 586 | nMax = 0 |
|
|
584 | 587 | phaseOffset = numpy.zeros((len(pairslist),n)) |
|
|
585 | 588 | |
|
|
586 | 589 | for i in range(n): |
|
|
587 | 590 | nMax += nLength |
|
|
588 | 591 | phaseCCF = -numpy.angle(self.__calculateCCF(voltage[:,nMin:nMax,:], pairslist, [0])) |
|
|
589 | 592 | phaseCCF = numpy.mean(phaseCCF, axis = 2) |
|
|
590 | 593 | phaseOffset[:,i] = phaseCCF.transpose() |
|
|
591 | 594 | nMin = nMax |
|
|
592 | 595 | # phaseDiff, phaseArrival = self.estimatePhaseDifference(voltage, pairslist) |
|
|
593 | 596 | |
|
|
594 | 597 | #Remove Outliers |
|
|
595 | 598 | factor = 2 |
|
|
596 | 599 | wt = phaseOffset - signal.medfilt(phaseOffset,(1,5)) |
|
|
597 | 600 | dw = numpy.std(wt,axis = 1) |
|
|
598 | 601 | dw = dw.reshape((dw.size,1)) |
|
|
599 | 602 | ind = numpy.where(numpy.logical_or(wt>dw*factor,wt<-dw*factor)) |
|
|
600 | 603 | phaseOffset[ind] = numpy.nan |
|
|
601 | 604 | phaseOffset = stats.nanmean(phaseOffset, axis=1) |
|
|
602 | 605 | |
|
|
603 | 606 | return phaseOffset |
|
|
604 | 607 | |
|
|
605 | 608 | def __shiftPhase(self, data, phaseShift): |
|
|
606 | 609 | #this will shift the phase of a complex number |
|
|
607 | 610 | dataShifted = numpy.abs(data) * numpy.exp((numpy.angle(data)+phaseShift)*1j) |
|
|
608 | 611 | return dataShifted |
|
|
609 | 612 | |
|
|
610 | 613 | def __estimatePhaseDifference(self, array, pairslist): |
|
|
611 | 614 | nChannel = array.shape[0] |
|
|
612 | 615 | nHeights = array.shape[2] |
|
|
613 | 616 | numPairs = len(pairslist) |
|
|
614 | 617 | # phaseCCF = numpy.zeros((nChannel, 5, nHeights)) |
|
|
615 | 618 | phaseCCF = numpy.angle(self.__calculateCCF(array, pairslist, [-2,-1,0,1,2])) |
|
|
616 | 619 | |
|
|
617 | 620 | #Correct phases |
|
|
618 | 621 | derPhaseCCF = phaseCCF[:,1:,:] - phaseCCF[:,0:-1,:] |
|
|
619 | 622 | indDer = numpy.where(numpy.abs(derPhaseCCF) > numpy.pi) |
|
|
620 | 623 | |
|
|
621 | 624 | if indDer[0].shape[0] > 0: |
|
|
622 | 625 | for i in range(indDer[0].shape[0]): |
|
|
623 | 626 | signo = -numpy.sign(derPhaseCCF[indDer[0][i],indDer[1][i],indDer[2][i]]) |
|
|
624 | 627 | phaseCCF[indDer[0][i],indDer[1][i]+1:,:] += signo*2*numpy.pi |
|
|
625 | 628 | |
|
|
626 | 629 | # for j in range(numSides): |
|
|
627 | 630 | # phaseCCFAux = self.calculateCCF(arrayCenter, arraySides[j,:,:], [-2,1,0,1,2]) |
|
|
628 | 631 | # phaseCCF[j,:,:] = numpy.angle(phaseCCFAux) |
|
|
629 | 632 | # |
|
|
630 | 633 | #Linear |
|
|
631 | 634 | phaseInt = numpy.zeros((numPairs,1)) |
|
|
632 | 635 | angAllCCF = phaseCCF[:,[0,1,3,4],0] |
|
|
633 | 636 | for j in range(numPairs): |
|
|
634 | 637 | fit = stats.linregress([-2,-1,1,2],angAllCCF[j,:]) |
|
|
635 | 638 | phaseInt[j] = fit[1] |
|
|
636 | 639 | #Phase Differences |
|
|
637 | 640 | phaseDiff = phaseInt - phaseCCF[:,2,:] |
|
|
638 | 641 | phaseArrival = phaseInt.reshape(phaseInt.size) |
|
|
639 | 642 | |
|
|
640 | 643 | #Dealias |
|
|
641 | 644 | indAlias = numpy.where(phaseArrival > numpy.pi) |
|
|
642 | 645 | phaseArrival[indAlias] -= 2*numpy.pi |
|
|
643 | 646 | indAlias = numpy.where(phaseArrival < -numpy.pi) |
|
|
644 | 647 | phaseArrival[indAlias] += 2*numpy.pi |
|
|
645 | 648 | |
|
|
646 | 649 | return phaseDiff, phaseArrival |
|
|
647 | 650 | |
|
|
648 | 651 | def __coherentDetection(self, volts, timeSegment, timeInterval, pairslist, thresh): |
|
|
649 | 652 | #this function will run the coherent detection used in Holdworth et al. 2004 and return the net power |
|
|
650 | 653 | #find the phase shifts of each channel over 1 second intervals |
|
|
651 | 654 | #only look at ranges below the beacon signal |
|
|
652 | 655 | numProfPerBlock = numpy.ceil(timeSegment/timeInterval) |
|
|
653 | 656 | numBlocks = int(volts.shape[1]/numProfPerBlock) |
|
|
654 | 657 | numHeights = volts.shape[2] |
|
|
655 | 658 | nChannel = volts.shape[0] |
|
|
656 | 659 | voltsCohDet = volts.copy() |
|
|
657 | 660 | |
|
|
658 | 661 | pairsarray = numpy.array(pairslist) |
|
|
659 | 662 | indSides = pairsarray[:,1] |
|
|
660 | 663 | # indSides = numpy.array(range(nChannel)) |
|
|
661 | 664 | # indSides = numpy.delete(indSides, indCenter) |
|
|
662 | 665 | # |
|
|
663 | 666 | # listCenter = numpy.array_split(volts[indCenter,:,:], numBlocks, 0) |
|
|
664 | 667 | listBlocks = numpy.array_split(volts, numBlocks, 1) |
|
|
665 | 668 | |
|
|
666 | 669 | startInd = 0 |
|
|
667 | 670 | endInd = 0 |
|
|
668 | 671 | |
|
|
669 | 672 | for i in range(numBlocks): |
|
|
670 | 673 | startInd = endInd |
|
|
671 | 674 | endInd = endInd + listBlocks[i].shape[1] |
|
|
672 | 675 | |
|
|
673 | 676 | arrayBlock = listBlocks[i] |
|
|
674 | 677 | # arrayBlockCenter = listCenter[i] |
|
|
675 | 678 | |
|
|
676 | 679 | #Estimate the Phase Difference |
|
|
677 | 680 | phaseDiff, aux = self.__estimatePhaseDifference(arrayBlock, pairslist) |
|
|
678 | 681 | #Phase Difference RMS |
|
|
679 | 682 | arrayPhaseRMS = numpy.abs(phaseDiff) |
|
|
680 | 683 | phaseRMSaux = numpy.sum(arrayPhaseRMS < thresh,0) |
|
|
681 | 684 | indPhase = numpy.where(phaseRMSaux==4) |
|
|
682 | 685 | #Shifting |
|
|
683 | 686 | if indPhase[0].shape[0] > 0: |
|
|
684 | 687 | for j in range(indSides.size): |
|
|
685 | 688 | arrayBlock[indSides[j],:,indPhase] = self.__shiftPhase(arrayBlock[indSides[j],:,indPhase], phaseDiff[j,indPhase].transpose()) |
|
|
686 | 689 | voltsCohDet[:,startInd:endInd,:] = arrayBlock |
|
|
687 | 690 | |
|
|
688 | 691 | return voltsCohDet |
|
|
689 | 692 | |
|
|
690 | 693 | def __calculateCCF(self, volts, pairslist ,laglist): |
|
|
691 | 694 | |
|
|
692 | 695 | nHeights = volts.shape[2] |
|
|
693 | 696 | nPoints = volts.shape[1] |
|
|
694 | 697 | voltsCCF = numpy.zeros((len(pairslist), len(laglist), nHeights),dtype = 'complex') |
|
|
695 | 698 | |
|
|
696 | 699 | for i in range(len(pairslist)): |
|
|
697 | 700 | volts1 = volts[pairslist[i][0]] |
|
|
698 | 701 | volts2 = volts[pairslist[i][1]] |
|
|
699 | 702 | |
|
|
700 | 703 | for t in range(len(laglist)): |
|
|
701 | 704 | idxT = laglist[t] |
|
|
702 | 705 | if idxT >= 0: |
|
|
703 | 706 | vStacked = numpy.vstack((volts2[idxT:,:], |
|
|
704 | 707 | numpy.zeros((idxT, nHeights),dtype='complex'))) |
|
|
705 | 708 | else: |
|
|
706 | 709 | vStacked = numpy.vstack((numpy.zeros((-idxT, nHeights),dtype='complex'), |
|
|
707 | 710 | volts2[:(nPoints + idxT),:])) |
|
|
708 | 711 | voltsCCF[i,t,:] = numpy.sum((numpy.conjugate(volts1)*vStacked),axis=0) |
|
|
709 | 712 | |
|
|
710 | 713 | vStacked = None |
|
|
711 | 714 | return voltsCCF |
|
|
712 | 715 | |
|
|
713 | 716 | def __getNoise(self, power, timeSegment, timeInterval): |
|
|
714 | 717 | numProfPerBlock = numpy.ceil(timeSegment/timeInterval) |
|
|
715 | 718 | numBlocks = int(power.shape[0]/numProfPerBlock) |
|
|
716 | 719 | numHeights = power.shape[1] |
|
|
717 | 720 | |
|
|
718 | 721 | listPower = numpy.array_split(power, numBlocks, 0) |
|
|
719 | 722 | noise = numpy.zeros((power.shape[0], power.shape[1])) |
|
|
720 | 723 | noise1 = numpy.zeros((power.shape[0], power.shape[1])) |
|
|
721 | 724 | |
|
|
722 | 725 | startInd = 0 |
|
|
723 | 726 | endInd = 0 |
|
|
724 | 727 | |
|
|
725 | 728 | for i in range(numBlocks): #split por canal |
|
|
726 | 729 | startInd = endInd |
|
|
727 | 730 | endInd = endInd + listPower[i].shape[0] |
|
|
728 | 731 | |
|
|
729 | 732 | arrayBlock = listPower[i] |
|
|
730 | 733 | noiseAux = numpy.mean(arrayBlock, 0) |
|
|
731 | 734 | # noiseAux = numpy.median(noiseAux) |
|
|
732 | 735 | # noiseAux = numpy.mean(arrayBlock) |
|
|
733 | 736 | noise[startInd:endInd,:] = noise[startInd:endInd,:] + noiseAux |
|
|
734 | 737 | |
|
|
735 | 738 | noiseAux1 = numpy.mean(arrayBlock) |
|
|
736 | 739 | noise1[startInd:endInd,:] = noise1[startInd:endInd,:] + noiseAux1 |
|
|
737 | 740 | |
|
|
738 | 741 | return noise, noise1 |
|
|
739 | 742 | |
|
|
740 | 743 | def __findMeteors(self, power, thresh): |
|
|
741 | 744 | nProf = power.shape[0] |
|
|
742 | 745 | nHeights = power.shape[1] |
|
|
743 | 746 | listMeteors = [] |
|
|
744 | 747 | |
|
|
745 | 748 | for i in range(nHeights): |
|
|
746 | 749 | powerAux = power[:,i] |
|
|
747 | 750 | threshAux = thresh[:,i] |
|
|
748 | 751 | |
|
|
749 | 752 | indUPthresh = numpy.where(powerAux > threshAux)[0] |
|
|
750 | 753 | indDNthresh = numpy.where(powerAux <= threshAux)[0] |
|
|
751 | 754 | |
|
|
752 | 755 | j = 0 |
|
|
753 | 756 | |
|
|
754 | 757 | while (j < indUPthresh.size - 2): |
|
|
755 | 758 | if (indUPthresh[j + 2] == indUPthresh[j] + 2): |
|
|
756 | 759 | indDNAux = numpy.where(indDNthresh > indUPthresh[j]) |
|
|
757 | 760 | indDNthresh = indDNthresh[indDNAux] |
|
|
758 | 761 | |
|
|
759 | 762 | if (indDNthresh.size > 0): |
|
|
760 | 763 | indEnd = indDNthresh[0] - 1 |
|
|
761 | 764 | indInit = indUPthresh[j] |
|
|
762 | 765 | |
|
|
763 | 766 | meteor = powerAux[indInit:indEnd + 1] |
|
|
764 | 767 | indPeak = meteor.argmax() + indInit |
|
|
765 | 768 | FLA = sum(numpy.conj(meteor)*numpy.hstack((meteor[1:],0))) |
|
|
766 | 769 | |
|
|
767 | 770 | listMeteors.append(numpy.array([i,indInit,indPeak,indEnd,FLA])) #CHEQUEAR!!!!! |
|
|
768 | 771 | j = numpy.where(indUPthresh == indEnd)[0] + 1 |
|
|
769 | 772 | else: j+=1 |
|
|
770 | 773 | else: j+=1 |
|
|
771 | 774 | |
|
|
772 | 775 | return listMeteors |
|
|
773 | 776 | |
|
|
774 | 777 | def __removeMultipleDetections(self,listMeteors, rangeLimit, timeLimit): |
|
|
775 | 778 | |
|
|
776 | 779 | arrayMeteors = numpy.asarray(listMeteors) |
|
|
777 | 780 | listMeteors1 = [] |
|
|
778 | 781 | |
|
|
779 | 782 | while arrayMeteors.shape[0] > 0: |
|
|
780 | 783 | FLAs = arrayMeteors[:,4] |
|
|
781 | 784 | maxFLA = FLAs.argmax() |
|
|
782 | 785 | listMeteors1.append(arrayMeteors[maxFLA,:]) |
|
|
783 | 786 | |
|
|
784 | 787 | MeteorInitTime = arrayMeteors[maxFLA,1] |
|
|
785 | 788 | MeteorEndTime = arrayMeteors[maxFLA,3] |
|
|
786 | 789 | MeteorHeight = arrayMeteors[maxFLA,0] |
|
|
787 | 790 | |
|
|
788 | 791 | #Check neighborhood |
|
|
789 | 792 | maxHeightIndex = MeteorHeight + rangeLimit |
|
|
790 | 793 | minHeightIndex = MeteorHeight - rangeLimit |
|
|
791 | 794 | minTimeIndex = MeteorInitTime - timeLimit |
|
|
792 | 795 | maxTimeIndex = MeteorEndTime + timeLimit |
|
|
793 | 796 | |
|
|
794 | 797 | #Check Heights |
|
|
795 | 798 | indHeight = numpy.logical_and(arrayMeteors[:,0] >= minHeightIndex, arrayMeteors[:,0] <= maxHeightIndex) |
|
|
796 | 799 | indTime = numpy.logical_and(arrayMeteors[:,3] >= minTimeIndex, arrayMeteors[:,1] <= maxTimeIndex) |
|
|
797 | 800 | indBoth = numpy.where(numpy.logical_and(indTime,indHeight)) |
|
|
798 | 801 | |
|
|
799 | 802 | arrayMeteors = numpy.delete(arrayMeteors, indBoth, axis = 0) |
|
|
800 | 803 | |
|
|
801 | 804 | return listMeteors1 |
|
|
802 | 805 | |
|
|
803 | 806 | def __meteorReestimation(self, listMeteors, volts, pairslist, thresh, noise, timeInterval,frequency): |
|
|
804 | 807 | numHeights = volts.shape[2] |
|
|
805 | 808 | nChannel = volts.shape[0] |
|
|
806 | 809 | |
|
|
807 | 810 | thresholdPhase = thresh[0] |
|
|
808 | 811 | thresholdNoise = thresh[1] |
|
|
809 | 812 | thresholdDB = float(thresh[2]) |
|
|
810 | 813 | |
|
|
811 | 814 | thresholdDB1 = 10**(thresholdDB/10) |
|
|
812 | 815 | pairsarray = numpy.array(pairslist) |
|
|
813 | 816 | indSides = pairsarray[:,1] |
|
|
814 | 817 | |
|
|
815 | 818 | pairslist1 = list(pairslist) |
|
|
816 | 819 | pairslist1.append((0,1)) |
|
|
817 | 820 | pairslist1.append((3,4)) |
|
|
818 | 821 | |
|
|
819 | 822 | listMeteors1 = [] |
|
|
820 | 823 | listPowerSeries = [] |
|
|
821 | 824 | listVoltageSeries = [] |
|
|
822 | 825 | #volts has the war data |
|
|
823 | 826 | |
|
|
824 | 827 | if frequency == 30e6: |
|
|
825 | 828 | timeLag = 45*10**-3 |
|
|
826 | 829 | else: |
|
|
827 | 830 | timeLag = 15*10**-3 |
|
|
828 | 831 | lag = numpy.ceil(timeLag/timeInterval) |
|
|
829 | 832 | |
|
|
830 | 833 | for i in range(len(listMeteors)): |
|
|
831 | 834 | |
|
|
832 | 835 | ###################### 3.6 - 3.7 PARAMETERS REESTIMATION ######################### |
|
|
833 | 836 | meteorAux = numpy.zeros(16) |
|
|
834 | 837 | |
|
|
835 | 838 | #Loading meteor Data (mHeight, mStart, mPeak, mEnd) |
|
|
836 | 839 | mHeight = listMeteors[i][0] |
|
|
837 | 840 | mStart = listMeteors[i][1] |
|
|
838 | 841 | mPeak = listMeteors[i][2] |
|
|
839 | 842 | mEnd = listMeteors[i][3] |
|
|
840 | 843 | |
|
|
841 | 844 | #get the volt data between the start and end times of the meteor |
|
|
842 | 845 | meteorVolts = volts[:,mStart:mEnd+1,mHeight] |
|
|
843 | 846 | meteorVolts = meteorVolts.reshape(meteorVolts.shape[0], meteorVolts.shape[1], 1) |
|
|
844 | 847 | |
|
|
845 | 848 | #3.6. Phase Difference estimation |
|
|
846 | 849 | phaseDiff, aux = self.__estimatePhaseDifference(meteorVolts, pairslist) |
|
|
847 | 850 | |
|
|
848 | 851 | #3.7. Phase difference removal & meteor start, peak and end times reestimated |
|
|
849 | 852 | #meteorVolts0.- all Channels, all Profiles |
|
|
850 | 853 | meteorVolts0 = volts[:,:,mHeight] |
|
|
851 | 854 | meteorThresh = noise[:,mHeight]*thresholdNoise |
|
|
852 | 855 | meteorNoise = noise[:,mHeight] |
|
|
853 | 856 | meteorVolts0[indSides,:] = self.__shiftPhase(meteorVolts0[indSides,:], phaseDiff) #Phase Shifting |
|
|
854 | 857 | powerNet0 = numpy.nansum(numpy.abs(meteorVolts0)**2, axis = 0) #Power |
|
|
855 | 858 | |
|
|
856 | 859 | #Times reestimation |
|
|
857 | 860 | mStart1 = numpy.where(powerNet0[:mPeak] < meteorThresh[:mPeak])[0] |
|
|
858 | 861 | if mStart1.size > 0: |
|
|
859 | 862 | mStart1 = mStart1[-1] + 1 |
|
|
860 | 863 | |
|
|
861 | 864 | else: |
|
|
862 | 865 | mStart1 = mPeak |
|
|
863 | 866 | |
|
|
864 | 867 | mEnd1 = numpy.where(powerNet0[mPeak:] < meteorThresh[mPeak:])[0][0] + mPeak - 1 |
|
|
865 | 868 | mEndDecayTime1 = numpy.where(powerNet0[mPeak:] < meteorNoise[mPeak:])[0] |
|
|
866 | 869 | if mEndDecayTime1.size == 0: |
|
|
867 | 870 | mEndDecayTime1 = powerNet0.size |
|
|
868 | 871 | else: |
|
|
869 | 872 | mEndDecayTime1 = mEndDecayTime1[0] + mPeak - 1 |
|
|
870 | 873 | # mPeak1 = meteorVolts0[mStart1:mEnd1 + 1].argmax() |
|
|
871 | 874 | |
|
|
872 | 875 | #meteorVolts1.- all Channels, from start to end |
|
|
873 | 876 | meteorVolts1 = meteorVolts0[:,mStart1:mEnd1 + 1] |
|
|
874 | 877 | meteorVolts2 = meteorVolts0[:,mPeak + lag:mEnd1 + 1] |
|
|
875 | 878 | if meteorVolts2.shape[1] == 0: |
|
|
876 | 879 | meteorVolts2 = meteorVolts0[:,mPeak:mEnd1 + 1] |
|
|
877 | 880 | meteorVolts1 = meteorVolts1.reshape(meteorVolts1.shape[0], meteorVolts1.shape[1], 1) |
|
|
878 | 881 | meteorVolts2 = meteorVolts2.reshape(meteorVolts2.shape[0], meteorVolts2.shape[1], 1) |
|
|
879 | 882 | ##################### END PARAMETERS REESTIMATION ######################### |
|
|
880 | 883 | |
|
|
881 | 884 | ##################### 3.8 PHASE DIFFERENCE REESTIMATION ######################## |
|
|
882 | 885 | # if mEnd1 - mStart1 > 4: #Error Number 6: echo less than 5 samples long; too short for analysis |
|
|
883 | 886 | if meteorVolts2.shape[1] > 0: |
|
|
884 | 887 | #Phase Difference re-estimation |
|
|
885 | 888 | phaseDiff1, phaseDiffint = self.__estimatePhaseDifference(meteorVolts2, pairslist1) #Phase Difference Estimation |
|
|
886 | 889 | # phaseDiff1, phaseDiffint = self.estimatePhaseDifference(meteorVolts2, pairslist) |
|
|
887 | 890 | meteorVolts2 = meteorVolts2.reshape(meteorVolts2.shape[0], meteorVolts2.shape[1]) |
|
|
888 | 891 | phaseDiff11 = numpy.reshape(phaseDiff1, (phaseDiff1.shape[0],1)) |
|
|
889 | 892 | meteorVolts2[indSides,:] = self.__shiftPhase(meteorVolts2[indSides,:], phaseDiff11[0:4]) #Phase Shifting |
|
|
890 | 893 | |
|
|
891 | 894 | #Phase Difference RMS |
|
|
892 | 895 | phaseRMS1 = numpy.sqrt(numpy.mean(numpy.square(phaseDiff1))) |
|
|
893 | 896 | powerNet1 = numpy.nansum(numpy.abs(meteorVolts1[:,:])**2,0) |
|
|
894 | 897 | #Data from Meteor |
|
|
895 | 898 | mPeak1 = powerNet1.argmax() + mStart1 |
|
|
896 | 899 | mPeakPower1 = powerNet1.max() |
|
|
897 | 900 | noiseAux = sum(noise[mStart1:mEnd1 + 1,mHeight]) |
|
|
898 | 901 | mSNR1 = (sum(powerNet1)-noiseAux)/noiseAux |
|
|
899 | 902 | Meteor1 = numpy.array([mHeight, mStart1, mPeak1, mEnd1, mPeakPower1, mSNR1, phaseRMS1]) |
|
|
900 | 903 | Meteor1 = numpy.hstack((Meteor1,phaseDiffint)) |
|
|
901 | 904 | PowerSeries = powerNet0[mStart1:mEndDecayTime1 + 1] |
|
|
902 | 905 | #Vectorize |
|
|
903 | 906 | meteorAux[0:7] = [mHeight, mStart1, mPeak1, mEnd1, mPeakPower1, mSNR1, phaseRMS1] |
|
|
904 | 907 | meteorAux[7:11] = phaseDiffint[0:4] |
|
|
905 | 908 | |
|
|
906 | 909 | #Rejection Criterions |
|
|
907 | 910 | if phaseRMS1 > thresholdPhase: #Error Number 17: Phase variation |
|
|
908 | 911 | meteorAux[-1] = 17 |
|
|
909 | 912 | elif mSNR1 < thresholdDB1: #Error Number 1: SNR < threshold dB |
|
|
910 | 913 | meteorAux[-1] = 1 |
|
|
911 | 914 | |
|
|
912 | 915 | |
|
|
913 | 916 | else: |
|
|
914 | 917 | meteorAux[0:4] = [mHeight, mStart, mPeak, mEnd] |
|
|
915 | 918 | meteorAux[-1] = 6 #Error Number 6: echo less than 5 samples long; too short for analysis |
|
|
916 | 919 | PowerSeries = 0 |
|
|
917 | 920 | |
|
|
918 | 921 | listMeteors1.append(meteorAux) |
|
|
919 | 922 | listPowerSeries.append(PowerSeries) |
|
|
920 | 923 | listVoltageSeries.append(meteorVolts1) |
|
|
921 | 924 | |
|
|
922 | 925 | return listMeteors1, listPowerSeries, listVoltageSeries |
|
|
923 | 926 | |
|
|
924 | 927 | def __estimateDecayTime(self, listMeteors, listPower, timeInterval, frequency): |
|
|
925 | 928 | |
|
|
926 | 929 | threshError = 10 |
|
|
927 | 930 | #Depending if it is 30 or 50 MHz |
|
|
928 | 931 | if frequency == 30e6: |
|
|
929 | 932 | timeLag = 45*10**-3 |
|
|
930 | 933 | else: |
|
|
931 | 934 | timeLag = 15*10**-3 |
|
|
932 | 935 | lag = numpy.ceil(timeLag/timeInterval) |
|
|
933 | 936 | |
|
|
934 | 937 | listMeteors1 = [] |
|
|
935 | 938 | |
|
|
936 | 939 | for i in range(len(listMeteors)): |
|
|
937 | 940 | meteorPower = listPower[i] |
|
|
938 | 941 | meteorAux = listMeteors[i] |
|
|
939 | 942 | |
|
|
940 | 943 | if meteorAux[-1] == 0: |
|
|
941 | 944 | |
|
|
942 | 945 | try: |
|
|
943 | 946 | indmax = meteorPower.argmax() |
|
|
944 | 947 | indlag = indmax + lag |
|
|
945 | 948 | |
|
|
946 | 949 | y = meteorPower[indlag:] |
|
|
947 | 950 | x = numpy.arange(0, y.size)*timeLag |
|
|
948 | 951 | |
|
|
949 | 952 | #first guess |
|
|
950 | 953 | a = y[0] |
|
|
951 | 954 | tau = timeLag |
|
|
952 | 955 | #exponential fit |
|
|
953 | 956 | popt, pcov = optimize.curve_fit(self.__exponential_function, x, y, p0 = [a, tau]) |
|
|
954 | 957 | y1 = self.__exponential_function(x, *popt) |
|
|
955 | 958 | #error estimation |
|
|
956 | 959 | error = sum((y - y1)**2)/(numpy.var(y)*(y.size - popt.size)) |
|
|
957 | 960 | |
|
|
958 | 961 | decayTime = popt[1] |
|
|
959 | 962 | riseTime = indmax*timeInterval |
|
|
960 | 963 | meteorAux[11:13] = [decayTime, error] |
|
|
961 | 964 | |
|
|
962 | 965 | #Table items 7, 8 and 11 |
|
|
963 | 966 | if (riseTime > 0.3): #Number 7: Echo rise exceeds 0.3s |
|
|
964 | 967 | meteorAux[-1] = 7 |
|
|
965 | 968 | elif (decayTime < 2*riseTime) : #Number 8: Echo decay time less than than twice rise time |
|
|
966 | 969 | meteorAux[-1] = 8 |
|
|
967 | 970 | if (error > threshError): #Number 11: Poor fit to amplitude for estimation of decay time |
|
|
968 | 971 | meteorAux[-1] = 11 |
|
|
969 | 972 | |
|
|
970 | 973 | |
|
|
971 | 974 | except: |
|
|
972 | 975 | meteorAux[-1] = 11 |
|
|
973 | 976 | |
|
|
974 | 977 | |
|
|
975 | 978 | listMeteors1.append(meteorAux) |
|
|
976 | 979 | |
|
|
977 | 980 | return listMeteors1 |
|
|
978 | 981 | |
|
|
979 | 982 | #Exponential Function |
|
|
980 | 983 | |
|
|
981 | 984 | def __exponential_function(self, x, a, tau): |
|
|
982 | 985 | y = a*numpy.exp(-x/tau) |
|
|
983 | 986 | return y |
|
|
984 | 987 | |
|
|
985 | 988 | def __getRadialVelocity(self, listMeteors, listVolts, radialStdThresh, pairslist, timeInterval): |
|
|
986 | 989 | |
|
|
987 | 990 | pairslist1 = list(pairslist) |
|
|
988 | 991 | pairslist1.append((0,1)) |
|
|
989 | 992 | pairslist1.append((3,4)) |
|
|
990 | 993 | numPairs = len(pairslist1) |
|
|
991 | 994 | #Time Lag |
|
|
992 | 995 | timeLag = 45*10**-3 |
|
|
993 | 996 | c = 3e8 |
|
|
994 | 997 | lag = numpy.ceil(timeLag/timeInterval) |
|
|
995 | 998 | freq = 30e6 |
|
|
996 | 999 | |
|
|
997 | 1000 | listMeteors1 = [] |
|
|
998 | 1001 | |
|
|
999 | 1002 | for i in range(len(listMeteors)): |
|
|
1000 | 1003 | meteorAux = listMeteors[i] |
|
|
1001 | 1004 | if meteorAux[-1] == 0: |
|
|
1002 | 1005 | mStart = listMeteors[i][1] |
|
|
1003 | 1006 | mPeak = listMeteors[i][2] |
|
|
1004 | 1007 | mLag = mPeak - mStart + lag |
|
|
1005 | 1008 | |
|
|
1006 | 1009 | #get the volt data between the start and end times of the meteor |
|
|
1007 | 1010 | meteorVolts = listVolts[i] |
|
|
1008 | 1011 | meteorVolts = meteorVolts.reshape(meteorVolts.shape[0], meteorVolts.shape[1], 1) |
|
|
1009 | 1012 | |
|
|
1010 | 1013 | #Get CCF |
|
|
1011 | 1014 | allCCFs = self.__calculateCCF(meteorVolts, pairslist1, [-2,-1,0,1,2]) |
|
|
1012 | 1015 | |
|
|
1013 | 1016 | #Method 2 |
|
|
1014 | 1017 | slopes = numpy.zeros(numPairs) |
|
|
1015 | 1018 | time = numpy.array([-2,-1,1,2])*timeInterval |
|
|
1016 | 1019 | angAllCCF = numpy.angle(allCCFs[:,[0,1,3,4],0]) |
|
|
1017 | 1020 | |
|
|
1018 | 1021 | #Correct phases |
|
|
1019 | 1022 | derPhaseCCF = angAllCCF[:,1:] - angAllCCF[:,0:-1] |
|
|
1020 | 1023 | indDer = numpy.where(numpy.abs(derPhaseCCF) > numpy.pi) |
|
|
1021 | 1024 | |
|
|
1022 | 1025 | if indDer[0].shape[0] > 0: |
|
|
1023 | 1026 | for i in range(indDer[0].shape[0]): |
|
|
1024 | 1027 | signo = -numpy.sign(derPhaseCCF[indDer[0][i],indDer[1][i]]) |
|
|
1025 | 1028 | angAllCCF[indDer[0][i],indDer[1][i]+1:] += signo*2*numpy.pi |
|
|
1026 | 1029 | |
|
|
1027 | 1030 | # fit = scipy.stats.linregress(numpy.array([-2,-1,1,2])*timeInterval, numpy.array([phaseLagN2s[i],phaseLagN1s[i],phaseLag1s[i],phaseLag2s[i]])) |
|
|
1028 | 1031 | for j in range(numPairs): |
|
|
1029 | 1032 | fit = stats.linregress(time, angAllCCF[j,:]) |
|
|
1030 | 1033 | slopes[j] = fit[0] |
|
|
1031 | 1034 | |
|
|
1032 | 1035 | #Remove Outlier |
|
|
1033 | 1036 | # indOut = numpy.argmax(numpy.abs(slopes - numpy.mean(slopes))) |
|
|
1034 | 1037 | # slopes = numpy.delete(slopes,indOut) |
|
|
1035 | 1038 | # indOut = numpy.argmax(numpy.abs(slopes - numpy.mean(slopes))) |
|
|
1036 | 1039 | # slopes = numpy.delete(slopes,indOut) |
|
|
1037 | 1040 | |
|
|
1038 | 1041 | radialVelocity = -numpy.mean(slopes)*(0.25/numpy.pi)*(c/freq) |
|
|
1039 | 1042 | radialError = numpy.std(slopes)*(0.25/numpy.pi)*(c/freq) |
|
|
1040 | 1043 | meteorAux[-2] = radialError |
|
|
1041 | 1044 | meteorAux[-3] = radialVelocity |
|
|
1042 | 1045 | |
|
|
1043 | 1046 | #Setting Error |
|
|
1044 | 1047 | #Number 15: Radial Drift velocity or projected horizontal velocity exceeds 200 m/s |
|
|
1045 | 1048 | if numpy.abs(radialVelocity) > 200: |
|
|
1046 | 1049 | meteorAux[-1] = 15 |
|
|
1047 | 1050 | #Number 12: Poor fit to CCF variation for estimation of radial drift velocity |
|
|
1048 | 1051 | elif radialError > radialStdThresh: |
|
|
1049 | 1052 | meteorAux[-1] = 12 |
|
|
1050 | 1053 | |
|
|
1051 | 1054 | listMeteors1.append(meteorAux) |
|
|
1052 | 1055 | return listMeteors1 |
|
|
1053 | 1056 | |
|
|
1054 | 1057 | def __setNewArrays(self, listMeteors, date, heiRang): |
|
|
1055 | 1058 | |
|
|
1056 | 1059 | #New arrays |
|
|
1057 | 1060 | arrayMeteors = numpy.array(listMeteors) |
|
|
1058 | 1061 | arrayParameters = numpy.zeros((len(listMeteors), 13)) |
|
|
1059 | 1062 | |
|
|
1060 | 1063 | #Date inclusion |
|
|
1061 | 1064 | # date = re.findall(r'\((.*?)\)', date) |
|
|
1062 | 1065 | # date = date[0].split(',') |
|
|
1063 | 1066 | # date = map(int, date) |
|
|
1064 | 1067 | # |
|
|
1065 | 1068 | # if len(date)<6: |
|
|
1066 | 1069 | # date.append(0) |
|
|
1067 | 1070 | # |
|
|
1068 | 1071 | # date = [date[0]*10000 + date[1]*100 + date[2], date[3]*10000 + date[4]*100 + date[5]] |
|
|
1069 | 1072 | # arrayDate = numpy.tile(date, (len(listMeteors), 1)) |
|
|
1070 | 1073 | arrayDate = numpy.tile(date, (len(listMeteors))) |
|
|
1071 | 1074 | |
|
|
1072 | 1075 | #Meteor array |
|
|
1073 | 1076 | # arrayMeteors[:,0] = heiRang[arrayMeteors[:,0].astype(int)] |
|
|
1074 | 1077 | # arrayMeteors = numpy.hstack((arrayDate, arrayMeteors)) |
|
|
1075 | 1078 | |
|
|
1076 | 1079 | #Parameters Array |
|
|
1077 | 1080 | arrayParameters[:,0] = arrayDate #Date |
|
|
1078 | 1081 | arrayParameters[:,1] = heiRang[arrayMeteors[:,0].astype(int)] #Range |
|
|
1079 | 1082 | arrayParameters[:,6:8] = arrayMeteors[:,-3:-1] #Radial velocity and its error |
|
|
1080 | 1083 | arrayParameters[:,8:12] = arrayMeteors[:,7:11] #Phases |
|
|
1081 | 1084 | arrayParameters[:,-1] = arrayMeteors[:,-1] #Error |
|
|
1082 | 1085 | |
|
|
1083 | 1086 | |
|
|
1084 | 1087 | return arrayParameters |
|
|
1085 | 1088 | |
|
|
1086 | 1089 | # def __getAOA(self, phases, pairsList, error, AOAthresh, azimuth): |
|
|
1087 | 1090 | # |
|
|
1088 | 1091 | # arrayAOA = numpy.zeros((phases.shape[0],3)) |
|
|
1089 | 1092 | # cosdir0, cosdir = self.__getDirectionCosines(phases, pairsList) |
|
|
1090 | 1093 | # |
|
|
1091 | 1094 | # arrayAOA[:,:2] = self.__calculateAOA(cosdir, azimuth) |
|
|
1092 | 1095 | # cosDirError = numpy.sum(numpy.abs(cosdir0 - cosdir), axis = 1) |
|
|
1093 | 1096 | # arrayAOA[:,2] = cosDirError |
|
|
1094 | 1097 | # |
|
|
1095 | 1098 | # azimuthAngle = arrayAOA[:,0] |
|
|
1096 | 1099 | # zenithAngle = arrayAOA[:,1] |
|
|
1097 | 1100 | # |
|
|
1098 | 1101 | # #Setting Error |
|
|
1099 | 1102 | # #Number 3: AOA not fesible |
|
|
1100 | 1103 | # indInvalid = numpy.where(numpy.logical_and((numpy.logical_or(numpy.isnan(zenithAngle), numpy.isnan(azimuthAngle))),error == 0))[0] |
|
|
1101 | 1104 | # error[indInvalid] = 3 |
|
|
1102 | 1105 | # #Number 4: Large difference in AOAs obtained from different antenna baselines |
|
|
1103 | 1106 | # indInvalid = numpy.where(numpy.logical_and(cosDirError > AOAthresh,error == 0))[0] |
|
|
1104 | 1107 | # error[indInvalid] = 4 |
|
|
1105 | 1108 | # return arrayAOA, error |
|
|
1106 | 1109 | # |
|
|
1107 | 1110 | # def __getDirectionCosines(self, arrayPhase, pairsList): |
|
|
1108 | 1111 | # |
|
|
1109 | 1112 | # #Initializing some variables |
|
|
1110 | 1113 | # ang_aux = numpy.array([-8,-7,-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6,7,8])*2*numpy.pi |
|
|
1111 | 1114 | # ang_aux = ang_aux.reshape(1,ang_aux.size) |
|
|
1112 | 1115 | # |
|
|
1113 | 1116 | # cosdir = numpy.zeros((arrayPhase.shape[0],2)) |
|
|
1114 | 1117 | # cosdir0 = numpy.zeros((arrayPhase.shape[0],2)) |
|
|
1115 | 1118 | # |
|
|
1116 | 1119 | # |
|
|
1117 | 1120 | # for i in range(2): |
|
|
1118 | 1121 | # #First Estimation |
|
|
1119 | 1122 | # phi0_aux = arrayPhase[:,pairsList[i][0]] + arrayPhase[:,pairsList[i][1]] |
|
|
1120 | 1123 | # #Dealias |
|
|
1121 | 1124 | # indcsi = numpy.where(phi0_aux > numpy.pi) |
|
|
1122 | 1125 | # phi0_aux[indcsi] -= 2*numpy.pi |
|
|
1123 | 1126 | # indcsi = numpy.where(phi0_aux < -numpy.pi) |
|
|
1124 | 1127 | # phi0_aux[indcsi] += 2*numpy.pi |
|
|
1125 | 1128 | # #Direction Cosine 0 |
|
|
1126 | 1129 | # cosdir0[:,i] = -(phi0_aux)/(2*numpy.pi*0.5) |
|
|
1127 | 1130 | # |
|
|
1128 | 1131 | # #Most-Accurate Second Estimation |
|
|
1129 | 1132 | # phi1_aux = arrayPhase[:,pairsList[i][0]] - arrayPhase[:,pairsList[i][1]] |
|
|
1130 | 1133 | # phi1_aux = phi1_aux.reshape(phi1_aux.size,1) |
|
|
1131 | 1134 | # #Direction Cosine 1 |
|
|
1132 | 1135 | # cosdir1 = -(phi1_aux + ang_aux)/(2*numpy.pi*4.5) |
|
|
1133 | 1136 | # |
|
|
1134 | 1137 | # #Searching the correct Direction Cosine |
|
|
1135 | 1138 | # cosdir0_aux = cosdir0[:,i] |
|
|
1136 | 1139 | # cosdir0_aux = cosdir0_aux.reshape(cosdir0_aux.size,1) |
|
|
1137 | 1140 | # #Minimum Distance |
|
|
1138 | 1141 | # cosDiff = (cosdir1 - cosdir0_aux)**2 |
|
|
1139 | 1142 | # indcos = cosDiff.argmin(axis = 1) |
|
|
1140 | 1143 | # #Saving Value obtained |
|
|
1141 | 1144 | # cosdir[:,i] = cosdir1[numpy.arange(len(indcos)),indcos] |
|
|
1142 | 1145 | # |
|
|
1143 | 1146 | # return cosdir0, cosdir |
|
|
1144 | 1147 | # |
|
|
1145 | 1148 | # def __calculateAOA(self, cosdir, azimuth): |
|
|
1146 | 1149 | # cosdirX = cosdir[:,0] |
|
|
1147 | 1150 | # cosdirY = cosdir[:,1] |
|
|
1148 | 1151 | # |
|
|
1149 | 1152 | # zenithAngle = numpy.arccos(numpy.sqrt(1 - cosdirX**2 - cosdirY**2))*180/numpy.pi |
|
|
1150 | 1153 | # azimuthAngle = numpy.arctan2(cosdirX,cosdirY)*180/numpy.pi + azimuth #0 deg north, 90 deg east |
|
|
1151 | 1154 | # angles = numpy.vstack((azimuthAngle, zenithAngle)).transpose() |
|
|
1152 | 1155 | # |
|
|
1153 | 1156 | # return angles |
|
|
1154 | 1157 | # |
|
|
1155 | 1158 | # def __getHeights(self, Ranges, zenith, error, minHeight, maxHeight): |
|
|
1156 | 1159 | # |
|
|
1157 | 1160 | # Ramb = 375 #Ramb = c/(2*PRF) |
|
|
1158 | 1161 | # Re = 6371 #Earth Radius |
|
|
1159 | 1162 | # heights = numpy.zeros(Ranges.shape) |
|
|
1160 | 1163 | # |
|
|
1161 | 1164 | # R_aux = numpy.array([0,1,2])*Ramb |
|
|
1162 | 1165 | # R_aux = R_aux.reshape(1,R_aux.size) |
|
|
1163 | 1166 | # |
|
|
1164 | 1167 | # Ranges = Ranges.reshape(Ranges.size,1) |
|
|
1165 | 1168 | # |
|
|
1166 | 1169 | # Ri = Ranges + R_aux |
|
|
1167 | 1170 | # hi = numpy.sqrt(Re**2 + Ri**2 + (2*Re*numpy.cos(zenith*numpy.pi/180)*Ri.transpose()).transpose()) - Re |
|
|
1168 | 1171 | # |
|
|
1169 | 1172 | # #Check if there is a height between 70 and 110 km |
|
|
1170 | 1173 | # h_bool = numpy.sum(numpy.logical_and(hi > minHeight, hi < maxHeight), axis = 1) |
|
|
1171 | 1174 | # ind_h = numpy.where(h_bool == 1)[0] |
|
|
1172 | 1175 | # |
|
|
1173 | 1176 | # hCorr = hi[ind_h, :] |
|
|
1174 | 1177 | # ind_hCorr = numpy.where(numpy.logical_and(hi > minHeight, hi < maxHeight)) |
|
|
1175 | 1178 | # |
|
|
1176 | 1179 | # hCorr = hi[ind_hCorr] |
|
|
1177 | 1180 | # heights[ind_h] = hCorr |
|
|
1178 | 1181 | # |
|
|
1179 | 1182 | # #Setting Error |
|
|
1180 | 1183 | # #Number 13: Height unresolvable echo: not valid height within 70 to 110 km |
|
|
1181 | 1184 | # #Number 14: Height ambiguous echo: more than one possible height within 70 to 110 km |
|
|
1182 | 1185 | # |
|
|
1183 | 1186 | # indInvalid2 = numpy.where(numpy.logical_and(h_bool > 1, error == 0))[0] |
|
|
1184 | 1187 | # error[indInvalid2] = 14 |
|
|
1185 | 1188 | # indInvalid1 = numpy.where(numpy.logical_and(h_bool == 0, error == 0))[0] |
|
|
1186 | 1189 | # error[indInvalid1] = 13 |
|
|
1187 | 1190 | # |
|
|
1188 | 1191 | # return heights, error |
|
|
1189 | 1192 | |
|
|
1190 | 1193 | def SpectralFitting(self, getSNR = True, path=None, file=None, groupList=None): |
|
|
1191 | 1194 | |
|
|
1192 | 1195 | ''' |
|
|
1193 | 1196 | Function GetMoments() |
|
|
1194 | 1197 | |
|
|
1195 | 1198 | Input: |
|
|
1196 | 1199 | Output: |
|
|
1197 | 1200 | Variables modified: |
|
|
1198 | 1201 | ''' |
|
|
1199 | 1202 | if path != None: |
|
|
1200 | 1203 | sys.path.append(path) |
|
|
1201 | 1204 | self.dataOut.library = importlib.import_module(file) |
|
|
1202 | 1205 | |
|
|
1203 | 1206 | #To be inserted as a parameter |
|
|
1204 | 1207 | groupArray = numpy.array(groupList) |
|
|
1205 | 1208 | # groupArray = numpy.array([[0,1],[2,3]]) |
|
|
1206 | 1209 | self.dataOut.groupList = groupArray |
|
|
1207 | 1210 | |
|
|
1208 | 1211 | nGroups = groupArray.shape[0] |
|
|
1209 | 1212 | nChannels = self.dataIn.nChannels |
|
|
1210 | 1213 | nHeights=self.dataIn.heightList.size |
|
|
1211 | 1214 | |
|
|
1212 | 1215 | #Parameters Array |
|
|
1213 | 1216 | self.dataOut.data_param = None |
|
|
1214 | 1217 | |
|
|
1215 | 1218 | #Set constants |
|
|
1216 | 1219 | constants = self.dataOut.library.setConstants(self.dataIn) |
|
|
1217 | 1220 | self.dataOut.constants = constants |
|
|
1218 | 1221 | M = self.dataIn.normFactor |
|
|
1219 | 1222 | N = self.dataIn.nFFTPoints |
|
|
1220 | 1223 | ippSeconds = self.dataIn.ippSeconds |
|
|
1221 | 1224 | K = self.dataIn.nIncohInt |
|
|
1222 | 1225 | pairsArray = numpy.array(self.dataIn.pairsList) |
|
|
1223 | 1226 | |
|
|
1224 | 1227 | #List of possible combinations |
|
|
1225 | 1228 | listComb = itertools.combinations(numpy.arange(groupArray.shape[1]),2) |
|
|
1226 | 1229 | indCross = numpy.zeros(len(list(listComb)), dtype = 'int') |
|
|
1227 | 1230 | |
|
|
1228 | 1231 | if getSNR: |
|
|
1229 | 1232 | listChannels = groupArray.reshape((groupArray.size)) |
|
|
1230 | 1233 | listChannels.sort() |
|
|
1231 | 1234 | noise = self.dataIn.getNoise() |
|
|
1232 | 1235 | self.dataOut.data_SNR = self.__getSNR(self.dataIn.data_spc[listChannels,:,:], noise[listChannels]) |
|
|
1233 | 1236 | |
|
|
1234 | 1237 | for i in range(nGroups): |
|
|
1235 | 1238 | coord = groupArray[i,:] |
|
|
1236 | 1239 | |
|
|
1237 | 1240 | #Input data array |
|
|
1238 | 1241 | data = self.dataIn.data_spc[coord,:,:]/(M*N) |
|
|
1239 | 1242 | data = data.reshape((data.shape[0]*data.shape[1],data.shape[2])) |
|
|
1240 | 1243 | |
|
|
1241 | 1244 | #Cross Spectra data array for Covariance Matrixes |
|
|
1242 | 1245 | ind = 0 |
|
|
1243 | 1246 | for pairs in listComb: |
|
|
1244 | 1247 | pairsSel = numpy.array([coord[x],coord[y]]) |
|
|
1245 | 1248 | indCross[ind] = int(numpy.where(numpy.all(pairsArray == pairsSel, axis = 1))[0][0]) |
|
|
1246 | 1249 | ind += 1 |
|
|
1247 | 1250 | dataCross = self.dataIn.data_cspc[indCross,:,:]/(M*N) |
|
|
1248 | 1251 | dataCross = dataCross**2/K |
|
|
1249 | 1252 | |
|
|
1250 | 1253 | for h in range(nHeights): |
|
|
1251 | 1254 | # print self.dataOut.heightList[h] |
|
|
1252 | 1255 | |
|
|
1253 | 1256 | #Input |
|
|
1254 | 1257 | d = data[:,h] |
|
|
1255 | 1258 | |
|
|
1256 | 1259 | #Covariance Matrix |
|
|
1257 | 1260 | D = numpy.diag(d**2/K) |
|
|
1258 | 1261 | ind = 0 |
|
|
1259 | 1262 | for pairs in listComb: |
|
|
1260 | 1263 | #Coordinates in Covariance Matrix |
|
|
1261 | 1264 | x = pairs[0] |
|
|
1262 | 1265 | y = pairs[1] |
|
|
1263 | 1266 | #Channel Index |
|
|
1264 | 1267 | S12 = dataCross[ind,:,h] |
|
|
1265 | 1268 | D12 = numpy.diag(S12) |
|
|
1266 | 1269 | #Completing Covariance Matrix with Cross Spectras |
|
|
1267 | 1270 | D[x*N:(x+1)*N,y*N:(y+1)*N] = D12 |
|
|
1268 | 1271 | D[y*N:(y+1)*N,x*N:(x+1)*N] = D12 |
|
|
1269 | 1272 | ind += 1 |
|
|
1270 | 1273 | Dinv=numpy.linalg.inv(D) |
|
|
1271 | 1274 | L=numpy.linalg.cholesky(Dinv) |
|
|
1272 | 1275 | LT=L.T |
|
|
1273 | 1276 | |
|
|
1274 | 1277 | dp = numpy.dot(LT,d) |
|
|
1275 | 1278 | |
|
|
1276 | 1279 | #Initial values |
|
|
1277 | 1280 | data_spc = self.dataIn.data_spc[coord,:,h] |
|
|
1278 | 1281 | |
|
|
1279 | 1282 | if (h>0)and(error1[3]<5): |
|
|
1280 | 1283 | p0 = self.dataOut.data_param[i,:,h-1] |
|
|
1281 | 1284 | else: |
|
|
1282 | 1285 | p0 = numpy.array(self.dataOut.library.initialValuesFunction(data_spc, constants, i)) |
|
|
1283 | 1286 | |
|
|
1284 | 1287 | try: |
|
|
1285 | 1288 | #Least Squares |
|
|
1286 | 1289 | minp,covp,infodict,mesg,ier = optimize.leastsq(self.__residFunction,p0,args=(dp,LT,constants),full_output=True) |
|
|
1287 | 1290 | # minp,covp = optimize.leastsq(self.__residFunction,p0,args=(dp,LT,constants)) |
|
|
1288 | 1291 | #Chi square error |
|
|
1289 | 1292 | error0 = numpy.sum(infodict['fvec']**2)/(2*N) |
|
|
1290 | 1293 | #Error with Jacobian |
|
|
1291 | 1294 | error1 = self.dataOut.library.errorFunction(minp,constants,LT) |
|
|
1292 | 1295 | except: |
|
|
1293 | 1296 | minp = p0*numpy.nan |
|
|
1294 | 1297 | error0 = numpy.nan |
|
|
1295 | 1298 | error1 = p0*numpy.nan |
|
|
1296 | 1299 | |
|
|
1297 | 1300 | #Save |
|
|
1298 | 1301 | if self.dataOut.data_param == None: |
|
|
1299 | 1302 | self.dataOut.data_param = numpy.zeros((nGroups, p0.size, nHeights))*numpy.nan |
|
|
1300 | 1303 | self.dataOut.data_error = numpy.zeros((nGroups, p0.size + 1, nHeights))*numpy.nan |
|
|
1301 | 1304 | |
|
|
1302 | 1305 | self.dataOut.data_error[i,:,h] = numpy.hstack((error0,error1)) |
|
|
1303 | 1306 | self.dataOut.data_param[i,:,h] = minp |
|
|
1304 | 1307 | return |
|
|
1305 | 1308 | |
|
|
1306 | 1309 | def __residFunction(self, p, dp, LT, constants): |
|
|
1307 | 1310 | |
|
|
1308 | 1311 | fm = self.dataOut.library.modelFunction(p, constants) |
|
|
1309 | 1312 | fmp=numpy.dot(LT,fm) |
|
|
1310 | 1313 | |
|
|
1311 | 1314 | return dp-fmp |
|
|
1312 | 1315 | |
|
|
1313 | 1316 | def __getSNR(self, z, noise): |
|
|
1314 | 1317 | |
|
|
1315 | 1318 | avg = numpy.average(z, axis=1) |
|
|
1316 | 1319 | SNR = (avg.T-noise)/noise |
|
|
1317 | 1320 | SNR = SNR.T |
|
|
1318 | 1321 | return SNR |
|
|
1319 | 1322 | |
|
|
1320 | 1323 | def __chisq(p,chindex,hindex): |
|
|
1321 | 1324 | #similar to Resid but calculates CHI**2 |
|
|
1322 | 1325 | [LT,d,fm]=setupLTdfm(p,chindex,hindex) |
|
|
1323 | 1326 | dp=numpy.dot(LT,d) |
|
|
1324 | 1327 | fmp=numpy.dot(LT,fm) |
|
|
1325 | 1328 | chisq=numpy.dot((dp-fmp).T,(dp-fmp)) |
|
|
1326 | 1329 | return chisq |
|
|
1327 | 1330 |
|
|
|
1331 | def NonSpecularMeteorDetection(self, mode, SNRthresh=8, phaseDerThresh=0.5, cohThresh=0.8, allData = False): | |
|
|
1332 | data_acf = self.dataOut.data_pre[0] | |
|
|
1333 | data_ccf = self.dataOut.data_pre[1] | |
|
|
1334 | ||
|
|
1335 | lamb = self.dataOut.C/self.dataOut.frequency | |
|
|
1336 | tSamp = self.dataOut.ippSeconds*self.dataOut.nCohInt | |
|
|
1337 | paramInterval = self.dataOut.paramInterval | |
|
|
1338 | ||
|
|
1339 | nChannels = data_acf.shape[0] | |
|
|
1340 | nLags = data_acf.shape[1] | |
|
|
1341 | nProfiles = data_acf.shape[2] | |
|
|
1342 | nHeights = self.dataOut.nHeights | |
|
|
1343 | nCohInt = self.dataOut.nCohInt | |
|
|
1344 | sec = numpy.round(nProfiles/self.dataOut.paramInterval) | |
|
|
1345 | heightList = self.dataOut.heightList | |
|
|
1346 | ippSeconds = self.dataOut.ippSeconds*self.dataOut.nCohInt*self.dataOut.nAvg | |
|
|
1347 | utctime = self.dataOut.utctime | |
|
|
1348 | ||
|
|
1349 | self.dataOut.abscissaList = numpy.arange(0,paramInterval+ippSeconds,ippSeconds) | |
|
|
1350 | ||
|
|
1351 | #------------------------ SNR -------------------------------------- | |
|
|
1352 | power = data_acf[:,0,:,:].real | |
|
|
1353 | noise = numpy.zeros(nChannels) | |
|
|
1354 | SNR = numpy.zeros(power.shape) | |
|
|
1355 | for i in range(nChannels): | |
|
|
1356 | noise[i] = hildebrand_sekhon(power[i,:], nCohInt) | |
|
|
1357 | SNR[i] = (power[i]-noise[i])/noise[i] | |
|
|
1358 | SNRm = numpy.nanmean(SNR, axis = 0) | |
|
|
1359 | SNRdB = 10*numpy.log10(SNR) | |
|
|
1360 | ||
|
|
1361 | if mode == 'SA': | |
|
|
1362 | nPairs = data_ccf.shape[0] | |
|
|
1363 | #---------------------- Coherence and Phase -------------------------- | |
|
|
1364 | phase = numpy.zeros(data_ccf[:,0,:,:].shape) | |
|
|
1365 | # phase1 = numpy.copy(phase) | |
|
|
1366 | coh1 = numpy.zeros(data_ccf[:,0,:,:].shape) | |
|
|
1367 | ||
|
|
1368 | for p in range(nPairs): | |
|
|
1369 | ch0 = self.dataOut.groupList[p][0] | |
|
|
1370 | ch1 = self.dataOut.groupList[p][1] | |
|
|
1371 | ccf = data_ccf[p,0,:,:]/numpy.sqrt(data_acf[ch0,0,:,:]*data_acf[ch1,0,:,:]) | |
|
|
1372 | phase[p,:,:] = ndimage.median_filter(numpy.angle(ccf), size = (5,1)) #median filter | |
|
|
1373 | # phase1[p,:,:] = numpy.angle(ccf) #median filter | |
|
|
1374 | coh1[p,:,:] = ndimage.median_filter(numpy.abs(ccf), 5) #median filter | |
|
|
1375 | # coh1[p,:,:] = numpy.abs(ccf) #median filter | |
|
|
1376 | coh = numpy.nanmax(coh1, axis = 0) | |
|
|
1377 | # struc = numpy.ones((5,1)) | |
|
|
1378 | # coh = ndimage.morphology.grey_dilation(coh, size=(10,1)) | |
|
|
1379 | #---------------------- Radial Velocity ---------------------------- | |
|
|
1380 | phaseAux = numpy.mean(numpy.angle(data_acf[:,1,:,:]), axis = 0) | |
|
|
1381 | velRad = phaseAux*lamb/(4*numpy.pi*tSamp) | |
|
|
1382 | ||
|
|
1383 | if allData: | |
|
|
1384 | boolMetFin = ~numpy.isnan(SNRm) | |
|
|
1385 | # coh[:-1,:] = numpy.nanmean(numpy.abs(phase[:,1:,:] - phase[:,:-1,:]),axis=0) | |
|
|
1386 | else: | |
|
|
1387 | #------------------------ Meteor mask --------------------------------- | |
|
|
1388 | # #SNR mask | |
|
|
1389 | # boolMet = (SNRdB>SNRthresh)#|(~numpy.isnan(SNRdB)) | |
|
|
1390 | # | |
|
|
1391 | # #Erase small objects | |
|
|
1392 | # boolMet1 = self.__erase_small(boolMet, 2*sec, 5) | |
|
|
1393 | # | |
|
|
1394 | # auxEEJ = numpy.sum(boolMet1,axis=0) | |
|
|
1395 | # indOver = auxEEJ>nProfiles*0.8 #Use this later | |
|
|
1396 | # indEEJ = numpy.where(indOver)[0] | |
|
|
1397 | # indNEEJ = numpy.where(~indOver)[0] | |
|
|
1398 | # | |
|
|
1399 | # boolMetFin = boolMet1 | |
|
|
1400 | # | |
|
|
1401 | # if indEEJ.size > 0: | |
|
|
1402 | # boolMet1[:,indEEJ] = False #Erase heights with EEJ | |
|
|
1403 | # | |
|
|
1404 | # boolMet2 = coh > cohThresh | |
|
|
1405 | # boolMet2 = self.__erase_small(boolMet2, 2*sec,5) | |
|
|
1406 | # | |
|
|
1407 | # #Final Meteor mask | |
|
|
1408 | # boolMetFin = boolMet1|boolMet2 | |
|
|
1409 | ||
|
|
1410 | #Coherence mask | |
|
|
1411 | boolMet1 = coh > 0.75 | |
|
|
1412 | struc = numpy.ones((30,1)) | |
|
|
1413 | boolMet1 = ndimage.morphology.binary_dilation(boolMet1, structure=struc) | |
|
|
1414 | ||
|
|
1415 | #Derivative mask | |
|
|
1416 | derPhase = numpy.nanmean(numpy.abs(phase[:,1:,:] - phase[:,:-1,:]),axis=0) | |
|
|
1417 | boolMet2 = derPhase < 0.2 | |
|
|
1418 | # boolMet2 = ndimage.morphology.binary_opening(boolMet2) | |
|
|
1419 | # boolMet2 = ndimage.morphology.binary_closing(boolMet2, structure = numpy.ones((10,1))) | |
|
|
1420 | boolMet2 = ndimage.median_filter(boolMet2,size=5) | |
|
|
1421 | boolMet2 = numpy.vstack((boolMet2,numpy.full((1,nHeights), True, dtype=bool))) | |
|
|
1422 | # #Final mask | |
|
|
1423 | # boolMetFin = boolMet2 | |
|
|
1424 | boolMetFin = boolMet1&boolMet2 | |
|
|
1425 | # boolMetFin = ndimage.morphology.binary_dilation(boolMetFin) | |
|
|
1426 | #Creating data_param | |
|
|
1427 | coordMet = numpy.where(boolMetFin) | |
|
|
1428 | ||
|
|
1429 | tmet = coordMet[0] | |
|
|
1430 | hmet = coordMet[1] | |
|
|
1431 | ||
|
|
1432 | data_param = numpy.zeros((tmet.size, 6 + nPairs)) | |
|
|
1433 | data_param[:,0] = utctime | |
|
|
1434 | data_param[:,1] = tmet | |
|
|
1435 | data_param[:,2] = hmet | |
|
|
1436 | data_param[:,3] = SNRm[tmet,hmet] | |
|
|
1437 | data_param[:,4] = velRad[tmet,hmet] | |
|
|
1438 | data_param[:,5] = coh[tmet,hmet] | |
|
|
1439 | data_param[:,6:] = phase[:,tmet,hmet].T | |
|
|
1440 | ||
|
|
1441 | elif mode == 'DBS': | |
|
|
1442 | self.dataOut.groupList = numpy.arange(nChannels) | |
|
|
1443 | ||
|
|
1444 | #Radial Velocities | |
|
|
1445 | # phase = numpy.angle(data_acf[:,1,:,:]) | |
|
|
1446 | phase = ndimage.median_filter(numpy.angle(data_acf[:,1,:,:]), size = (1,5,1)) | |
|
|
1447 | velRad = phase*lamb/(4*numpy.pi*tSamp) | |
|
|
1448 | ||
|
|
1449 | #Spectral width | |
|
|
1450 | acf1 = ndimage.median_filter(numpy.abs(data_acf[:,1,:,:]), size = (1,5,1)) | |
|
|
1451 | acf2 = ndimage.median_filter(numpy.abs(data_acf[:,2,:,:]), size = (1,5,1)) | |
|
|
1452 | ||
|
|
1453 | spcWidth = (lamb/(2*numpy.sqrt(6)*numpy.pi*tSamp))*numpy.sqrt(numpy.log(acf1/acf2)) | |
|
|
1454 | # velRad = ndimage.median_filter(velRad, size = (1,5,1)) | |
|
|
1455 | if allData: | |
|
|
1456 | boolMetFin = ~numpy.isnan(SNRdB) | |
|
|
1457 | else: | |
|
|
1458 | #SNR | |
|
|
1459 | boolMet1 = (SNRdB>SNRthresh) #SNR mask | |
|
|
1460 | boolMet1 = ndimage.median_filter(boolMet1, size=(1,5,5)) | |
|
|
1461 | ||
|
|
1462 | #Radial velocity | |
|
|
1463 | boolMet2 = numpy.abs(velRad) < 30 | |
|
|
1464 | boolMet2 = ndimage.median_filter(boolMet2, (1,5,5)) | |
|
|
1465 | ||
|
|
1466 | #Spectral Width | |
|
|
1467 | boolMet3 = spcWidth < 30 | |
|
|
1468 | boolMet3 = ndimage.median_filter(boolMet3, (1,5,5)) | |
|
|
1469 | # boolMetFin = self.__erase_small(boolMet1, 10,5) | |
|
|
1470 | boolMetFin = boolMet1&boolMet2&boolMet3 | |
|
|
1471 | ||
|
|
1472 | #Creating data_param | |
|
|
1473 | coordMet = numpy.where(boolMetFin) | |
|
|
1474 | ||
|
|
1475 | cmet = coordMet[0] | |
|
|
1476 | tmet = coordMet[1] | |
|
|
1477 | hmet = coordMet[2] | |
|
|
1478 | ||
|
|
1479 | data_param = numpy.zeros((tmet.size, 7)) | |
|
|
1480 | data_param[:,0] = utctime | |
|
|
1481 | data_param[:,1] = cmet | |
|
|
1482 | data_param[:,2] = tmet | |
|
|
1483 | data_param[:,3] = hmet | |
|
|
1484 | data_param[:,4] = SNR[cmet,tmet,hmet].T | |
|
|
1485 | data_param[:,5] = velRad[cmet,tmet,hmet].T | |
|
|
1486 | data_param[:,6] = spcWidth[cmet,tmet,hmet].T | |
|
|
1487 | ||
|
|
1488 | # self.dataOut.data_param = data_int | |
|
|
1489 | if len(data_param) == 0: | |
|
|
1490 | self.dataOut.flagNoData = True | |
|
|
1491 | else: | |
|
|
1492 | self.dataOut.data_param = data_param | |
|
|
1493 | ||
|
|
1494 | def __erase_small(self, binArray, threshX, threshY): | |
|
|
1495 | labarray, numfeat = ndimage.measurements.label(binArray) | |
|
|
1496 | binArray1 = numpy.copy(binArray) | |
|
|
1497 | ||
|
|
1498 | for i in range(1,numfeat + 1): | |
|
|
1499 | auxBin = (labarray==i) | |
|
|
1500 | auxSize = auxBin.sum() | |
|
|
1501 | ||
|
|
1502 | x,y = numpy.where(auxBin) | |
|
|
1503 | widthX = x.max() - x.min() | |
|
|
1504 | widthY = y.max() - y.min() | |
|
|
1505 | ||
|
|
1506 | #width X: 3 seg -> 12.5*3 | |
|
|
1507 | #width Y: | |
|
|
1508 | ||
|
|
1509 | if (auxSize < 50) or (widthX < threshX) or (widthY < threshY): | |
|
|
1510 | binArray1[auxBin] = False | |
|
|
1511 | ||
|
|
1512 | return binArray1 | |
|
|
1513 | ||
|
|
1514 | def WeirdEcho(self): | |
|
|
1515 | # data_pre = self.dataOut.data_pre | |
|
|
1516 | # nHeights = self.dataOut.nHeights | |
|
|
1517 | # # nProfiles = self.dataOut.data_pre.shape[1] | |
|
|
1518 | # # data_param = numpy.zeros((len(pairslist),nProfiles,nHeights)) | |
|
|
1519 | # data_param = numpy.zeros((len(pairslist),nHeights)) | |
|
|
1520 | # | |
|
|
1521 | # for i in range(len(pairslist)): | |
|
|
1522 | # chan0 = data_pre[pairslist[i][0],:] | |
|
|
1523 | # chan1 = data_pre[pairslist[i][1],:] | |
|
|
1524 | # #calcular correlacion cruzada | |
|
|
1525 | # #magnitud es coherencia | |
|
|
1526 | # #fase es dif fase | |
|
|
1527 | # correl = chan0*numpy.conj(chan1) | |
|
|
1528 | # coherence = numpy.abs(correl)/(numpy.abs(chan0)*numpy.abs(chan1)) | |
|
|
1529 | # phase = numpy.angle(correl) | |
|
|
1530 | # # data_param[2*i,:,:] = coherence | |
|
|
1531 | # data_param[i,:] = phase | |
|
|
1532 | # | |
|
|
1533 | # self.dataOut.data_param = data_param | |
|
|
1534 | # self.dataOut.groupList = pairslist | |
|
|
1535 | data_cspc = self.dataOut.data_pre[1] | |
|
|
1536 | ccf = numpy.average(data_cspc,axis=1) | |
|
|
1537 | phases = numpy.angle(ccf).T | |
|
|
1538 | ||
|
|
1539 | meteorOps = MeteorOperations() | |
|
|
1540 | pairsList = ((0,1),(2,3)) | |
|
|
1541 | jph = numpy.array([0,0,0,0]) | |
|
|
1542 | AOAthresh = numpy.pi/8 | |
|
|
1543 | azimuth = 45 | |
|
|
1544 | error = numpy.zeros((phases.shape[0],1)) | |
|
|
1545 | AOA,error = meteorOps.getAOA(phases, pairsList, error, AOAthresh, azimuth) | |
|
|
1546 | self.dataOut.data_param = AOA.T | |
|
|
1328 | 1547 | |
|
|
1329 | 1548 | class WindProfiler(Operation): |
|
|
1330 | 1549 | |
|
|
1331 | 1550 | __isConfig = False |
|
|
1332 | 1551 | |
|
|
1333 | 1552 | __initime = None |
|
|
1334 | 1553 | __lastdatatime = None |
|
|
1335 | 1554 | __integrationtime = None |
|
|
1336 | 1555 | |
|
|
1337 | 1556 | __buffer = None |
|
|
1338 | 1557 | |
|
|
1339 | 1558 | __dataReady = False |
|
|
1340 | 1559 | |
|
|
1341 | 1560 | __firstdata = None |
|
|
1342 | 1561 | |
|
|
1343 | 1562 | n = None |
|
|
1344 | 1563 | |
|
|
1345 | 1564 | def __init__(self): |
|
|
1346 | 1565 | Operation.__init__(self) |
|
|
1347 | 1566 | |
|
|
1348 | 1567 | def __calculateCosDir(self, elev, azim): |
|
|
1349 | 1568 | zen = (90 - elev)*numpy.pi/180 |
|
|
1350 | 1569 | azim = azim*numpy.pi/180 |
|
|
1351 | 1570 | cosDirX = numpy.sqrt((1-numpy.cos(zen)**2)/((1+numpy.tan(azim)**2))) |
|
|
1352 | 1571 | cosDirY = numpy.sqrt(1-numpy.cos(zen)**2-cosDirX**2) |
|
|
1353 | 1572 | |
|
|
1354 | 1573 | signX = numpy.sign(numpy.cos(azim)) |
|
|
1355 | 1574 | signY = numpy.sign(numpy.sin(azim)) |
|
|
1356 | 1575 | |
|
|
1357 | 1576 | cosDirX = numpy.copysign(cosDirX, signX) |
|
|
1358 | 1577 | cosDirY = numpy.copysign(cosDirY, signY) |
|
|
1359 | 1578 | return cosDirX, cosDirY |
|
|
1360 | 1579 | |
|
|
1361 | 1580 | def __calculateAngles(self, theta_x, theta_y, azimuth): |
|
|
1362 | 1581 | |
|
|
1363 | 1582 | dir_cosw = numpy.sqrt(1-theta_x**2-theta_y**2) |
|
|
1364 | 1583 | zenith_arr = numpy.arccos(dir_cosw) |
|
|
1365 | 1584 | azimuth_arr = numpy.arctan2(theta_x,theta_y) + azimuth*math.pi/180 |
|
|
1366 | 1585 | |
|
|
1367 | 1586 | dir_cosu = numpy.sin(azimuth_arr)*numpy.sin(zenith_arr) |
|
|
1368 | 1587 | dir_cosv = numpy.cos(azimuth_arr)*numpy.sin(zenith_arr) |
|
|
1369 | 1588 | |
|
|
1370 | 1589 | return azimuth_arr, zenith_arr, dir_cosu, dir_cosv, dir_cosw |
|
|
1371 | 1590 | |
|
|
1372 | 1591 | def __calculateMatA(self, dir_cosu, dir_cosv, dir_cosw, horOnly): |
|
|
1373 | 1592 | |
|
|
1374 | 1593 | # |
|
|
1375 | 1594 | if horOnly: |
|
|
1376 | 1595 | A = numpy.c_[dir_cosu,dir_cosv] |
|
|
1377 | 1596 | else: |
|
|
1378 | 1597 | A = numpy.c_[dir_cosu,dir_cosv,dir_cosw] |
|
|
1379 | 1598 | A = numpy.asmatrix(A) |
|
|
1380 | 1599 | A1 = numpy.linalg.inv(A.transpose()*A)*A.transpose() |
|
|
1381 | 1600 | |
|
|
1382 | 1601 | return A1 |
|
|
1383 | 1602 | |
|
|
1384 | 1603 | def __correctValues(self, heiRang, phi, velRadial, SNR): |
|
|
1385 | 1604 | listPhi = phi.tolist() |
|
|
1386 | 1605 | maxid = listPhi.index(max(listPhi)) |
|
|
1387 | 1606 | minid = listPhi.index(min(listPhi)) |
|
|
1388 | 1607 | |
|
|
1389 | 1608 | rango = range(len(phi)) |
|
|
1390 | 1609 | # rango = numpy.delete(rango,maxid) |
|
|
1391 | 1610 | |
|
|
1392 | 1611 | heiRang1 = heiRang*math.cos(phi[maxid]) |
|
|
1393 | 1612 | heiRangAux = heiRang*math.cos(phi[minid]) |
|
|
1394 | 1613 | indOut = (heiRang1 < heiRangAux[0]).nonzero() |
|
|
1395 | 1614 | heiRang1 = numpy.delete(heiRang1,indOut) |
|
|
1396 | 1615 | |
|
|
1397 | 1616 | velRadial1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
|
1398 | 1617 | SNR1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
|
1399 | 1618 | |
|
|
1400 | 1619 | for i in rango: |
|
|
1401 | 1620 | x = heiRang*math.cos(phi[i]) |
|
|
1402 | 1621 | y1 = velRadial[i,:] |
|
|
1403 | 1622 | f1 = interpolate.interp1d(x,y1,kind = 'cubic') |
|
|
1404 | 1623 | |
|
|
1405 | 1624 | x1 = heiRang1 |
|
|
1406 | 1625 | y11 = f1(x1) |
|
|
1407 | 1626 | |
|
|
1408 | 1627 | y2 = SNR[i,:] |
|
|
1409 | 1628 | f2 = interpolate.interp1d(x,y2,kind = 'cubic') |
|
|
1410 | 1629 | y21 = f2(x1) |
|
|
1411 | 1630 | |
|
|
1412 | 1631 | velRadial1[i,:] = y11 |
|
|
1413 | 1632 | SNR1[i,:] = y21 |
|
|
1414 | 1633 | |
|
|
1415 | 1634 | return heiRang1, velRadial1, SNR1 |
|
|
1416 | 1635 | |
|
|
1417 | 1636 | def __calculateVelUVW(self, A, velRadial): |
|
|
1418 | 1637 | |
|
|
1419 | 1638 | #Operacion Matricial |
|
|
1420 | 1639 | # velUVW = numpy.zeros((velRadial.shape[1],3)) |
|
|
1421 | 1640 | # for ind in range(velRadial.shape[1]): |
|
|
1422 | 1641 | # velUVW[ind,:] = numpy.dot(A,velRadial[:,ind]) |
|
|
1423 | 1642 | # velUVW = velUVW.transpose() |
|
|
1424 | 1643 | velUVW = numpy.zeros((A.shape[0],velRadial.shape[1])) |
|
|
1425 | 1644 | velUVW[:,:] = numpy.dot(A,velRadial) |
|
|
1426 | 1645 | |
|
|
1427 | 1646 | |
|
|
1428 | 1647 | return velUVW |
|
|
1429 | 1648 | |
|
|
1430 | 1649 | def techniqueDBS(self, velRadial0, dirCosx, disrCosy, azimuth, correct, horizontalOnly, heiRang, SNR0): |
|
|
1431 | 1650 | """ |
|
|
1432 | 1651 | Function that implements Doppler Beam Swinging (DBS) technique. |
|
|
1433 | 1652 | |
|
|
1434 | 1653 | Input: Radial velocities, Direction cosines (x and y) of the Beam, Antenna azimuth, |
|
|
1435 | 1654 | Direction correction (if necessary), Ranges and SNR |
|
|
1436 | 1655 | |
|
|
1437 | 1656 | Output: Winds estimation (Zonal, Meridional and Vertical) |
|
|
1438 | 1657 | |
|
|
1439 | 1658 | Parameters affected: Winds, height range, SNR |
|
|
1440 | 1659 | """ |
|
|
1441 | 1660 | azimuth_arr, zenith_arr, dir_cosu, dir_cosv, dir_cosw = self.__calculateAngles(dirCosx, disrCosy, azimuth) |
|
|
1442 | 1661 | heiRang1, velRadial1, SNR1 = self.__correctValues(heiRang, zenith_arr, correct*velRadial0, SNR0) |
|
|
1443 | 1662 | A = self.__calculateMatA(dir_cosu, dir_cosv, dir_cosw, horizontalOnly) |
|
|
1444 | 1663 | |
|
|
1445 | 1664 | #Calculo de Componentes de la velocidad con DBS |
|
|
1446 | 1665 | winds = self.__calculateVelUVW(A,velRadial1) |
|
|
1447 | 1666 | |
|
|
1448 | 1667 | return winds, heiRang1, SNR1 |
|
|
1449 | 1668 | |
|
|
1450 | 1669 | def __calculateDistance(self, posx, posy, pairsCrossCorr, pairsList, pairs, azimuth = None): |
|
|
1451 | 1670 | |
|
|
1452 | 1671 | posx = numpy.asarray(posx) |
|
|
1453 | 1672 | posy = numpy.asarray(posy) |
|
|
1454 | 1673 | |
|
|
1455 | 1674 | #Rotacion Inversa para alinear con el azimuth |
|
|
1456 | 1675 | if azimuth!= None: |
|
|
1457 | 1676 | azimuth = azimuth*math.pi/180 |
|
|
1458 | 1677 | posx1 = posx*math.cos(azimuth) + posy*math.sin(azimuth) |
|
|
1459 | 1678 | posy1 = -posx*math.sin(azimuth) + posy*math.cos(azimuth) |
|
|
1460 | 1679 | else: |
|
|
1461 | 1680 | posx1 = posx |
|
|
1462 | 1681 | posy1 = posy |
|
|
1463 | 1682 | |
|
|
1464 | 1683 | #Calculo de Distancias |
|
|
1465 | 1684 | distx = numpy.zeros(pairsCrossCorr.size) |
|
|
1466 | 1685 | disty = numpy.zeros(pairsCrossCorr.size) |
|
|
1467 | 1686 | dist = numpy.zeros(pairsCrossCorr.size) |
|
|
1468 | 1687 | ang = numpy.zeros(pairsCrossCorr.size) |
|
|
1469 | 1688 | |
|
|
1470 | 1689 | for i in range(pairsCrossCorr.size): |
|
|
1471 | 1690 | distx[i] = posx1[pairsList[pairsCrossCorr[i]][1]] - posx1[pairsList[pairsCrossCorr[i]][0]] |
|
|
1472 | 1691 | disty[i] = posy1[pairsList[pairsCrossCorr[i]][1]] - posy1[pairsList[pairsCrossCorr[i]][0]] |
|
|
1473 | 1692 | dist[i] = numpy.sqrt(distx[i]**2 + disty[i]**2) |
|
|
1474 | 1693 | ang[i] = numpy.arctan2(disty[i],distx[i]) |
|
|
1475 | 1694 | #Calculo de Matrices |
|
|
1476 | 1695 | nPairs = len(pairs) |
|
|
1477 | 1696 | ang1 = numpy.zeros((nPairs, 2, 1)) |
|
|
1478 | 1697 | dist1 = numpy.zeros((nPairs, 2, 1)) |
|
|
1479 | 1698 | |
|
|
1480 | 1699 | for j in range(nPairs): |
|
|
1481 | 1700 | dist1[j,0,0] = dist[pairs[j][0]] |
|
|
1482 | 1701 | dist1[j,1,0] = dist[pairs[j][1]] |
|
|
1483 | 1702 | ang1[j,0,0] = ang[pairs[j][0]] |
|
|
1484 | 1703 | ang1[j,1,0] = ang[pairs[j][1]] |
|
|
1485 | 1704 | |
|
|
1486 | 1705 | return distx,disty, dist1,ang1 |
|
|
1487 | 1706 | |
|
|
1488 | 1707 | def __calculateVelVer(self, phase, lagTRange, _lambda): |
|
|
1489 | 1708 | |
|
|
1490 | 1709 | Ts = lagTRange[1] - lagTRange[0] |
|
|
1491 | 1710 | velW = -_lambda*phase/(4*math.pi*Ts) |
|
|
1492 | 1711 | |
|
|
1493 | 1712 | return velW |
|
|
1494 | 1713 | |
|
|
1495 | 1714 | def __calculateVelHorDir(self, dist, tau1, tau2, ang): |
|
|
1496 | 1715 | nPairs = tau1.shape[0] |
|
|
1497 | 1716 | vel = numpy.zeros((nPairs,3,tau1.shape[2])) |
|
|
1498 | 1717 | |
|
|
1499 | 1718 | angCos = numpy.cos(ang) |
|
|
1500 | 1719 | angSin = numpy.sin(ang) |
|
|
1501 | 1720 | |
|
|
1502 | 1721 | vel0 = dist*tau1/(2*tau2**2) |
|
|
1503 | 1722 | vel[:,0,:] = (vel0*angCos).sum(axis = 1) |
|
|
1504 | 1723 | vel[:,1,:] = (vel0*angSin).sum(axis = 1) |
|
|
1505 | 1724 | |
|
|
1506 | 1725 | ind = numpy.where(numpy.isinf(vel)) |
|
|
1507 | 1726 | vel[ind] = numpy.nan |
|
|
1508 | 1727 | |
|
|
1509 | 1728 | return vel |
|
|
1510 | 1729 | |
|
|
1511 | 1730 | def __getPairsAutoCorr(self, pairsList, nChannels): |
|
|
1512 | 1731 | |
|
|
1513 | 1732 | pairsAutoCorr = numpy.zeros(nChannels, dtype = 'int')*numpy.nan |
|
|
1514 | 1733 | |
|
|
1515 | 1734 | for l in range(len(pairsList)): |
|
|
1516 | 1735 | firstChannel = pairsList[l][0] |
|
|
1517 | 1736 | secondChannel = pairsList[l][1] |
|
|
1518 | 1737 | |
|
|
1519 | 1738 | #Obteniendo pares de Autocorrelacion |
|
|
1520 | 1739 | if firstChannel == secondChannel: |
|
|
1521 | 1740 | pairsAutoCorr[firstChannel] = int(l) |
|
|
1522 | 1741 | |
|
|
1523 | 1742 | pairsAutoCorr = pairsAutoCorr.astype(int) |
|
|
1524 | 1743 | |
|
|
1525 | 1744 | pairsCrossCorr = range(len(pairsList)) |
|
|
1526 | 1745 | pairsCrossCorr = numpy.delete(pairsCrossCorr,pairsAutoCorr) |
|
|
1527 | 1746 | |
|
|
1528 | 1747 | return pairsAutoCorr, pairsCrossCorr |
|
|
1529 | 1748 | |
|
|
1530 | 1749 | def techniqueSA(self, pairsSelected, pairsList, nChannels, tau, azimuth, _lambda, position_x, position_y, lagTRange, correctFactor): |
|
|
1531 | 1750 | """ |
|
|
1532 | 1751 | Function that implements Spaced Antenna (SA) technique. |
|
|
1533 | 1752 | |
|
|
1534 | 1753 | Input: Radial velocities, Direction cosines (x and y) of the Beam, Antenna azimuth, |
|
|
1535 | 1754 | Direction correction (if necessary), Ranges and SNR |
|
|
1536 | 1755 | |
|
|
1537 | 1756 | Output: Winds estimation (Zonal, Meridional and Vertical) |
|
|
1538 | 1757 | |
|
|
1539 | 1758 | Parameters affected: Winds |
|
|
1540 | 1759 | """ |
|
|
1541 | 1760 | #Cross Correlation pairs obtained |
|
|
1542 | 1761 | pairsAutoCorr, pairsCrossCorr = self.__getPairsAutoCorr(pairsList, nChannels) |
|
|
1543 | 1762 | pairsArray = numpy.array(pairsList)[pairsCrossCorr] |
|
|
1544 | 1763 | pairsSelArray = numpy.array(pairsSelected) |
|
|
1545 | 1764 | pairs = [] |
|
|
1546 | 1765 | |
|
|
1547 | 1766 | #Wind estimation pairs obtained |
|
|
1548 | 1767 | for i in range(pairsSelArray.shape[0]/2): |
|
|
1549 | 1768 | ind1 = numpy.where(numpy.all(pairsArray == pairsSelArray[2*i], axis = 1))[0][0] |
|
|
1550 | 1769 | ind2 = numpy.where(numpy.all(pairsArray == pairsSelArray[2*i + 1], axis = 1))[0][0] |
|
|
1551 | 1770 | pairs.append((ind1,ind2)) |
|
|
1552 | 1771 | |
|
|
1553 | 1772 | indtau = tau.shape[0]/2 |
|
|
1554 | 1773 | tau1 = tau[:indtau,:] |
|
|
1555 | 1774 | tau2 = tau[indtau:-1,:] |
|
|
1556 | 1775 | tau1 = tau1[pairs,:] |
|
|
1557 | 1776 | tau2 = tau2[pairs,:] |
|
|
1558 | 1777 | phase1 = tau[-1,:] |
|
|
1559 | 1778 | |
|
|
1560 | 1779 | #--------------------------------------------------------------------- |
|
|
1561 | 1780 | #Metodo Directo |
|
|
1562 | 1781 | distx, disty, dist, ang = self.__calculateDistance(position_x, position_y, pairsCrossCorr, pairsList, pairs,azimuth) |
|
|
1563 | 1782 | winds = self.__calculateVelHorDir(dist, tau1, tau2, ang) |
|
|
1564 | 1783 | winds = stats.nanmean(winds, axis=0) |
|
|
1565 | 1784 | #--------------------------------------------------------------------- |
|
|
1566 | 1785 | #Metodo General |
|
|
1567 | 1786 | # distx, disty, dist = self.calculateDistance(position_x,position_y,pairsCrossCorr, pairsList, azimuth) |
|
|
1568 | 1787 | # #Calculo Coeficientes de Funcion de Correlacion |
|
|
1569 | 1788 | # F,G,A,B,H = self.calculateCoef(tau1,tau2,distx,disty,n) |
|
|
1570 | 1789 | # #Calculo de Velocidades |
|
|
1571 | 1790 | # winds = self.calculateVelUV(F,G,A,B,H) |
|
|
1572 | 1791 | |
|
|
1573 | 1792 | #--------------------------------------------------------------------- |
|
|
1574 | 1793 | winds[2,:] = self.__calculateVelVer(phase1, lagTRange, _lambda) |
|
|
1575 | 1794 | winds = correctFactor*winds |
|
|
1576 | 1795 | return winds |
|
|
1577 | 1796 | |
|
|
1578 | 1797 | def __checkTime(self, currentTime, paramInterval, outputInterval): |
|
|
1579 | 1798 | |
|
|
1580 | 1799 | dataTime = currentTime + paramInterval |
|
|
1581 | 1800 | deltaTime = dataTime - self.__initime |
|
|
1582 | 1801 | |
|
|
1583 | 1802 | if deltaTime >= outputInterval or deltaTime < 0: |
|
|
1584 | 1803 | self.__dataReady = True |
|
|
1585 | 1804 | return |
|
|
1586 | 1805 | |
|
|
1587 | 1806 | def techniqueMeteors(self, arrayMeteor, meteorThresh, heightMin, heightMax): |
|
|
1588 | 1807 | ''' |
|
|
1589 | 1808 | Function that implements winds estimation technique with detected meteors. |
|
|
1590 | 1809 | |
|
|
1591 | 1810 | Input: Detected meteors, Minimum meteor quantity to wind estimation |
|
|
1592 | 1811 | |
|
|
1593 | 1812 | Output: Winds estimation (Zonal and Meridional) |
|
|
1594 | 1813 | |
|
|
1595 | 1814 | Parameters affected: Winds |
|
|
1596 | 1815 | ''' |
|
|
1597 | 1816 | # print arrayMeteor.shape |
|
|
1598 | 1817 | #Settings |
|
|
1599 | 1818 | nInt = (heightMax - heightMin)/2 |
|
|
1600 | 1819 | # print nInt |
|
|
1601 | 1820 | nInt = int(nInt) |
|
|
1602 | 1821 | # print nInt |
|
|
1603 | 1822 | winds = numpy.zeros((2,nInt))*numpy.nan |
|
|
1604 | 1823 | |
|
|
1605 | 1824 | #Filter errors |
|
|
1606 | 1825 | error = numpy.where(arrayMeteor[:,-1] == 0)[0] |
|
|
1607 | 1826 | finalMeteor = arrayMeteor[error,:] |
|
|
1608 | 1827 | |
|
|
1609 | 1828 | #Meteor Histogram |
|
|
1610 | 1829 | finalHeights = finalMeteor[:,2] |
|
|
1611 | 1830 | hist = numpy.histogram(finalHeights, bins = nInt, range = (heightMin,heightMax)) |
|
|
1612 | 1831 | nMeteorsPerI = hist[0] |
|
|
1613 | 1832 | heightPerI = hist[1] |
|
|
1614 | 1833 | |
|
|
1615 | 1834 | #Sort of meteors |
|
|
1616 | 1835 | indSort = finalHeights.argsort() |
|
|
1617 | 1836 | finalMeteor2 = finalMeteor[indSort,:] |
|
|
1618 | 1837 | |
|
|
1619 | 1838 | # Calculating winds |
|
|
1620 | 1839 | ind1 = 0 |
|
|
1621 | 1840 | ind2 = 0 |
|
|
1622 | 1841 | |
|
|
1623 | 1842 | for i in range(nInt): |
|
|
1624 | 1843 | nMet = nMeteorsPerI[i] |
|
|
1625 | 1844 | ind1 = ind2 |
|
|
1626 | 1845 | ind2 = ind1 + nMet |
|
|
1627 | 1846 | |
|
|
1628 | 1847 | meteorAux = finalMeteor2[ind1:ind2,:] |
|
|
1629 | 1848 | |
|
|
1630 | 1849 | if meteorAux.shape[0] >= meteorThresh: |
|
|
1631 | 1850 | vel = meteorAux[:, 6] |
|
|
1632 | 1851 | zen = meteorAux[:, 4]*numpy.pi/180 |
|
|
1633 | 1852 | azim = meteorAux[:, 3]*numpy.pi/180 |
|
|
1634 | 1853 | |
|
|
1635 | 1854 | n = numpy.cos(zen) |
|
|
1636 | 1855 | # m = (1 - n**2)/(1 - numpy.tan(azim)**2) |
|
|
1637 | 1856 | # l = m*numpy.tan(azim) |
|
|
1638 | 1857 | l = numpy.sin(zen)*numpy.sin(azim) |
|
|
1639 | 1858 | m = numpy.sin(zen)*numpy.cos(azim) |
|
|
1640 | 1859 | |
|
|
1641 | 1860 | A = numpy.vstack((l, m)).transpose() |
|
|
1642 | 1861 | A1 = numpy.dot(numpy.linalg.inv( numpy.dot(A.transpose(),A) ),A.transpose()) |
|
|
1643 | 1862 | windsAux = numpy.dot(A1, vel) |
|
|
1644 | 1863 | |
|
|
1645 | 1864 | winds[0,i] = windsAux[0] |
|
|
1646 | 1865 | winds[1,i] = windsAux[1] |
|
|
1647 | 1866 | |
|
|
1648 | 1867 | return winds, heightPerI[:-1] |
|
|
1649 | 1868 | |
|
|
1869 | def techniqueNSM_SA(self, **kwargs): | |
|
|
1870 | metArray = kwargs['metArray'] | |
|
|
1871 | heightList = kwargs['heightList'] | |
|
|
1872 | timeList = kwargs['timeList'] | |
|
|
1873 | ||
|
|
1874 | rx_location = kwargs['rx_location'] | |
|
|
1875 | groupList = kwargs['groupList'] | |
|
|
1876 | azimuth = kwargs['azimuth'] | |
|
|
1877 | dfactor = kwargs['dfactor'] | |
|
|
1878 | k = kwargs['k'] | |
|
|
1879 | ||
|
|
1880 | azimuth1, dist = self.__calculateAzimuth1(rx_location, groupList, azimuth) | |
|
|
1881 | d = dist*dfactor | |
|
|
1882 | #Phase calculation | |
|
|
1883 | metArray1 = self.__getPhaseSlope(metArray, heightList, timeList) | |
|
|
1884 | ||
|
|
1885 | metArray1[:,-2] = metArray1[:,-2]*metArray1[:,2]*1000/(k*d[metArray1[:,1].astype(int)]) #angles into velocities | |
|
|
1886 | ||
|
|
1887 | velEst = numpy.zeros((heightList.size,2))*numpy.nan | |
|
|
1888 | azimuth1 = azimuth1*numpy.pi/180 | |
|
|
1889 | ||
|
|
1890 | for i in range(heightList.size): | |
|
|
1891 | h = heightList[i] | |
|
|
1892 | indH = numpy.where((metArray1[:,2] == h)&(numpy.abs(metArray1[:,-2]) < 100))[0] | |
|
|
1893 | metHeight = metArray1[indH,:] | |
|
|
1894 | if metHeight.shape[0] >= 2: | |
|
|
1895 | velAux = numpy.asmatrix(metHeight[:,-2]).T #Radial Velocities | |
|
|
1896 | iazim = metHeight[:,1].astype(int) | |
|
|
1897 | azimAux = numpy.asmatrix(azimuth1[iazim]).T #Azimuths | |
|
|
1898 | A = numpy.hstack((numpy.cos(azimAux),numpy.sin(azimAux))) | |
|
|
1899 | A = numpy.asmatrix(A) | |
|
|
1900 | A1 = numpy.linalg.pinv(A.transpose()*A)*A.transpose() | |
|
|
1901 | velHor = numpy.dot(A1,velAux) | |
|
|
1902 | ||
|
|
1903 | velEst[i,:] = numpy.squeeze(velHor) | |
|
|
1904 | return velEst | |
|
|
1905 | ||
|
|
1906 | def __getPhaseSlope(self, metArray, heightList, timeList): | |
|
|
1907 | meteorList = [] | |
|
|
1908 | #utctime sec1 height SNR velRad ph0 ph1 ph2 coh0 coh1 coh2 | |
|
|
1909 | #Putting back together the meteor matrix | |
|
|
1910 | utctime = metArray[:,0] | |
|
|
1911 | uniqueTime = numpy.unique(utctime) | |
|
|
1912 | ||
|
|
1913 | phaseDerThresh = 0.5 | |
|
|
1914 | ippSeconds = timeList[1] - timeList[0] | |
|
|
1915 | sec = numpy.where(timeList>1)[0][0] | |
|
|
1916 | nPairs = metArray.shape[1] - 6 | |
|
|
1917 | nHeights = len(heightList) | |
|
|
1918 | ||
|
|
1919 | for t in uniqueTime: | |
|
|
1920 | metArray1 = metArray[utctime==t,:] | |
|
|
1921 | # phaseDerThresh = numpy.pi/4 #reducir Phase thresh | |
|
|
1922 | tmet = metArray1[:,1].astype(int) | |
|
|
1923 | hmet = metArray1[:,2].astype(int) | |
|
|
1924 | ||
|
|
1925 | metPhase = numpy.zeros((nPairs, heightList.size, timeList.size - 1)) | |
|
|
1926 | metPhase[:,:] = numpy.nan | |
|
|
1927 | metPhase[:,hmet,tmet] = metArray1[:,6:].T | |
|
|
1928 | ||
|
|
1929 | #Delete short trails | |
|
|
1930 | metBool = ~numpy.isnan(metPhase[0,:,:]) | |
|
|
1931 | heightVect = numpy.sum(metBool, axis = 1) | |
|
|
1932 | metBool[heightVect<sec,:] = False | |
|
|
1933 | metPhase[:,heightVect<sec,:] = numpy.nan | |
|
|
1934 | ||
|
|
1935 | #Derivative | |
|
|
1936 | metDer = numpy.abs(metPhase[:,:,1:] - metPhase[:,:,:-1]) | |
|
|
1937 | phDerAux = numpy.dstack((numpy.full((nPairs,nHeights,1), False, dtype=bool),metDer > phaseDerThresh)) | |
|
|
1938 | metPhase[phDerAux] = numpy.nan | |
|
|
1939 | ||
|
|
1940 | #--------------------------METEOR DETECTION ----------------------------------------- | |
|
|
1941 | indMet = numpy.where(numpy.any(metBool,axis=1))[0] | |
|
|
1942 | ||
|
|
1943 | for p in numpy.arange(nPairs): | |
|
|
1944 | phase = metPhase[p,:,:] | |
|
|
1945 | phDer = metDer[p,:,:] | |
|
|
1946 | ||
|
|
1947 | for h in indMet: | |
|
|
1948 | height = heightList[h] | |
|
|
1949 | phase1 = phase[h,:] #82 | |
|
|
1950 | phDer1 = phDer[h,:] | |
|
|
1951 | ||
|
|
1952 | phase1[~numpy.isnan(phase1)] = numpy.unwrap(phase1[~numpy.isnan(phase1)]) #Unwrap | |
|
|
1953 | ||
|
|
1954 | indValid = numpy.where(~numpy.isnan(phase1))[0] | |
|
|
1955 | initMet = indValid[0] | |
|
|
1956 | endMet = 0 | |
|
|
1957 | ||
|
|
1958 | for i in range(len(indValid)-1): | |
|
|
1959 | ||
|
|
1960 | #Time difference | |
|
|
1961 | inow = indValid[i] | |
|
|
1962 | inext = indValid[i+1] | |
|
|
1963 | idiff = inext - inow | |
|
|
1964 | #Phase difference | |
|
|
1965 | phDiff = numpy.abs(phase1[inext] - phase1[inow]) | |
|
|
1966 | ||
|
|
1967 | if idiff>sec or phDiff>numpy.pi/4 or inext==indValid[-1]: #End of Meteor | |
|
|
1968 | sizeTrail = inow - initMet + 1 | |
|
|
1969 | if sizeTrail>3*sec: #Too short meteors | |
|
|
1970 | x = numpy.arange(initMet,inow+1)*ippSeconds | |
|
|
1971 | y = phase1[initMet:inow+1] | |
|
|
1972 | ynnan = ~numpy.isnan(y) | |
|
|
1973 | x = x[ynnan] | |
|
|
1974 | y = y[ynnan] | |
|
|
1975 | slope, intercept, r_value, p_value, std_err = stats.linregress(x,y) | |
|
|
1976 | ylin = x*slope + intercept | |
|
|
1977 | rsq = r_value**2 | |
|
|
1978 | if rsq > 0.5: | |
|
|
1979 | vel = slope#*height*1000/(k*d) | |
|
|
1980 | estAux = numpy.array([utctime,p,height, vel, rsq]) | |
|
|
1981 | meteorList.append(estAux) | |
|
|
1982 | initMet = inext | |
|
|
1983 | metArray2 = numpy.array(meteorList) | |
|
|
1984 | ||
|
|
1985 | return metArray2 | |
|
|
1986 | ||
|
|
1987 | def __calculateAzimuth1(self, rx_location, pairslist, azimuth0): | |
|
|
1988 | ||
|
|
1989 | azimuth1 = numpy.zeros(len(pairslist)) | |
|
|
1990 | dist = numpy.zeros(len(pairslist)) | |
|
|
1991 | ||
|
|
1992 | for i in range(len(rx_location)): | |
|
|
1993 | ch0 = pairslist[i][0] | |
|
|
1994 | ch1 = pairslist[i][1] | |
|
|
1995 | ||
|
|
1996 | diffX = rx_location[ch0][0] - rx_location[ch1][0] | |
|
|
1997 | diffY = rx_location[ch0][1] - rx_location[ch1][1] | |
|
|
1998 | azimuth1[i] = numpy.arctan2(diffY,diffX)*180/numpy.pi | |
|
|
1999 | dist[i] = numpy.sqrt(diffX**2 + diffY**2) | |
|
|
2000 | ||
|
|
2001 | azimuth1 -= azimuth0 | |
|
|
2002 | return azimuth1, dist | |
|
|
2003 | ||
|
|
2004 | def techniqueNSM_DBS(self, **kwargs): | |
|
|
2005 | metArray = kwargs['metArray'] | |
|
|
2006 | heightList = kwargs['heightList'] | |
|
|
2007 | timeList = kwargs['timeList'] | |
|
|
2008 | zenithList = kwargs['zenithList'] | |
|
|
2009 | nChan = numpy.max(cmet) + 1 | |
|
|
2010 | nHeights = len(heightList) | |
|
|
2011 | ||
|
|
2012 | utctime = metArray[:,0] | |
|
|
2013 | cmet = metArray[:,1] | |
|
|
2014 | hmet = metArray1[:,3].astype(int) | |
|
|
2015 | h1met = heightList[hmet]*zenithList[cmet] | |
|
|
2016 | vmet = metArray1[:,5] | |
|
|
2017 | ||
|
|
2018 | for i in range(nHeights - 1): | |
|
|
2019 | hmin = heightList[i] | |
|
|
2020 | hmax = heightList[i + 1] | |
|
|
2021 | ||
|
|
2022 | vthisH = vmet[(h1met>=hmin) & (h1met<hmax)] | |
|
|
2023 | ||
|
|
2024 | ||
|
|
2025 | ||
|
|
2026 | return data_output | |
|
|
2027 | ||
|
|
1650 | 2028 | def run(self, dataOut, technique, **kwargs): |
|
|
1651 | 2029 | |
|
|
1652 | 2030 | param = dataOut.data_param |
|
|
1653 | 2031 | if dataOut.abscissaList != None: |
|
|
1654 | 2032 | absc = dataOut.abscissaList[:-1] |
|
|
1655 | 2033 | noise = dataOut.noise |
|
|
1656 | 2034 | heightList = dataOut.heightList |
|
|
1657 | 2035 | SNR = dataOut.data_SNR |
|
|
1658 | 2036 | |
|
|
1659 | 2037 | if technique == 'DBS': |
|
|
1660 | 2038 | |
|
|
1661 | 2039 | if kwargs.has_key('dirCosx') and kwargs.has_key('dirCosy'): |
|
|
1662 | 2040 | theta_x = numpy.array(kwargs['dirCosx']) |
|
|
1663 | 2041 | theta_y = numpy.array(kwargs['dirCosy']) |
|
|
1664 | 2042 | else: |
|
|
1665 | 2043 | elev = numpy.array(kwargs['elevation']) |
|
|
1666 | 2044 | azim = numpy.array(kwargs['azimuth']) |
|
|
1667 | 2045 | theta_x, theta_y = self.__calculateCosDir(elev, azim) |
|
|
1668 | 2046 | azimuth = kwargs['correctAzimuth'] |
|
|
1669 | 2047 | if kwargs.has_key('horizontalOnly'): |
|
|
1670 | 2048 | horizontalOnly = kwargs['horizontalOnly'] |
|
|
1671 | 2049 | else: horizontalOnly = False |
|
|
1672 | 2050 | if kwargs.has_key('correctFactor'): |
|
|
1673 | 2051 | correctFactor = kwargs['correctFactor'] |
|
|
1674 | 2052 | else: correctFactor = 1 |
|
|
1675 | 2053 | if kwargs.has_key('channelList'): |
|
|
1676 | 2054 | channelList = kwargs['channelList'] |
|
|
1677 | 2055 | if len(channelList) == 2: |
|
|
1678 | 2056 | horizontalOnly = True |
|
|
1679 | 2057 | arrayChannel = numpy.array(channelList) |
|
|
1680 | 2058 | param = param[arrayChannel,:,:] |
|
|
1681 | 2059 | theta_x = theta_x[arrayChannel] |
|
|
1682 | 2060 | theta_y = theta_y[arrayChannel] |
|
|
1683 | 2061 | |
|
|
1684 | 2062 | velRadial0 = param[:,1,:] #Radial velocity |
|
|
1685 | 2063 | dataOut.data_output, dataOut.heightList, dataOut.data_SNR = self.techniqueDBS(velRadial0, theta_x, theta_y, azimuth, correctFactor, horizontalOnly, heightList, SNR) #DBS Function |
|
|
1686 | 2064 | dataOut.utctimeInit = dataOut.utctime |
|
|
1687 |
dataOut.outputInterval = dataOut. |
|
|
|
2065 | dataOut.outputInterval = dataOut.paramInterval | |
|
|
1688 | 2066 | |
|
|
1689 | 2067 | elif technique == 'SA': |
|
|
1690 | 2068 | |
|
|
1691 | 2069 | #Parameters |
|
|
1692 | 2070 | position_x = kwargs['positionX'] |
|
|
1693 | 2071 | position_y = kwargs['positionY'] |
|
|
1694 | 2072 | azimuth = kwargs['azimuth'] |
|
|
1695 | 2073 | |
|
|
1696 | 2074 | if kwargs.has_key('crosspairsList'): |
|
|
1697 | 2075 | pairs = kwargs['crosspairsList'] |
|
|
1698 | 2076 | else: |
|
|
1699 | 2077 | pairs = None |
|
|
1700 | 2078 | |
|
|
1701 | 2079 | if kwargs.has_key('correctFactor'): |
|
|
1702 | 2080 | correctFactor = kwargs['correctFactor'] |
|
|
1703 | 2081 | else: |
|
|
1704 | 2082 | correctFactor = 1 |
|
|
1705 | 2083 | |
|
|
1706 | 2084 | tau = dataOut.data_param |
|
|
1707 | 2085 | _lambda = dataOut.C/dataOut.frequency |
|
|
1708 | 2086 | pairsList = dataOut.groupList |
|
|
1709 | 2087 | nChannels = dataOut.nChannels |
|
|
1710 | 2088 | |
|
|
1711 | 2089 | dataOut.data_output = self.techniqueSA(pairs, pairsList, nChannels, tau, azimuth, _lambda, position_x, position_y, absc, correctFactor) |
|
|
1712 | 2090 | dataOut.utctimeInit = dataOut.utctime |
|
|
1713 | 2091 | dataOut.outputInterval = dataOut.timeInterval |
|
|
1714 | 2092 | |
|
|
1715 | 2093 | elif technique == 'Meteors': |
|
|
1716 | 2094 | dataOut.flagNoData = True |
|
|
1717 | 2095 | self.__dataReady = False |
|
|
1718 | 2096 | |
|
|
1719 | 2097 | if kwargs.has_key('nHours'): |
|
|
1720 | 2098 | nHours = kwargs['nHours'] |
|
|
1721 | 2099 | else: |
|
|
1722 | 2100 | nHours = 1 |
|
|
1723 | 2101 | |
|
|
1724 | 2102 | if kwargs.has_key('meteorsPerBin'): |
|
|
1725 | 2103 | meteorThresh = kwargs['meteorsPerBin'] |
|
|
1726 | 2104 | else: |
|
|
1727 | 2105 | meteorThresh = 6 |
|
|
1728 | 2106 | |
|
|
1729 | 2107 | if kwargs.has_key('hmin'): |
|
|
1730 | 2108 | hmin = kwargs['hmin'] |
|
|
1731 | 2109 | else: hmin = 70 |
|
|
1732 | 2110 | if kwargs.has_key('hmax'): |
|
|
1733 | 2111 | hmax = kwargs['hmax'] |
|
|
1734 | 2112 | else: hmax = 110 |
|
|
1735 | 2113 | |
|
|
1736 | 2114 | dataOut.outputInterval = nHours*3600 |
|
|
1737 | 2115 | |
|
|
1738 | 2116 | if self.__isConfig == False: |
|
|
1739 | 2117 | # self.__initime = dataOut.datatime.replace(minute = 0, second = 0, microsecond = 03) |
|
|
1740 | 2118 | #Get Initial LTC time |
|
|
1741 | 2119 | self.__initime = datetime.datetime.utcfromtimestamp(dataOut.utctime) |
|
|
1742 | 2120 | self.__initime = (self.__initime.replace(minute = 0, second = 0, microsecond = 0) - datetime.datetime(1970, 1, 1)).total_seconds() |
|
|
1743 | 2121 | |
|
|
1744 | 2122 | self.__isConfig = True |
|
|
1745 | 2123 | |
|
|
1746 | 2124 | if self.__buffer == None: |
|
|
1747 | 2125 | self.__buffer = dataOut.data_param |
|
|
1748 | 2126 | self.__firstdata = copy.copy(dataOut) |
|
|
1749 | 2127 | |
|
|
1750 | 2128 | else: |
|
|
1751 | 2129 | self.__buffer = numpy.vstack((self.__buffer, dataOut.data_param)) |
|
|
1752 | 2130 | |
|
|
1753 | 2131 | self.__checkTime(dataOut.utctime, dataOut.paramInterval, dataOut.outputInterval) #Check if the buffer is ready |
|
|
1754 | 2132 | |
|
|
1755 | 2133 | if self.__dataReady: |
|
|
1756 | 2134 | dataOut.utctimeInit = self.__initime |
|
|
1757 | 2135 | |
|
|
1758 | 2136 | self.__initime += dataOut.outputInterval #to erase time offset |
|
|
1759 | 2137 | |
|
|
1760 | 2138 | dataOut.data_output, dataOut.heightList = self.techniqueMeteors(self.__buffer, meteorThresh, hmin, hmax) |
|
|
1761 | 2139 | dataOut.flagNoData = False |
|
|
1762 | 2140 | self.__buffer = None |
|
|
1763 | 2141 | |
|
|
2142 | elif technique == 'Meteors1': | |
|
|
2143 | dataOut.flagNoData = True | |
|
|
2144 | self.__dataReady = False | |
|
|
2145 | ||
|
|
2146 | if kwargs.has_key('nMins'): | |
|
|
2147 | nMins = kwargs['nMins'] | |
|
|
2148 | else: nMins = 20 | |
|
|
2149 | if kwargs.has_key('rx_location'): | |
|
|
2150 | rx_location = kwargs['rx_location'] | |
|
|
2151 | else: rx_location = [(0,1),(1,1),(1,0)] | |
|
|
2152 | if kwargs.has_key('azimuth'): | |
|
|
2153 | azimuth = kwargs['azimuth'] | |
|
|
2154 | else: azimuth = 51 | |
|
|
2155 | if kwargs.has_key('dfactor'): | |
|
|
2156 | dfactor = kwargs['dfactor'] | |
|
|
2157 | if kwargs.has_key('mode'): | |
|
|
2158 | mode = kwargs['mode'] | |
|
|
2159 | else: mode = 'SA' | |
|
|
2160 | ||
|
|
2161 | #Borrar luego esto | |
|
|
2162 | if dataOut.groupList == None: | |
|
|
2163 | dataOut.groupList = [(0,1),(0,2),(1,2)] | |
|
|
2164 | groupList = dataOut.groupList | |
|
|
2165 | C = 3e8 | |
|
|
2166 | freq = 50e6 | |
|
|
2167 | lamb = C/freq | |
|
|
2168 | k = 2*numpy.pi/lamb | |
|
|
2169 | ||
|
|
2170 | timeList = dataOut.abscissaList | |
|
|
2171 | heightList = dataOut.heightList | |
|
|
2172 | ||
|
|
2173 | if self.__isConfig == False: | |
|
|
2174 | dataOut.outputInterval = nMins*60 | |
|
|
2175 | # self.__initime = dataOut.datatime.replace(minute = 0, second = 0, microsecond = 03) | |
|
|
2176 | #Get Initial LTC time | |
|
|
2177 | initime = datetime.datetime.utcfromtimestamp(dataOut.utctime) | |
|
|
2178 | minuteAux = initime.minute | |
|
|
2179 | minuteNew = int(numpy.floor(minuteAux/nMins)*nMins) | |
|
|
2180 | self.__initime = (initime.replace(minute = minuteNew, second = 0, microsecond = 0) - datetime.datetime(1970, 1, 1)).total_seconds() | |
|
|
2181 | ||
|
|
2182 | self.__isConfig = True | |
|
|
2183 | ||
|
|
2184 | if self.__buffer == None: | |
|
|
2185 | self.__buffer = dataOut.data_param | |
|
|
2186 | self.__firstdata = copy.copy(dataOut) | |
|
|
2187 | ||
|
|
2188 | else: | |
|
|
2189 | self.__buffer = numpy.vstack((self.__buffer, dataOut.data_param)) | |
|
|
2190 | ||
|
|
2191 | self.__checkTime(dataOut.utctime, dataOut.paramInterval, dataOut.outputInterval) #Check if the buffer is ready | |
|
|
2192 | ||
|
|
2193 | if self.__dataReady: | |
|
|
2194 | dataOut.utctimeInit = self.__initime | |
|
|
2195 | self.__initime += dataOut.outputInterval #to erase time offset | |
|
|
2196 | ||
|
|
2197 | metArray = self.__buffer | |
|
|
2198 | if mode == 'SA': | |
|
|
2199 | dataOut.data_output = self.techniqueNSM_SA(rx_location=rx_location, groupList=groupList, azimuth=azimuth, dfactor=dfactor, k=k,metArray=metArray, heightList=heightList,timeList=timeList) | |
|
|
2200 | elif mode == 'DBS': | |
|
|
2201 | dataOut.data_output = self.techniqueNSM_DBS(metArray=metArray,heightList=heightList,timeList=timeList) | |
|
|
2202 | dataOut.data_output = dataOut.data_output.T | |
|
|
2203 | dataOut.flagNoData = False | |
|
|
2204 | self.__buffer = None | |
|
|
2205 | ||
|
|
1764 | 2206 | return |
|
|
1765 | 2207 | |
|
|
1766 | 2208 | class EWDriftsEstimation(Operation): |
|
|
1767 | 2209 | |
|
|
1768 | ||
|
|
1769 | 2210 | def __init__(self): |
|
|
1770 | 2211 | Operation.__init__(self) |
|
|
1771 | 2212 | |
|
|
1772 | 2213 | def __correctValues(self, heiRang, phi, velRadial, SNR): |
|
|
1773 | 2214 | listPhi = phi.tolist() |
|
|
1774 | 2215 | maxid = listPhi.index(max(listPhi)) |
|
|
1775 | 2216 | minid = listPhi.index(min(listPhi)) |
|
|
1776 | 2217 | |
|
|
1777 | 2218 | rango = range(len(phi)) |
|
|
1778 | 2219 | # rango = numpy.delete(rango,maxid) |
|
|
1779 | 2220 | |
|
|
1780 | 2221 | heiRang1 = heiRang*math.cos(phi[maxid]) |
|
|
1781 | 2222 | heiRangAux = heiRang*math.cos(phi[minid]) |
|
|
1782 | 2223 | indOut = (heiRang1 < heiRangAux[0]).nonzero() |
|
|
1783 | 2224 | heiRang1 = numpy.delete(heiRang1,indOut) |
|
|
1784 | 2225 | |
|
|
1785 | 2226 | velRadial1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
|
1786 | 2227 | SNR1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
|
1787 | 2228 | |
|
|
1788 | 2229 | for i in rango: |
|
|
1789 | 2230 | x = heiRang*math.cos(phi[i]) |
|
|
1790 | 2231 | y1 = velRadial[i,:] |
|
|
1791 | 2232 | f1 = interpolate.interp1d(x,y1,kind = 'cubic') |
|
|
1792 | 2233 | |
|
|
1793 | 2234 | x1 = heiRang1 |
|
|
1794 | 2235 | y11 = f1(x1) |
|
|
1795 | 2236 | |
|
|
1796 | 2237 | y2 = SNR[i,:] |
|
|
1797 | 2238 | f2 = interpolate.interp1d(x,y2,kind = 'cubic') |
|
|
1798 | 2239 | y21 = f2(x1) |
|
|
1799 | 2240 | |
|
|
1800 | 2241 | velRadial1[i,:] = y11 |
|
|
1801 | 2242 | SNR1[i,:] = y21 |
|
|
1802 | 2243 | |
|
|
1803 | 2244 | return heiRang1, velRadial1, SNR1 |
|
|
1804 | 2245 | |
|
|
1805 | 2246 | def run(self, dataOut, zenith, zenithCorrection): |
|
|
1806 | 2247 | heiRang = dataOut.heightList |
|
|
1807 | 2248 | velRadial = dataOut.data_param[:,3,:] |
|
|
1808 | 2249 | SNR = dataOut.data_SNR |
|
|
1809 | 2250 | |
|
|
1810 | 2251 | zenith = numpy.array(zenith) |
|
|
1811 | 2252 | zenith -= zenithCorrection |
|
|
1812 | 2253 | zenith *= numpy.pi/180 |
|
|
1813 | 2254 | |
|
|
1814 | 2255 | heiRang1, velRadial1, SNR1 = self.__correctValues(heiRang, numpy.abs(zenith), velRadial, SNR) |
|
|
1815 | 2256 | |
|
|
1816 | 2257 | alp = zenith[0] |
|
|
1817 | 2258 | bet = zenith[1] |
|
|
1818 | 2259 | |
|
|
1819 | 2260 | w_w = velRadial1[0,:] |
|
|
1820 | 2261 | w_e = velRadial1[1,:] |
|
|
1821 | 2262 | |
|
|
1822 | 2263 | w = (w_w*numpy.sin(bet) - w_e*numpy.sin(alp))/(numpy.cos(alp)*numpy.sin(bet) - numpy.cos(bet)*numpy.sin(alp)) |
|
|
1823 | 2264 | u = (w_w*numpy.cos(bet) - w_e*numpy.cos(alp))/(numpy.sin(alp)*numpy.cos(bet) - numpy.sin(bet)*numpy.cos(alp)) |
|
|
1824 | 2265 | |
|
|
1825 | 2266 | winds = numpy.vstack((u,w)) |
|
|
1826 | 2267 | |
|
|
1827 | 2268 | dataOut.heightList = heiRang1 |
|
|
1828 | 2269 | dataOut.data_output = winds |
|
|
1829 | 2270 | dataOut.data_SNR = SNR1 |
|
|
1830 | 2271 | |
|
|
1831 | 2272 | dataOut.utctimeInit = dataOut.utctime |
|
|
1832 | 2273 | dataOut.outputInterval = dataOut.timeInterval |
|
|
1833 | 2274 | return |
|
|
1834 | 2275 | |
|
|
1835 | 2276 | class PhaseCalibration(Operation): |
|
|
1836 | 2277 | |
|
|
1837 | 2278 | __buffer = None |
|
|
1838 | 2279 | |
|
|
1839 | 2280 | __initime = None |
|
|
1840 | 2281 | |
|
|
1841 | 2282 | __dataReady = False |
|
|
1842 | 2283 | |
|
|
1843 | 2284 | __isConfig = False |
|
|
1844 | 2285 | |
|
|
1845 | 2286 | def __checkTime(self, currentTime, initTime, paramInterval, outputInterval): |
|
|
1846 | 2287 | |
|
|
1847 | 2288 | dataTime = currentTime + paramInterval |
|
|
1848 | 2289 | deltaTime = dataTime - initTime |
|
|
1849 | 2290 | |
|
|
1850 | 2291 | if deltaTime >= outputInterval or deltaTime < 0: |
|
|
1851 | 2292 | return True |
|
|
1852 | 2293 | |
|
|
1853 | 2294 | return False |
|
|
1854 | 2295 | |
|
|
1855 | 2296 | def __getGammas(self, pairs, k, d, phases): |
|
|
1856 | 2297 | gammas = numpy.zeros(2) |
|
|
1857 | 2298 | |
|
|
1858 | 2299 | for i in range(len(pairs)): |
|
|
1859 | 2300 | |
|
|
1860 | 2301 | pairi = pairs[i] |
|
|
1861 | 2302 | |
|
|
1862 | 2303 | #Calculating gamma |
|
|
1863 | 2304 | jdcos = phases[:,pairi[1]]/(k*d[pairi[1]]) |
|
|
1864 | 2305 | jgamma = numpy.angle(numpy.exp(1j*(k*d[pairi[0]]*jdcos - phases[:,pairi[0]]))) |
|
|
1865 | 2306 | |
|
|
1866 | 2307 | #Revised distribution |
|
|
1867 | 2308 | jgammaArray = numpy.hstack((jgamma,jgamma+0.5*numpy.pi,jgamma-0.5*numpy.pi)) |
|
|
1868 | 2309 | |
|
|
1869 | 2310 | #Histogram |
|
|
1870 | 2311 | nBins = 64.0 |
|
|
1871 | 2312 | rmin = -0.5*numpy.pi |
|
|
1872 | 2313 | rmax = 0.5*numpy.pi |
|
|
1873 | 2314 | phaseHisto = numpy.histogram(jgammaArray, bins=nBins, range=(rmin,rmax)) |
|
|
1874 | 2315 | |
|
|
1875 | 2316 | meteorsY = phaseHisto[0] |
|
|
1876 | 2317 | phasesX = phaseHisto[1][:-1] |
|
|
1877 | 2318 | width = phasesX[1] - phasesX[0] |
|
|
1878 | 2319 | phasesX += width/2 |
|
|
1879 | 2320 | |
|
|
1880 | 2321 | #Gaussian aproximation |
|
|
1881 | 2322 | bpeak = meteorsY.argmax() |
|
|
1882 | 2323 | peak = meteorsY.max() |
|
|
1883 | 2324 | jmin = bpeak - 5 |
|
|
1884 | 2325 | jmax = bpeak + 5 + 1 |
|
|
1885 | 2326 | |
|
|
1886 | 2327 | if jmin<0: |
|
|
1887 | 2328 | jmin = 0 |
|
|
1888 | 2329 | jmax = 6 |
|
|
1889 | 2330 | elif jmax > meteorsY.size: |
|
|
1890 | 2331 | jmin = meteorsY.size - 6 |
|
|
1891 | 2332 | jmax = meteorsY.size |
|
|
1892 | 2333 | |
|
|
1893 | 2334 | x0 = numpy.array([peak,bpeak,50]) |
|
|
1894 | 2335 | coeff = optimize.leastsq(self.__residualFunction, x0, args=(meteorsY[jmin:jmax], phasesX[jmin:jmax])) |
|
|
1895 | 2336 | |
|
|
1896 | 2337 | #Gammas |
|
|
1897 | 2338 | gammas[i] = coeff[0][1] |
|
|
1898 | 2339 | |
|
|
1899 | 2340 | return gammas |
|
|
1900 | 2341 | |
|
|
1901 | 2342 | def __residualFunction(self, coeffs, y, t): |
|
|
1902 | 2343 | |
|
|
1903 | 2344 | return y - self.__gauss_function(t, coeffs) |
|
|
1904 | 2345 | |
|
|
1905 | 2346 | def __gauss_function(self, t, coeffs): |
|
|
1906 | 2347 | |
|
|
1907 | 2348 | return coeffs[0]*numpy.exp(-0.5*((t - coeffs[1]) / coeffs[2])**2) |
|
|
1908 | 2349 | |
|
|
1909 | 2350 | def __getPhases(self, azimuth, h, pairsList, d, gammas, meteorsArray): |
|
|
1910 | 2351 | meteorOps = MeteorOperations() |
|
|
1911 | 2352 | nchan = 4 |
|
|
1912 | 2353 | pairx = pairsList[0] |
|
|
1913 | 2354 | pairy = pairsList[1] |
|
|
1914 | 2355 | center_xangle = 0 |
|
|
1915 | 2356 | center_yangle = 0 |
|
|
1916 | 2357 | range_angle = numpy.array([10*numpy.pi,numpy.pi,numpy.pi/2,numpy.pi/4]) |
|
|
1917 | 2358 | ntimes = len(range_angle) |
|
|
1918 | 2359 | |
|
|
1919 | 2360 | nstepsx = 20.0 |
|
|
1920 | 2361 | nstepsy = 20.0 |
|
|
1921 | 2362 | |
|
|
1922 | 2363 | for iz in range(ntimes): |
|
|
1923 | 2364 | min_xangle = -range_angle[iz]/2 + center_xangle |
|
|
1924 | 2365 | max_xangle = range_angle[iz]/2 + center_xangle |
|
|
1925 | 2366 | min_yangle = -range_angle[iz]/2 + center_yangle |
|
|
1926 | 2367 | max_yangle = range_angle[iz]/2 + center_yangle |
|
|
1927 | 2368 | |
|
|
1928 | 2369 | inc_x = (max_xangle-min_xangle)/nstepsx |
|
|
1929 | 2370 | inc_y = (max_yangle-min_yangle)/nstepsy |
|
|
1930 | 2371 | |
|
|
1931 | 2372 | alpha_y = numpy.arange(nstepsy)*inc_y + min_yangle |
|
|
1932 | 2373 | alpha_x = numpy.arange(nstepsx)*inc_x + min_xangle |
|
|
1933 | 2374 | penalty = numpy.zeros((nstepsx,nstepsy)) |
|
|
1934 | 2375 | jph_array = numpy.zeros((nchan,nstepsx,nstepsy)) |
|
|
1935 | 2376 | jph = numpy.zeros(nchan) |
|
|
1936 | 2377 | |
|
|
1937 | 2378 | # Iterations looking for the offset |
|
|
1938 | 2379 | for iy in range(int(nstepsy)): |
|
|
1939 | 2380 | for ix in range(int(nstepsx)): |
|
|
1940 | 2381 | jph[pairy[1]] = alpha_y[iy] |
|
|
1941 | 2382 | jph[pairy[0]] = -gammas[1] + alpha_y[iy]*d[pairy[0]]/d[pairy[1]] |
|
|
1942 | 2383 | |
|
|
1943 | 2384 | jph[pairx[1]] = alpha_x[ix] |
|
|
1944 | 2385 | jph[pairx[0]] = -gammas[0] + alpha_x[ix]*d[pairx[0]]/d[pairx[1]] |
|
|
1945 | 2386 | |
|
|
1946 | 2387 | jph_array[:,ix,iy] = jph |
|
|
1947 | 2388 | |
|
|
1948 | 2389 | meteorsArray1 = meteorOps.getMeteorParams(meteorsArray, azimuth, h, pairsList, jph) |
|
|
1949 | 2390 | error = meteorsArray1[:,-1] |
|
|
1950 | 2391 | ind1 = numpy.where(error==0)[0] |
|
|
1951 | 2392 | penalty[ix,iy] = ind1.size |
|
|
1952 | 2393 | |
|
|
1953 | 2394 | i,j = numpy.unravel_index(penalty.argmax(), penalty.shape) |
|
|
1954 | 2395 | phOffset = jph_array[:,i,j] |
|
|
1955 | 2396 | |
|
|
1956 | 2397 | center_xangle = phOffset[pairx[1]] |
|
|
1957 | 2398 | center_yangle = phOffset[pairy[1]] |
|
|
1958 | 2399 | |
|
|
1959 | 2400 | phOffset = numpy.angle(numpy.exp(1j*jph_array[:,i,j])) |
|
|
1960 | 2401 | phOffset = phOffset*180/numpy.pi |
|
|
1961 | 2402 | return phOffset |
|
|
1962 | 2403 | |
|
|
1963 | 2404 | |
|
|
1964 | 2405 | def run(self, dataOut, hmin, hmax, direction25X=-1, direction20X=1, direction25Y=-1, direction20Y=1, nHours = 1): |
|
|
1965 | 2406 | |
|
|
1966 | 2407 | dataOut.flagNoData = True |
|
|
1967 | 2408 | self.__dataReady = False |
|
|
1968 | 2409 | dataOut.outputInterval = nHours*3600 |
|
|
1969 | 2410 | |
|
|
1970 | 2411 | if self.__isConfig == False: |
|
|
1971 | 2412 | # self.__initime = dataOut.datatime.replace(minute = 0, second = 0, microsecond = 03) |
|
|
1972 | 2413 | #Get Initial LTC time |
|
|
1973 | 2414 | self.__initime = datetime.datetime.utcfromtimestamp(dataOut.utctime) |
|
|
1974 | 2415 | self.__initime = (self.__initime.replace(minute = 0, second = 0, microsecond = 0) - datetime.datetime(1970, 1, 1)).total_seconds() |
|
|
1975 | 2416 | |
|
|
1976 | 2417 | self.__isConfig = True |
|
|
1977 | 2418 | |
|
|
1978 | 2419 | if self.__buffer == None: |
|
|
1979 | 2420 | self.__buffer = dataOut.data_param.copy() |
|
|
1980 | 2421 | |
|
|
1981 | 2422 | else: |
|
|
1982 | 2423 | self.__buffer = numpy.vstack((self.__buffer, dataOut.data_param)) |
|
|
1983 | 2424 | |
|
|
1984 | 2425 | self.__dataReady = self.__checkTime(dataOut.utctime, self.__initime, dataOut.paramInterval, dataOut.outputInterval) #Check if the buffer is ready |
|
|
1985 | 2426 | |
|
|
1986 | 2427 | if self.__dataReady: |
|
|
1987 | 2428 | dataOut.utctimeInit = self.__initime |
|
|
1988 | 2429 | self.__initime += dataOut.outputInterval #to erase time offset |
|
|
1989 | 2430 | |
|
|
1990 | 2431 | freq = dataOut.frequency |
|
|
1991 | 2432 | c = dataOut.C #m/s |
|
|
1992 | 2433 | lamb = c/freq |
|
|
1993 | 2434 | k = 2*numpy.pi/lamb |
|
|
1994 | 2435 | azimuth = 0 |
|
|
1995 | 2436 | h = (hmin, hmax) |
|
|
1996 | 2437 | pairs = ((0,1),(2,3)) |
|
|
1997 | 2438 | distances = [direction25X*2.5*lamb, direction20X*2*lamb, direction25Y*2.5*lamb, direction20Y*2*lamb] |
|
|
1998 | 2439 | |
|
|
1999 | 2440 | meteorsArray = self.__buffer |
|
|
2000 | 2441 | error = meteorsArray[:,-1] |
|
|
2001 | 2442 | boolError = (error==0)|(error==3)|(error==4)|(error==13)|(error==14) |
|
|
2002 | 2443 | ind1 = numpy.where(boolError)[0] |
|
|
2003 | 2444 | meteorsArray = meteorsArray[ind1,:] |
|
|
2004 | 2445 | meteorsArray[:,-1] = 0 |
|
|
2005 | 2446 | phases = meteorsArray[:,8:12] |
|
|
2006 | 2447 | |
|
|
2007 | 2448 | #Calculate Gammas |
|
|
2008 | 2449 | gammas = self.__getGammas(pairs, k, distances, phases) |
|
|
2009 | 2450 | # gammas = numpy.array([-21.70409463,45.76935864])*numpy.pi/180 |
|
|
2010 | 2451 | #Calculate Phases |
|
|
2011 | 2452 | phasesOff = self.__getPhases(azimuth, h, pairs, distances, gammas, meteorsArray) |
|
|
2012 | 2453 | phasesOff = phasesOff.reshape((1,phasesOff.size)) |
|
|
2013 | 2454 | dataOut.data_output = -phasesOff |
|
|
2014 | 2455 | dataOut.flagNoData = False |
|
|
2015 | 2456 | self.__buffer = None |
|
|
2016 | 2457 | |
|
|
2017 | 2458 | |
|
|
2018 | 2459 | return |
|
|
2019 | 2460 | |
|
|
2020 | 2461 | class MeteorOperations(): |
|
|
2021 | 2462 | |
|
|
2022 | 2463 | def __init__(self): |
|
|
2023 | 2464 | |
|
|
2024 | 2465 | return |
|
|
2025 | 2466 | |
|
|
2026 | 2467 | def getMeteorParams(self, arrayParameters0, azimuth, h, pairsList, jph): |
|
|
2027 | 2468 | |
|
|
2028 | 2469 | arrayParameters = arrayParameters0.copy() |
|
|
2029 | 2470 | hmin = h[0] |
|
|
2030 | 2471 | hmax = h[1] |
|
|
2031 | 2472 | |
|
|
2032 | 2473 | #Calculate AOA (Error N 3, 4) |
|
|
2033 | 2474 | #JONES ET AL. 1998 |
|
|
2034 | 2475 | AOAthresh = numpy.pi/8 |
|
|
2035 | 2476 | error = arrayParameters[:,-1] |
|
|
2036 | 2477 | phases = -arrayParameters[:,8:12] + jph |
|
|
2037 | 2478 | # phases = numpy.unwrap(phases) |
|
|
2038 | 2479 | arrayParameters[:,3:6], arrayParameters[:,-1] = self.__getAOA(phases, pairsList, error, AOAthresh, azimuth) |
|
|
2039 | 2480 | |
|
|
2040 | 2481 | #Calculate Heights (Error N 13 and 14) |
|
|
2041 | 2482 | error = arrayParameters[:,-1] |
|
|
2042 | 2483 | Ranges = arrayParameters[:,1] |
|
|
2043 | 2484 | zenith = arrayParameters[:,4] |
|
|
2044 | 2485 | arrayParameters[:,2], arrayParameters[:,-1] = self.__getHeights(Ranges, zenith, error, hmin, hmax) |
|
|
2045 | 2486 | |
|
|
2046 | 2487 | #----------------------- Get Final data ------------------------------------ |
|
|
2047 | 2488 | # error = arrayParameters[:,-1] |
|
|
2048 | 2489 | # ind1 = numpy.where(error==0)[0] |
|
|
2049 | 2490 | # arrayParameters = arrayParameters[ind1,:] |
|
|
2050 | 2491 | |
|
|
2051 | 2492 | return arrayParameters |
|
|
2052 | 2493 | |
|
|
2053 |
def |
|
|
|
2494 | def getAOA(self, phases, pairsList, error, AOAthresh, azimuth): | |
|
|
2054 | 2495 | |
|
|
2055 | 2496 | arrayAOA = numpy.zeros((phases.shape[0],3)) |
|
|
2056 | 2497 | cosdir0, cosdir = self.__getDirectionCosines(phases, pairsList) |
|
|
2057 | 2498 | |
|
|
2058 | 2499 | arrayAOA[:,:2] = self.__calculateAOA(cosdir, azimuth) |
|
|
2059 | 2500 | cosDirError = numpy.sum(numpy.abs(cosdir0 - cosdir), axis = 1) |
|
|
2060 | 2501 | arrayAOA[:,2] = cosDirError |
|
|
2061 | 2502 | |
|
|
2062 | 2503 | azimuthAngle = arrayAOA[:,0] |
|
|
2063 | 2504 | zenithAngle = arrayAOA[:,1] |
|
|
2064 | 2505 | |
|
|
2065 | 2506 | #Setting Error |
|
|
2066 | 2507 | indError = numpy.where(numpy.logical_or(error == 3, error == 4))[0] |
|
|
2067 | 2508 | error[indError] = 0 |
|
|
2068 | 2509 | #Number 3: AOA not fesible |
|
|
2069 | 2510 | indInvalid = numpy.where(numpy.logical_and((numpy.logical_or(numpy.isnan(zenithAngle), numpy.isnan(azimuthAngle))),error == 0))[0] |
|
|
2070 | 2511 | error[indInvalid] = 3 |
|
|
2071 | 2512 | #Number 4: Large difference in AOAs obtained from different antenna baselines |
|
|
2072 | 2513 | indInvalid = numpy.where(numpy.logical_and(cosDirError > AOAthresh,error == 0))[0] |
|
|
2073 | 2514 | error[indInvalid] = 4 |
|
|
2074 | 2515 | return arrayAOA, error |
|
|
2075 | 2516 | |
|
|
2076 | 2517 | def __getDirectionCosines(self, arrayPhase, pairsList): |
|
|
2077 | 2518 | |
|
|
2078 | 2519 | #Initializing some variables |
|
|
2079 | 2520 | ang_aux = numpy.array([-8,-7,-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6,7,8])*2*numpy.pi |
|
|
2080 | 2521 | ang_aux = ang_aux.reshape(1,ang_aux.size) |
|
|
2081 | 2522 | |
|
|
2082 | 2523 | cosdir = numpy.zeros((arrayPhase.shape[0],2)) |
|
|
2083 | 2524 | cosdir0 = numpy.zeros((arrayPhase.shape[0],2)) |
|
|
2084 | 2525 | |
|
|
2085 | 2526 | |
|
|
2086 | 2527 | for i in range(2): |
|
|
2087 | 2528 | #First Estimation |
|
|
2088 | 2529 | phi0_aux = arrayPhase[:,pairsList[i][0]] + arrayPhase[:,pairsList[i][1]] |
|
|
2089 | 2530 | #Dealias |
|
|
2090 | 2531 | indcsi = numpy.where(phi0_aux > numpy.pi) |
|
|
2091 | 2532 | phi0_aux[indcsi] -= 2*numpy.pi |
|
|
2092 | 2533 | indcsi = numpy.where(phi0_aux < -numpy.pi) |
|
|
2093 | 2534 | phi0_aux[indcsi] += 2*numpy.pi |
|
|
2094 | 2535 | #Direction Cosine 0 |
|
|
2095 | 2536 | cosdir0[:,i] = -(phi0_aux)/(2*numpy.pi*0.5) |
|
|
2096 | 2537 | |
|
|
2097 | 2538 | #Most-Accurate Second Estimation |
|
|
2098 | 2539 | phi1_aux = arrayPhase[:,pairsList[i][0]] - arrayPhase[:,pairsList[i][1]] |
|
|
2099 | 2540 | phi1_aux = phi1_aux.reshape(phi1_aux.size,1) |
|
|
2100 | 2541 | #Direction Cosine 1 |
|
|
2101 | 2542 | cosdir1 = -(phi1_aux + ang_aux)/(2*numpy.pi*4.5) |
|
|
2102 | 2543 | |
|
|
2103 | 2544 | #Searching the correct Direction Cosine |
|
|
2104 | 2545 | cosdir0_aux = cosdir0[:,i] |
|
|
2105 | 2546 | cosdir0_aux = cosdir0_aux.reshape(cosdir0_aux.size,1) |
|
|
2106 | 2547 | #Minimum Distance |
|
|
2107 | 2548 | cosDiff = (cosdir1 - cosdir0_aux)**2 |
|
|
2108 | 2549 | indcos = cosDiff.argmin(axis = 1) |
|
|
2109 | 2550 | #Saving Value obtained |
|
|
2110 | 2551 | cosdir[:,i] = cosdir1[numpy.arange(len(indcos)),indcos] |
|
|
2111 | 2552 | |
|
|
2112 | 2553 | return cosdir0, cosdir |
|
|
2113 | 2554 | |
|
|
2114 | 2555 | def __calculateAOA(self, cosdir, azimuth): |
|
|
2115 | 2556 | cosdirX = cosdir[:,0] |
|
|
2116 | 2557 | cosdirY = cosdir[:,1] |
|
|
2117 | 2558 | |
|
|
2118 | 2559 | zenithAngle = numpy.arccos(numpy.sqrt(1 - cosdirX**2 - cosdirY**2))*180/numpy.pi |
|
|
2119 | 2560 | azimuthAngle = numpy.arctan2(cosdirX,cosdirY)*180/numpy.pi + azimuth #0 deg north, 90 deg east |
|
|
2120 | 2561 | angles = numpy.vstack((azimuthAngle, zenithAngle)).transpose() |
|
|
2121 | 2562 | |
|
|
2122 | 2563 | return angles |
|
|
2123 | 2564 | |
|
|
2124 | 2565 | def __getHeights(self, Ranges, zenith, error, minHeight, maxHeight): |
|
|
2125 | 2566 | |
|
|
2126 | 2567 | Ramb = 375 #Ramb = c/(2*PRF) |
|
|
2127 | 2568 | Re = 6371 #Earth Radius |
|
|
2128 | 2569 | heights = numpy.zeros(Ranges.shape) |
|
|
2129 | 2570 | |
|
|
2130 | 2571 | R_aux = numpy.array([0,1,2])*Ramb |
|
|
2131 | 2572 | R_aux = R_aux.reshape(1,R_aux.size) |
|
|
2132 | 2573 | |
|
|
2133 | 2574 | Ranges = Ranges.reshape(Ranges.size,1) |
|
|
2134 | 2575 | |
|
|
2135 | 2576 | Ri = Ranges + R_aux |
|
|
2136 | 2577 | hi = numpy.sqrt(Re**2 + Ri**2 + (2*Re*numpy.cos(zenith*numpy.pi/180)*Ri.transpose()).transpose()) - Re |
|
|
2137 | 2578 | |
|
|
2138 | 2579 | #Check if there is a height between 70 and 110 km |
|
|
2139 | 2580 | h_bool = numpy.sum(numpy.logical_and(hi > minHeight, hi < maxHeight), axis = 1) |
|
|
2140 | 2581 | ind_h = numpy.where(h_bool == 1)[0] |
|
|
2141 | 2582 | |
|
|
2142 | 2583 | hCorr = hi[ind_h, :] |
|
|
2143 | 2584 | ind_hCorr = numpy.where(numpy.logical_and(hi > minHeight, hi < maxHeight)) |
|
|
2144 | 2585 | |
|
|
2145 | 2586 | hCorr = hi[ind_hCorr] |
|
|
2146 | 2587 | heights[ind_h] = hCorr |
|
|
2147 | 2588 | |
|
|
2148 | 2589 | #Setting Error |
|
|
2149 | 2590 | #Number 13: Height unresolvable echo: not valid height within 70 to 110 km |
|
|
2150 | 2591 | #Number 14: Height ambiguous echo: more than one possible height within 70 to 110 km |
|
|
2151 | 2592 | indError = numpy.where(numpy.logical_or(error == 13, error == 14))[0] |
|
|
2152 | 2593 | error[indError] = 0 |
|
|
2153 | 2594 | indInvalid2 = numpy.where(numpy.logical_and(h_bool > 1, error == 0))[0] |
|
|
2154 | 2595 | error[indInvalid2] = 14 |
|
|
2155 | 2596 | indInvalid1 = numpy.where(numpy.logical_and(h_bool == 0, error == 0))[0] |
|
|
2156 | 2597 | error[indInvalid1] = 13 |
|
|
2157 | 2598 | |
|
|
2158 | 2599 | return heights, error No newline at end of file |
General Comments 0
You need to be logged in to leave comments.
Login now