| @@ -1,1770 +1,1780 | |||||
| 1 | import numpy |
|
1 | import numpy | |
| 2 | import math |
|
2 | import math | |
| 3 | from scipy import optimize |
|
3 | from scipy import optimize | |
| 4 | from scipy import interpolate |
|
4 | from scipy import interpolate | |
| 5 | from scipy import signal |
|
5 | from scipy import signal | |
| 6 | from scipy import stats |
|
6 | from scipy import stats | |
| 7 | import re |
|
7 | import re | |
| 8 | import datetime |
|
8 | import datetime | |
| 9 | import copy |
|
9 | import copy | |
| 10 | import sys |
|
10 | import sys | |
| 11 | import importlib |
|
11 | import importlib | |
| 12 | import itertools |
|
12 | import itertools | |
| 13 |
|
13 | |||
| 14 | from jroproc_base import ProcessingUnit, Operation |
|
14 | from jroproc_base import ProcessingUnit, Operation | |
| 15 | from model.data.jrodata import Parameters |
|
15 | from model.data.jrodata import Parameters | |
| 16 |
|
16 | |||
| 17 |
|
17 | |||
| 18 | class ParametersProc(ProcessingUnit): |
|
18 | class ParametersProc(ProcessingUnit): | |
| 19 |
|
19 | |||
| 20 | nSeconds = None |
|
20 | nSeconds = None | |
| 21 |
|
21 | |||
| 22 | def __init__(self): |
|
22 | def __init__(self): | |
| 23 | ProcessingUnit.__init__(self) |
|
23 | ProcessingUnit.__init__(self) | |
| 24 |
|
24 | |||
| 25 | # self.objectDict = {} |
|
25 | # self.objectDict = {} | |
| 26 | self.buffer = None |
|
26 | self.buffer = None | |
| 27 | self.firstdatatime = None |
|
27 | self.firstdatatime = None | |
| 28 | self.profIndex = 0 |
|
28 | self.profIndex = 0 | |
| 29 | self.dataOut = Parameters() |
|
29 | self.dataOut = Parameters() | |
| 30 |
|
30 | |||
| 31 | def __updateObjFromInput(self): |
|
31 | def __updateObjFromInput(self): | |
| 32 |
|
32 | |||
| 33 | self.dataOut.inputUnit = self.dataIn.type |
|
33 | self.dataOut.inputUnit = self.dataIn.type | |
| 34 |
|
34 | |||
| 35 | self.dataOut.timeZone = self.dataIn.timeZone |
|
35 | self.dataOut.timeZone = self.dataIn.timeZone | |
| 36 | self.dataOut.dstFlag = self.dataIn.dstFlag |
|
36 | self.dataOut.dstFlag = self.dataIn.dstFlag | |
| 37 | self.dataOut.errorCount = self.dataIn.errorCount |
|
37 | self.dataOut.errorCount = self.dataIn.errorCount | |
| 38 | self.dataOut.useLocalTime = self.dataIn.useLocalTime |
|
38 | self.dataOut.useLocalTime = self.dataIn.useLocalTime | |
| 39 |
|
39 | |||
| 40 | self.dataOut.radarControllerHeaderObj = self.dataIn.radarControllerHeaderObj.copy() |
|
40 | self.dataOut.radarControllerHeaderObj = self.dataIn.radarControllerHeaderObj.copy() | |
| 41 | self.dataOut.systemHeaderObj = self.dataIn.systemHeaderObj.copy() |
|
41 | self.dataOut.systemHeaderObj = self.dataIn.systemHeaderObj.copy() | |
| 42 | self.dataOut.channelList = self.dataIn.channelList |
|
42 | self.dataOut.channelList = self.dataIn.channelList | |
| 43 | self.dataOut.heightList = self.dataIn.heightList |
|
43 | self.dataOut.heightList = self.dataIn.heightList | |
| 44 | self.dataOut.dtype = numpy.dtype([('real','<f4'),('imag','<f4')]) |
|
44 | self.dataOut.dtype = numpy.dtype([('real','<f4'),('imag','<f4')]) | |
| 45 | # self.dataOut.nHeights = self.dataIn.nHeights |
|
45 | # self.dataOut.nHeights = self.dataIn.nHeights | |
| 46 | # self.dataOut.nChannels = self.dataIn.nChannels |
|
46 | # self.dataOut.nChannels = self.dataIn.nChannels | |
| 47 | self.dataOut.nBaud = self.dataIn.nBaud |
|
47 | self.dataOut.nBaud = self.dataIn.nBaud | |
| 48 | self.dataOut.nCode = self.dataIn.nCode |
|
48 | self.dataOut.nCode = self.dataIn.nCode | |
| 49 | self.dataOut.code = self.dataIn.code |
|
49 | self.dataOut.code = self.dataIn.code | |
| 50 | # self.dataOut.nProfiles = self.dataOut.nFFTPoints |
|
50 | # self.dataOut.nProfiles = self.dataOut.nFFTPoints | |
| 51 | self.dataOut.flagTimeBlock = self.dataIn.flagTimeBlock |
|
51 | self.dataOut.flagTimeBlock = self.dataIn.flagTimeBlock | |
| 52 | self.dataOut.utctime = self.firstdatatime |
|
52 | self.dataOut.utctime = self.firstdatatime | |
| 53 | self.dataOut.flagDecodeData = self.dataIn.flagDecodeData #asumo q la data esta decodificada |
|
53 | self.dataOut.flagDecodeData = self.dataIn.flagDecodeData #asumo q la data esta decodificada | |
| 54 | self.dataOut.flagDeflipData = self.dataIn.flagDeflipData #asumo q la data esta sin flip |
|
54 | self.dataOut.flagDeflipData = self.dataIn.flagDeflipData #asumo q la data esta sin flip | |
| 55 | # self.dataOut.nCohInt = self.dataIn.nCohInt |
|
55 | # self.dataOut.nCohInt = self.dataIn.nCohInt | |
| 56 | # self.dataOut.nIncohInt = 1 |
|
56 | # self.dataOut.nIncohInt = 1 | |
| 57 | self.dataOut.ippSeconds = self.dataIn.ippSeconds |
|
57 | self.dataOut.ippSeconds = self.dataIn.ippSeconds | |
| 58 | # self.dataOut.windowOfFilter = self.dataIn.windowOfFilter |
|
58 | # self.dataOut.windowOfFilter = self.dataIn.windowOfFilter | |
| 59 | self.dataOut.timeInterval = self.dataIn.timeInterval |
|
59 | self.dataOut.timeInterval = self.dataIn.timeInterval | |
| 60 | self.dataOut.heightList = self.dataIn.getHeiRange() |
|
60 | self.dataOut.heightList = self.dataIn.getHeiRange() | |
| 61 | self.dataOut.frequency = self.dataIn.frequency |
|
61 | self.dataOut.frequency = self.dataIn.frequency | |
| 62 |
|
62 | |||
| 63 | def run(self, nSeconds = None, nProfiles = None): |
|
63 | def run(self, nSeconds = None, nProfiles = None): | |
| 64 |
|
64 | |||
| 65 |
|
65 | |||
| 66 |
|
66 | |||
| 67 | if self.firstdatatime == None: |
|
67 | if self.firstdatatime == None: | |
| 68 | self.firstdatatime = self.dataIn.utctime |
|
68 | self.firstdatatime = self.dataIn.utctime | |
| 69 |
|
69 | |||
| 70 | #---------------------- Voltage Data --------------------------- |
|
70 | #---------------------- Voltage Data --------------------------- | |
| 71 |
|
71 | |||
| 72 | if self.dataIn.type == "Voltage": |
|
72 | if self.dataIn.type == "Voltage": | |
| 73 | self.dataOut.flagNoData = True |
|
73 | self.dataOut.flagNoData = True | |
| 74 | if nSeconds != None: |
|
74 | if nSeconds != None: | |
| 75 | self.nSeconds = nSeconds |
|
75 | self.nSeconds = nSeconds | |
| 76 | self.nProfiles= int(numpy.floor(nSeconds/(self.dataIn.ippSeconds*self.dataIn.nCohInt))) |
|
76 | self.nProfiles= int(numpy.floor(nSeconds/(self.dataIn.ippSeconds*self.dataIn.nCohInt))) | |
| 77 |
|
77 | |||
| 78 | if self.buffer == None: |
|
78 | if self.buffer == None: | |
| 79 | self.buffer = numpy.zeros((self.dataIn.nChannels, |
|
79 | self.buffer = numpy.zeros((self.dataIn.nChannels, | |
| 80 | self.nProfiles, |
|
80 | self.nProfiles, | |
| 81 | self.dataIn.nHeights), |
|
81 | self.dataIn.nHeights), | |
| 82 | dtype='complex') |
|
82 | dtype='complex') | |
| 83 |
|
83 | |||
| 84 | self.buffer[:,self.profIndex,:] = self.dataIn.data.copy() |
|
84 | self.buffer[:,self.profIndex,:] = self.dataIn.data.copy() | |
| 85 | self.profIndex += 1 |
|
85 | self.profIndex += 1 | |
| 86 |
|
86 | |||
| 87 | if self.profIndex == self.nProfiles: |
|
87 | if self.profIndex == self.nProfiles: | |
| 88 |
|
88 | |||
| 89 | self.__updateObjFromInput() |
|
89 | self.__updateObjFromInput() | |
| 90 | self.dataOut.data_pre = self.buffer.copy() |
|
90 | self.dataOut.data_pre = self.buffer.copy() | |
| 91 | self.dataOut.paramInterval = nSeconds |
|
91 | self.dataOut.paramInterval = nSeconds | |
| 92 | self.dataOut.flagNoData = False |
|
92 | self.dataOut.flagNoData = False | |
| 93 |
|
93 | |||
| 94 | self.buffer = None |
|
94 | self.buffer = None | |
| 95 | self.firstdatatime = None |
|
95 | self.firstdatatime = None | |
| 96 | self.profIndex = 0 |
|
96 | self.profIndex = 0 | |
| 97 | return |
|
97 | return | |
| 98 |
|
98 | |||
| 99 | #---------------------- Spectra Data --------------------------- |
|
99 | #---------------------- Spectra Data --------------------------- | |
| 100 |
|
100 | |||
| 101 | if self.dataIn.type == "Spectra": |
|
101 | if self.dataIn.type == "Spectra": | |
| 102 | self.dataOut.data_pre = self.dataIn.data_spc.copy() |
|
102 | self.dataOut.data_pre = self.dataIn.data_spc.copy() | |
| 103 | self.dataOut.abscissaList = self.dataIn.getVelRange(1) |
|
103 | self.dataOut.abscissaList = self.dataIn.getVelRange(1) | |
| 104 | self.dataOut.noise = self.dataIn.getNoise() |
|
104 | self.dataOut.noise = self.dataIn.getNoise() | |
| 105 | self.dataOut.normFactor = self.dataIn.normFactor |
|
105 | self.dataOut.normFactor = self.dataIn.normFactor | |
|
|
106 | self.dataOut.groupList = self.dataIn.pairsList | |||
| 106 | self.dataOut.flagNoData = False |
|
107 | self.dataOut.flagNoData = False | |
| 107 |
|
108 | |||
| 108 | #---------------------- Correlation Data --------------------------- |
|
109 | #---------------------- Correlation Data --------------------------- | |
| 109 |
|
110 | |||
| 110 | if self.dataIn.type == "Correlation": |
|
111 | if self.dataIn.type == "Correlation": | |
| 111 | lagRRange = self.dataIn.lagR |
|
112 | lagRRange = self.dataIn.lagR | |
| 112 | indR = numpy.where(lagRRange == 0)[0][0] |
|
113 | indR = numpy.where(lagRRange == 0)[0][0] | |
| 113 |
|
114 | |||
| 114 | self.dataOut.data_pre = self.dataIn.data_corr.copy()[:,:,indR,:] |
|
115 | self.dataOut.data_pre = self.dataIn.data_corr.copy()[:,:,indR,:] | |
| 115 | self.dataOut.abscissaList = self.dataIn.getLagTRange(1) |
|
116 | self.dataOut.abscissaList = self.dataIn.getLagTRange(1) | |
| 116 | self.dataOut.noise = self.dataIn.noise |
|
117 | self.dataOut.noise = self.dataIn.noise | |
| 117 | self.dataOut.normFactor = self.dataIn.normFactor |
|
118 | self.dataOut.normFactor = self.dataIn.normFactor | |
| 118 | self.dataOut.data_SNR = self.dataIn.SNR |
|
119 | self.dataOut.data_SNR = self.dataIn.SNR | |
| 119 | self.dataOut.groupList = self.dataIn.pairsList |
|
120 | self.dataOut.groupList = self.dataIn.pairsList | |
| 120 | self.dataOut.flagNoData = False |
|
121 | self.dataOut.flagNoData = False | |
| 121 |
|
122 | |||
| 122 | #---------------------- Correlation Data --------------------------- |
|
123 | #---------------------- Correlation Data --------------------------- | |
| 123 |
|
124 | |||
| 124 | if self.dataIn.type == "Parameters": |
|
125 | if self.dataIn.type == "Parameters": | |
| 125 | self.dataOut.copy(self.dataIn) |
|
126 | self.dataOut.copy(self.dataIn) | |
| 126 | self.dataOut.flagNoData = False |
|
127 | self.dataOut.flagNoData = False | |
| 127 |
|
128 | |||
| 128 | return True |
|
129 | return True | |
| 129 |
|
130 | |||
| 130 | self.__updateObjFromInput() |
|
131 | self.__updateObjFromInput() | |
| 131 | self.firstdatatime = None |
|
132 | self.firstdatatime = None | |
| 132 | self.dataOut.utctimeInit = self.dataIn.utctime |
|
133 | self.dataOut.utctimeInit = self.dataIn.utctime | |
| 133 | self.dataOut.outputInterval = self.dataIn.timeInterval |
|
134 | self.dataOut.outputInterval = self.dataIn.timeInterval | |
| 134 |
|
135 | |||
| 135 | #------------------- Get Moments ---------------------------------- |
|
136 | #------------------- Get Moments ---------------------------------- | |
| 136 | def GetMoments(self, channelList = None): |
|
137 | def GetMoments(self, channelList = None): | |
| 137 | ''' |
|
138 | ''' | |
| 138 | Function GetMoments() |
|
139 | Function GetMoments() | |
| 139 |
|
140 | |||
| 140 | Input: |
|
141 | Input: | |
| 141 | channelList : simple channel list to select e.g. [2,3,7] |
|
142 | channelList : simple channel list to select e.g. [2,3,7] | |
| 142 | self.dataOut.data_pre |
|
143 | self.dataOut.data_pre | |
| 143 | self.dataOut.abscissaList |
|
144 | self.dataOut.abscissaList | |
| 144 | self.dataOut.noise |
|
145 | self.dataOut.noise | |
| 145 |
|
146 | |||
| 146 | Affected: |
|
147 | Affected: | |
| 147 | self.dataOut.data_param |
|
148 | self.dataOut.data_param | |
| 148 | self.dataOut.data_SNR |
|
149 | self.dataOut.data_SNR | |
| 149 |
|
150 | |||
| 150 | ''' |
|
151 | ''' | |
| 151 | data = self.dataOut.data_pre |
|
152 | data = self.dataOut.data_pre | |
| 152 | absc = self.dataOut.abscissaList[:-1] |
|
153 | absc = self.dataOut.abscissaList[:-1] | |
| 153 | noise = self.dataOut.noise |
|
154 | noise = self.dataOut.noise | |
| 154 |
|
155 | |||
| 155 | data_param = numpy.zeros((data.shape[0], 4, data.shape[2])) |
|
156 | data_param = numpy.zeros((data.shape[0], 4, data.shape[2])) | |
| 156 |
|
157 | |||
| 157 | if channelList== None: |
|
158 | if channelList== None: | |
| 158 | channelList = self.dataIn.channelList |
|
159 | channelList = self.dataIn.channelList | |
| 159 | self.dataOut.channelList = channelList |
|
160 | self.dataOut.channelList = channelList | |
| 160 |
|
161 | |||
| 161 | for ind in channelList: |
|
162 | for ind in channelList: | |
| 162 | data_param[ind,:,:] = self.__calculateMoments(data[ind,:,:], absc, noise[ind]) |
|
163 | data_param[ind,:,:] = self.__calculateMoments(data[ind,:,:], absc, noise[ind]) | |
| 163 |
|
164 | |||
| 164 | self.dataOut.data_param = data_param[:,1:,:] |
|
165 | self.dataOut.data_param = data_param[:,1:,:] | |
| 165 | self.dataOut.data_SNR = data_param[:,0] |
|
166 | self.dataOut.data_SNR = data_param[:,0] | |
| 166 | return |
|
167 | return | |
| 167 |
|
168 | |||
| 168 | 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): |
|
169 | 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): | |
| 169 |
|
170 | |||
| 170 | if (nicoh == None): nicoh = 1 |
|
171 | if (nicoh == None): nicoh = 1 | |
| 171 | if (graph == None): graph = 0 |
|
172 | if (graph == None): graph = 0 | |
| 172 | if (smooth == None): smooth = 0 |
|
173 | if (smooth == None): smooth = 0 | |
| 173 | elif (self.smooth < 3): smooth = 0 |
|
174 | elif (self.smooth < 3): smooth = 0 | |
| 174 |
|
175 | |||
| 175 | if (type1 == None): type1 = 0 |
|
176 | if (type1 == None): type1 = 0 | |
| 176 | if (fwindow == None): fwindow = numpy.zeros(oldfreq.size) + 1 |
|
177 | if (fwindow == None): fwindow = numpy.zeros(oldfreq.size) + 1 | |
| 177 | if (snrth == None): snrth = -3 |
|
178 | if (snrth == None): snrth = -3 | |
| 178 | if (dc == None): dc = 0 |
|
179 | if (dc == None): dc = 0 | |
| 179 | if (aliasing == None): aliasing = 0 |
|
180 | if (aliasing == None): aliasing = 0 | |
| 180 | if (oldfd == None): oldfd = 0 |
|
181 | if (oldfd == None): oldfd = 0 | |
| 181 | if (wwauto == None): wwauto = 0 |
|
182 | if (wwauto == None): wwauto = 0 | |
| 182 |
|
183 | |||
| 183 | if (n0 < 1.e-20): n0 = 1.e-20 |
|
184 | if (n0 < 1.e-20): n0 = 1.e-20 | |
| 184 |
|
185 | |||
| 185 | freq = oldfreq |
|
186 | freq = oldfreq | |
| 186 | vec_power = numpy.zeros(oldspec.shape[1]) |
|
187 | vec_power = numpy.zeros(oldspec.shape[1]) | |
| 187 | vec_fd = numpy.zeros(oldspec.shape[1]) |
|
188 | vec_fd = numpy.zeros(oldspec.shape[1]) | |
| 188 | vec_w = numpy.zeros(oldspec.shape[1]) |
|
189 | vec_w = numpy.zeros(oldspec.shape[1]) | |
| 189 | vec_snr = numpy.zeros(oldspec.shape[1]) |
|
190 | vec_snr = numpy.zeros(oldspec.shape[1]) | |
| 190 |
|
191 | |||
| 191 | for ind in range(oldspec.shape[1]): |
|
192 | for ind in range(oldspec.shape[1]): | |
| 192 |
|
193 | |||
| 193 | spec = oldspec[:,ind] |
|
194 | spec = oldspec[:,ind] | |
| 194 | aux = spec*fwindow |
|
195 | aux = spec*fwindow | |
| 195 | max_spec = aux.max() |
|
196 | max_spec = aux.max() | |
| 196 | m = list(aux).index(max_spec) |
|
197 | m = list(aux).index(max_spec) | |
| 197 |
|
198 | |||
| 198 | #Smooth |
|
199 | #Smooth | |
| 199 | if (smooth == 0): spec2 = spec |
|
200 | if (smooth == 0): spec2 = spec | |
| 200 | else: spec2 = scipy.ndimage.filters.uniform_filter1d(spec,size=smooth) |
|
201 | else: spec2 = scipy.ndimage.filters.uniform_filter1d(spec,size=smooth) | |
| 201 |
|
202 | |||
| 202 | # Calculo de Momentos |
|
203 | # Calculo de Momentos | |
| 203 | bb = spec2[range(m,spec2.size)] |
|
204 | bb = spec2[range(m,spec2.size)] | |
| 204 | bb = (bb<n0).nonzero() |
|
205 | bb = (bb<n0).nonzero() | |
| 205 | bb = bb[0] |
|
206 | bb = bb[0] | |
| 206 |
|
207 | |||
| 207 | ss = spec2[range(0,m + 1)] |
|
208 | ss = spec2[range(0,m + 1)] | |
| 208 | ss = (ss<n0).nonzero() |
|
209 | ss = (ss<n0).nonzero() | |
| 209 | ss = ss[0] |
|
210 | ss = ss[0] | |
| 210 |
|
211 | |||
| 211 | if (bb.size == 0): |
|
212 | if (bb.size == 0): | |
| 212 | bb0 = spec.size - 1 - m |
|
213 | bb0 = spec.size - 1 - m | |
| 213 | else: |
|
214 | else: | |
| 214 | bb0 = bb[0] - 1 |
|
215 | bb0 = bb[0] - 1 | |
| 215 | if (bb0 < 0): |
|
216 | if (bb0 < 0): | |
| 216 | bb0 = 0 |
|
217 | bb0 = 0 | |
| 217 |
|
218 | |||
| 218 | if (ss.size == 0): ss1 = 1 |
|
219 | if (ss.size == 0): ss1 = 1 | |
| 219 | else: ss1 = max(ss) + 1 |
|
220 | else: ss1 = max(ss) + 1 | |
| 220 |
|
221 | |||
| 221 | if (ss1 > m): ss1 = m |
|
222 | if (ss1 > m): ss1 = m | |
| 222 |
|
223 | |||
| 223 | valid = numpy.asarray(range(int(m + bb0 - ss1 + 1))) + ss1 |
|
224 | valid = numpy.asarray(range(int(m + bb0 - ss1 + 1))) + ss1 | |
| 224 | power = ((spec2[valid] - n0)*fwindow[valid]).sum() |
|
225 | power = ((spec2[valid] - n0)*fwindow[valid]).sum() | |
| 225 | fd = ((spec2[valid]- n0)*freq[valid]*fwindow[valid]).sum()/power |
|
226 | fd = ((spec2[valid]- n0)*freq[valid]*fwindow[valid]).sum()/power | |
| 226 | w = math.sqrt(((spec2[valid] - n0)*fwindow[valid]*(freq[valid]- fd)**2).sum()/power) |
|
227 | w = math.sqrt(((spec2[valid] - n0)*fwindow[valid]*(freq[valid]- fd)**2).sum()/power) | |
| 227 | snr = (spec2.mean()-n0)/n0 |
|
228 | snr = (spec2.mean()-n0)/n0 | |
| 228 |
|
229 | |||
| 229 | if (snr < 1.e-20) : |
|
230 | if (snr < 1.e-20) : | |
| 230 | snr = 1.e-20 |
|
231 | snr = 1.e-20 | |
| 231 |
|
232 | |||
| 232 | vec_power[ind] = power |
|
233 | vec_power[ind] = power | |
| 233 | vec_fd[ind] = fd |
|
234 | vec_fd[ind] = fd | |
| 234 | vec_w[ind] = w |
|
235 | vec_w[ind] = w | |
| 235 | vec_snr[ind] = snr |
|
236 | vec_snr[ind] = snr | |
| 236 |
|
237 | |||
| 237 | moments = numpy.vstack((vec_snr, vec_power, vec_fd, vec_w)) |
|
238 | moments = numpy.vstack((vec_snr, vec_power, vec_fd, vec_w)) | |
| 238 | return moments |
|
239 | return moments | |
| 239 |
|
240 | |||
|
|
241 | #------------------ Get SA Parameters -------------------------- | |||
|
|
242 | def GetSAParameters(self): | |||
|
|
243 | data = self.dataOut.data_pre | |||
|
|
244 | crossdata = self.dataIn.data_cspc | |||
|
|
245 | a = 1 | |||
|
|
246 | ||||
|
|
247 | ||||
|
|
248 | ||||
| 240 | #------------------- Get Lags ---------------------------------- |
|
249 | #------------------- Get Lags ---------------------------------- | |
| 241 |
|
250 | |||
| 242 | def GetLags(self): |
|
251 | def GetLags(self): | |
| 243 | ''' |
|
252 | ''' | |
| 244 | Function GetMoments() |
|
253 | Function GetMoments() | |
| 245 |
|
254 | |||
| 246 | Input: |
|
255 | Input: | |
| 247 | self.dataOut.data_pre |
|
256 | self.dataOut.data_pre | |
| 248 | self.dataOut.abscissaList |
|
257 | self.dataOut.abscissaList | |
| 249 | self.dataOut.noise |
|
258 | self.dataOut.noise | |
| 250 | self.dataOut.normFactor |
|
259 | self.dataOut.normFactor | |
| 251 | self.dataOut.data_SNR |
|
260 | self.dataOut.data_SNR | |
| 252 | self.dataOut.groupList |
|
261 | self.dataOut.groupList | |
| 253 | self.dataOut.nChannels |
|
262 | self.dataOut.nChannels | |
| 254 |
|
263 | |||
| 255 | Affected: |
|
264 | Affected: | |
| 256 | self.dataOut.data_param |
|
265 | self.dataOut.data_param | |
| 257 |
|
266 | |||
| 258 | ''' |
|
267 | ''' | |
|
|
268 | ||||
| 259 | data = self.dataOut.data_pre |
|
269 | data = self.dataOut.data_pre | |
| 260 | normFactor = self.dataOut.normFactor |
|
270 | normFactor = self.dataOut.normFactor | |
| 261 | nHeights = self.dataOut.nHeights |
|
271 | nHeights = self.dataOut.nHeights | |
| 262 | absc = self.dataOut.abscissaList[:-1] |
|
272 | absc = self.dataOut.abscissaList[:-1] | |
| 263 | noise = self.dataOut.noise |
|
273 | noise = self.dataOut.noise | |
| 264 | SNR = self.dataOut.data_SNR |
|
274 | SNR = self.dataOut.data_SNR | |
| 265 | pairsList = self.dataOut.groupList |
|
275 | pairsList = self.dataOut.groupList | |
| 266 | nChannels = self.dataOut.nChannels |
|
276 | nChannels = self.dataOut.nChannels | |
| 267 | pairsAutoCorr, pairsCrossCorr = self.__getPairsAutoCorr(pairsList, nChannels) |
|
277 | pairsAutoCorr, pairsCrossCorr = self.__getPairsAutoCorr(pairsList, nChannels) | |
| 268 | self.dataOut.data_param = numpy.zeros((len(pairsCrossCorr)*2 + 1, nHeights)) |
|
278 | self.dataOut.data_param = numpy.zeros((len(pairsCrossCorr)*2 + 1, nHeights)) | |
| 269 |
|
279 | |||
| 270 | dataNorm = numpy.abs(data) |
|
280 | dataNorm = numpy.abs(data) | |
| 271 | for l in range(len(pairsList)): |
|
281 | for l in range(len(pairsList)): | |
| 272 | dataNorm[l,:,:] = dataNorm[l,:,:]/normFactor[l,:] |
|
282 | dataNorm[l,:,:] = dataNorm[l,:,:]/normFactor[l,:] | |
| 273 |
|
283 | |||
| 274 | self.dataOut.data_param[:-1,:] = self.__calculateTaus(dataNorm, pairsCrossCorr, pairsAutoCorr, absc) |
|
284 | self.dataOut.data_param[:-1,:] = self.__calculateTaus(dataNorm, pairsCrossCorr, pairsAutoCorr, absc) | |
| 275 | self.dataOut.data_param[-1,:] = self.__calculateLag1Phase(data, pairsAutoCorr, absc) |
|
285 | self.dataOut.data_param[-1,:] = self.__calculateLag1Phase(data, pairsAutoCorr, absc) | |
| 276 | return |
|
286 | return | |
| 277 |
|
287 | |||
| 278 | def __getPairsAutoCorr(self, pairsList, nChannels): |
|
288 | def __getPairsAutoCorr(self, pairsList, nChannels): | |
| 279 |
|
289 | |||
| 280 | pairsAutoCorr = numpy.zeros(nChannels, dtype = 'int')*numpy.nan |
|
290 | pairsAutoCorr = numpy.zeros(nChannels, dtype = 'int')*numpy.nan | |
| 281 |
|
291 | |||
| 282 | for l in range(len(pairsList)): |
|
292 | for l in range(len(pairsList)): | |
| 283 | firstChannel = pairsList[l][0] |
|
293 | firstChannel = pairsList[l][0] | |
| 284 | secondChannel = pairsList[l][1] |
|
294 | secondChannel = pairsList[l][1] | |
| 285 |
|
295 | |||
| 286 | #Obteniendo pares de Autocorrelacion |
|
296 | #Obteniendo pares de Autocorrelacion | |
| 287 | if firstChannel == secondChannel: |
|
297 | if firstChannel == secondChannel: | |
| 288 | pairsAutoCorr[firstChannel] = int(l) |
|
298 | pairsAutoCorr[firstChannel] = int(l) | |
| 289 |
|
299 | |||
| 290 | pairsAutoCorr = pairsAutoCorr.astype(int) |
|
300 | pairsAutoCorr = pairsAutoCorr.astype(int) | |
| 291 |
|
301 | |||
| 292 | pairsCrossCorr = range(len(pairsList)) |
|
302 | pairsCrossCorr = range(len(pairsList)) | |
| 293 | pairsCrossCorr = numpy.delete(pairsCrossCorr,pairsAutoCorr) |
|
303 | pairsCrossCorr = numpy.delete(pairsCrossCorr,pairsAutoCorr) | |
| 294 |
|
304 | |||
| 295 | return pairsAutoCorr, pairsCrossCorr |
|
305 | return pairsAutoCorr, pairsCrossCorr | |
| 296 |
|
306 | |||
| 297 | def __calculateTaus(self, data, pairsCrossCorr, pairsAutoCorr, lagTRange): |
|
307 | def __calculateTaus(self, data, pairsCrossCorr, pairsAutoCorr, lagTRange): | |
| 298 |
|
308 | |||
| 299 | Pt0 = data.shape[1]/2 |
|
309 | Pt0 = data.shape[1]/2 | |
| 300 | #Funcion de Autocorrelacion |
|
310 | #Funcion de Autocorrelacion | |
| 301 | dataAutoCorr = stats.nanmean(data[pairsAutoCorr,:,:], axis = 0) |
|
311 | dataAutoCorr = stats.nanmean(data[pairsAutoCorr,:,:], axis = 0) | |
| 302 |
|
312 | |||
| 303 | #Obtencion Indice de TauCross |
|
313 | #Obtencion Indice de TauCross | |
| 304 | indCross = data[pairsCrossCorr,:,:].argmax(axis = 1) |
|
314 | indCross = data[pairsCrossCorr,:,:].argmax(axis = 1) | |
| 305 | #Obtencion Indice de TauAuto |
|
315 | #Obtencion Indice de TauAuto | |
| 306 | indAuto = numpy.zeros(indCross.shape,dtype = 'int') |
|
316 | indAuto = numpy.zeros(indCross.shape,dtype = 'int') | |
| 307 | CCValue = data[pairsCrossCorr,Pt0,:] |
|
317 | CCValue = data[pairsCrossCorr,Pt0,:] | |
| 308 | for i in range(pairsCrossCorr.size): |
|
318 | for i in range(pairsCrossCorr.size): | |
| 309 | indAuto[i,:] = numpy.abs(dataAutoCorr - CCValue[i,:]).argmin(axis = 0) |
|
319 | indAuto[i,:] = numpy.abs(dataAutoCorr - CCValue[i,:]).argmin(axis = 0) | |
| 310 |
|
320 | |||
| 311 | #Obtencion de TauCross y TauAuto |
|
321 | #Obtencion de TauCross y TauAuto | |
| 312 | tauCross = lagTRange[indCross] |
|
322 | tauCross = lagTRange[indCross] | |
| 313 | tauAuto = lagTRange[indAuto] |
|
323 | tauAuto = lagTRange[indAuto] | |
| 314 |
|
324 | |||
| 315 | Nan1, Nan2 = numpy.where(tauCross == lagTRange[0]) |
|
325 | Nan1, Nan2 = numpy.where(tauCross == lagTRange[0]) | |
| 316 |
|
326 | |||
| 317 | tauCross[Nan1,Nan2] = numpy.nan |
|
327 | tauCross[Nan1,Nan2] = numpy.nan | |
| 318 | tauAuto[Nan1,Nan2] = numpy.nan |
|
328 | tauAuto[Nan1,Nan2] = numpy.nan | |
| 319 | tau = numpy.vstack((tauCross,tauAuto)) |
|
329 | tau = numpy.vstack((tauCross,tauAuto)) | |
| 320 |
|
330 | |||
| 321 | return tau |
|
331 | return tau | |
| 322 |
|
332 | |||
| 323 | def __calculateLag1Phase(self, data, pairs, lagTRange): |
|
333 | def __calculateLag1Phase(self, data, pairs, lagTRange): | |
| 324 | data1 = stats.nanmean(data[pairs,:,:], axis = 0) |
|
334 | data1 = stats.nanmean(data[pairs,:,:], axis = 0) | |
| 325 | lag1 = numpy.where(lagTRange == 0)[0][0] + 1 |
|
335 | lag1 = numpy.where(lagTRange == 0)[0][0] + 1 | |
| 326 |
|
336 | |||
| 327 | phase = numpy.angle(data1[lag1,:]) |
|
337 | phase = numpy.angle(data1[lag1,:]) | |
| 328 |
|
338 | |||
| 329 | return phase |
|
339 | return phase | |
| 330 | #------------------- Detect Meteors ------------------------------ |
|
340 | #------------------- Detect Meteors ------------------------------ | |
| 331 |
|
341 | |||
| 332 | def DetectMeteors(self, hei_ref = None, tauindex = 0, |
|
342 | def DetectMeteors(self, hei_ref = None, tauindex = 0, | |
| 333 | predefinedPhaseShifts = None, centerReceiverIndex = 2, |
|
343 | predefinedPhaseShifts = None, centerReceiverIndex = 2, | |
| 334 | cohDetection = False, cohDet_timeStep = 1, cohDet_thresh = 25, |
|
344 | cohDetection = False, cohDet_timeStep = 1, cohDet_thresh = 25, | |
| 335 | noise_timeStep = 4, noise_multiple = 4, |
|
345 | noise_timeStep = 4, noise_multiple = 4, | |
| 336 | multDet_timeLimit = 1, multDet_rangeLimit = 3, |
|
346 | multDet_timeLimit = 1, multDet_rangeLimit = 3, | |
| 337 | phaseThresh = 20, SNRThresh = 8, |
|
347 | phaseThresh = 20, SNRThresh = 8, | |
| 338 | hmin = 70, hmax=110, azimuth = 0) : |
|
348 | hmin = 70, hmax=110, azimuth = 0) : | |
| 339 |
|
349 | |||
| 340 | ''' |
|
350 | ''' | |
| 341 | Function DetectMeteors() |
|
351 | Function DetectMeteors() | |
| 342 | Project developed with paper: |
|
352 | Project developed with paper: | |
| 343 | HOLDSWORTH ET AL. 2004 |
|
353 | HOLDSWORTH ET AL. 2004 | |
| 344 |
|
354 | |||
| 345 | Input: |
|
355 | Input: | |
| 346 | self.dataOut.data_pre |
|
356 | self.dataOut.data_pre | |
| 347 |
|
357 | |||
| 348 | centerReceiverIndex: From the channels, which is the center receiver |
|
358 | centerReceiverIndex: From the channels, which is the center receiver | |
| 349 |
|
359 | |||
| 350 | hei_ref: Height reference for the Beacon signal extraction |
|
360 | hei_ref: Height reference for the Beacon signal extraction | |
| 351 | tauindex: |
|
361 | tauindex: | |
| 352 | predefinedPhaseShifts: Predefined phase offset for the voltge signals |
|
362 | predefinedPhaseShifts: Predefined phase offset for the voltge signals | |
| 353 |
|
363 | |||
| 354 | cohDetection: Whether to user Coherent detection or not |
|
364 | cohDetection: Whether to user Coherent detection or not | |
| 355 | cohDet_timeStep: Coherent Detection calculation time step |
|
365 | cohDet_timeStep: Coherent Detection calculation time step | |
| 356 | cohDet_thresh: Coherent Detection phase threshold to correct phases |
|
366 | cohDet_thresh: Coherent Detection phase threshold to correct phases | |
| 357 |
|
367 | |||
| 358 | noise_timeStep: Noise calculation time step |
|
368 | noise_timeStep: Noise calculation time step | |
| 359 | noise_multiple: Noise multiple to define signal threshold |
|
369 | noise_multiple: Noise multiple to define signal threshold | |
| 360 |
|
370 | |||
| 361 | multDet_timeLimit: Multiple Detection Removal time limit in seconds |
|
371 | multDet_timeLimit: Multiple Detection Removal time limit in seconds | |
| 362 | multDet_rangeLimit: Multiple Detection Removal range limit in km |
|
372 | multDet_rangeLimit: Multiple Detection Removal range limit in km | |
| 363 |
|
373 | |||
| 364 | phaseThresh: Maximum phase difference between receiver to be consider a meteor |
|
374 | phaseThresh: Maximum phase difference between receiver to be consider a meteor | |
| 365 | SNRThresh: Minimum SNR threshold of the meteor signal to be consider a meteor |
|
375 | SNRThresh: Minimum SNR threshold of the meteor signal to be consider a meteor | |
| 366 |
|
376 | |||
| 367 | hmin: Minimum Height of the meteor to use it in the further wind estimations |
|
377 | hmin: Minimum Height of the meteor to use it in the further wind estimations | |
| 368 | hmax: Maximum Height of the meteor to use it in the further wind estimations |
|
378 | hmax: Maximum Height of the meteor to use it in the further wind estimations | |
| 369 | azimuth: Azimuth angle correction |
|
379 | azimuth: Azimuth angle correction | |
| 370 |
|
380 | |||
| 371 | Affected: |
|
381 | Affected: | |
| 372 | self.dataOut.data_param |
|
382 | self.dataOut.data_param | |
| 373 |
|
383 | |||
| 374 | Rejection Criteria (Errors): |
|
384 | Rejection Criteria (Errors): | |
| 375 | 0: No error; analysis OK |
|
385 | 0: No error; analysis OK | |
| 376 | 1: SNR < SNR threshold |
|
386 | 1: SNR < SNR threshold | |
| 377 | 2: angle of arrival (AOA) ambiguously determined |
|
387 | 2: angle of arrival (AOA) ambiguously determined | |
| 378 | 3: AOA estimate not feasible |
|
388 | 3: AOA estimate not feasible | |
| 379 | 4: Large difference in AOAs obtained from different antenna baselines |
|
389 | 4: Large difference in AOAs obtained from different antenna baselines | |
| 380 | 5: echo at start or end of time series |
|
390 | 5: echo at start or end of time series | |
| 381 | 6: echo less than 5 examples long; too short for analysis |
|
391 | 6: echo less than 5 examples long; too short for analysis | |
| 382 | 7: echo rise exceeds 0.3s |
|
392 | 7: echo rise exceeds 0.3s | |
| 383 | 8: echo decay time less than twice rise time |
|
393 | 8: echo decay time less than twice rise time | |
| 384 | 9: large power level before echo |
|
394 | 9: large power level before echo | |
| 385 | 10: large power level after echo |
|
395 | 10: large power level after echo | |
| 386 | 11: poor fit to amplitude for estimation of decay time |
|
396 | 11: poor fit to amplitude for estimation of decay time | |
| 387 | 12: poor fit to CCF phase variation for estimation of radial drift velocity |
|
397 | 12: poor fit to CCF phase variation for estimation of radial drift velocity | |
| 388 | 13: height unresolvable echo: not valid height within 70 to 110 km |
|
398 | 13: height unresolvable echo: not valid height within 70 to 110 km | |
| 389 | 14: height ambiguous echo: more then one possible height within 70 to 110 km |
|
399 | 14: height ambiguous echo: more then one possible height within 70 to 110 km | |
| 390 | 15: radial drift velocity or projected horizontal velocity exceeds 200 m/s |
|
400 | 15: radial drift velocity or projected horizontal velocity exceeds 200 m/s | |
| 391 | 16: oscilatory echo, indicating event most likely not an underdense echo |
|
401 | 16: oscilatory echo, indicating event most likely not an underdense echo | |
| 392 |
|
402 | |||
| 393 | 17: phase difference in meteor Reestimation |
|
403 | 17: phase difference in meteor Reestimation | |
| 394 |
|
404 | |||
| 395 | Data Storage: |
|
405 | Data Storage: | |
| 396 | Meteors for Wind Estimation (8): |
|
406 | Meteors for Wind Estimation (8): | |
| 397 | Day Hour | Range Height |
|
407 | Day Hour | Range Height | |
| 398 | Azimuth Zenith errorCosDir |
|
408 | Azimuth Zenith errorCosDir | |
| 399 | VelRad errorVelRad |
|
409 | VelRad errorVelRad | |
| 400 | TypeError |
|
410 | TypeError | |
| 401 |
|
411 | |||
| 402 | ''' |
|
412 | ''' | |
| 403 | #Get Beacon signal |
|
413 | #Get Beacon signal | |
| 404 | newheis = numpy.where(self.dataOut.heightList>self.dataOut.radarControllerHeaderObj.Taus[tauindex]) |
|
414 | newheis = numpy.where(self.dataOut.heightList>self.dataOut.radarControllerHeaderObj.Taus[tauindex]) | |
| 405 |
|
415 | |||
| 406 | if hei_ref != None: |
|
416 | if hei_ref != None: | |
| 407 | newheis = numpy.where(self.dataOut.heightList>hei_ref) |
|
417 | newheis = numpy.where(self.dataOut.heightList>hei_ref) | |
| 408 |
|
418 | |||
| 409 | heiRang = self.dataOut.getHeiRange() |
|
419 | heiRang = self.dataOut.getHeiRange() | |
| 410 | #Pairs List |
|
420 | #Pairs List | |
| 411 | pairslist = [] |
|
421 | pairslist = [] | |
| 412 | nChannel = self.dataOut.nChannels |
|
422 | nChannel = self.dataOut.nChannels | |
| 413 | for i in range(nChannel): |
|
423 | for i in range(nChannel): | |
| 414 | if i != centerReceiverIndex: |
|
424 | if i != centerReceiverIndex: | |
| 415 | pairslist.append((centerReceiverIndex,i)) |
|
425 | pairslist.append((centerReceiverIndex,i)) | |
| 416 |
|
426 | |||
| 417 | #****************REMOVING HARDWARE PHASE DIFFERENCES*************** |
|
427 | #****************REMOVING HARDWARE PHASE DIFFERENCES*************** | |
| 418 | # see if the user put in pre defined phase shifts |
|
428 | # see if the user put in pre defined phase shifts | |
| 419 | voltsPShift = self.dataOut.data_pre.copy() |
|
429 | voltsPShift = self.dataOut.data_pre.copy() | |
| 420 |
|
430 | |||
| 421 | if predefinedPhaseShifts != None: |
|
431 | if predefinedPhaseShifts != None: | |
| 422 | hardwarePhaseShifts = numpy.array(predefinedPhaseShifts)*numpy.pi/180 |
|
432 | hardwarePhaseShifts = numpy.array(predefinedPhaseShifts)*numpy.pi/180 | |
| 423 | else: |
|
433 | else: | |
| 424 | #get hardware phase shifts using beacon signal |
|
434 | #get hardware phase shifts using beacon signal | |
| 425 | hardwarePhaseShifts = self.__getHardwarePhaseDiff(self.dataOut.data_pre, pairslist, newheis, 10) |
|
435 | hardwarePhaseShifts = self.__getHardwarePhaseDiff(self.dataOut.data_pre, pairslist, newheis, 10) | |
| 426 | hardwarePhaseShifts = numpy.insert(hardwarePhaseShifts,centerReceiverIndex,0) |
|
436 | hardwarePhaseShifts = numpy.insert(hardwarePhaseShifts,centerReceiverIndex,0) | |
| 427 |
|
437 | |||
| 428 | voltsPShift = numpy.zeros((self.dataOut.data_pre.shape[0],self.dataOut.data_pre.shape[1],self.dataOut.data_pre.shape[2]), dtype = 'complex') |
|
438 | voltsPShift = numpy.zeros((self.dataOut.data_pre.shape[0],self.dataOut.data_pre.shape[1],self.dataOut.data_pre.shape[2]), dtype = 'complex') | |
| 429 | for i in range(self.dataOut.data_pre.shape[0]): |
|
439 | for i in range(self.dataOut.data_pre.shape[0]): | |
| 430 | voltsPShift[i,:,:] = self.__shiftPhase(self.dataOut.data_pre[i,:,:], hardwarePhaseShifts[i]) |
|
440 | voltsPShift[i,:,:] = self.__shiftPhase(self.dataOut.data_pre[i,:,:], hardwarePhaseShifts[i]) | |
| 431 | #******************END OF REMOVING HARDWARE PHASE DIFFERENCES********* |
|
441 | #******************END OF REMOVING HARDWARE PHASE DIFFERENCES********* | |
| 432 |
|
442 | |||
| 433 | #Remove DC |
|
443 | #Remove DC | |
| 434 | voltsDC = numpy.mean(voltsPShift,1) |
|
444 | voltsDC = numpy.mean(voltsPShift,1) | |
| 435 | voltsDC = numpy.mean(voltsDC,1) |
|
445 | voltsDC = numpy.mean(voltsDC,1) | |
| 436 | for i in range(voltsDC.shape[0]): |
|
446 | for i in range(voltsDC.shape[0]): | |
| 437 | voltsPShift[i] = voltsPShift[i] - voltsDC[i] |
|
447 | voltsPShift[i] = voltsPShift[i] - voltsDC[i] | |
| 438 |
|
448 | |||
| 439 | #Don't considerate last heights, theyre used to calculate Hardware Phase Shift |
|
449 | #Don't considerate last heights, theyre used to calculate Hardware Phase Shift | |
| 440 | voltsPShift = voltsPShift[:,:,:newheis[0][0]] |
|
450 | voltsPShift = voltsPShift[:,:,:newheis[0][0]] | |
| 441 |
|
451 | |||
| 442 | #************ FIND POWER OF DATA W/COH OR NON COH DETECTION (3.4) ********** |
|
452 | #************ FIND POWER OF DATA W/COH OR NON COH DETECTION (3.4) ********** | |
| 443 | #Coherent Detection |
|
453 | #Coherent Detection | |
| 444 | if cohDetection: |
|
454 | if cohDetection: | |
| 445 | #use coherent detection to get the net power |
|
455 | #use coherent detection to get the net power | |
| 446 | cohDet_thresh = cohDet_thresh*numpy.pi/180 |
|
456 | cohDet_thresh = cohDet_thresh*numpy.pi/180 | |
| 447 | voltsPShift = self.__coherentDetection(voltsPShift, cohDet_timeStep, self.dataOut.timeInterval, pairslist, cohDet_thresh) |
|
457 | voltsPShift = self.__coherentDetection(voltsPShift, cohDet_timeStep, self.dataOut.timeInterval, pairslist, cohDet_thresh) | |
| 448 |
|
458 | |||
| 449 | #Non-coherent detection! |
|
459 | #Non-coherent detection! | |
| 450 | powerNet = numpy.nansum(numpy.abs(voltsPShift[:,:,:])**2,0) |
|
460 | powerNet = numpy.nansum(numpy.abs(voltsPShift[:,:,:])**2,0) | |
| 451 | #********** END OF COH/NON-COH POWER CALCULATION********************** |
|
461 | #********** END OF COH/NON-COH POWER CALCULATION********************** | |
| 452 |
|
462 | |||
| 453 | #********** FIND THE NOISE LEVEL AND POSSIBLE METEORS **************** |
|
463 | #********** FIND THE NOISE LEVEL AND POSSIBLE METEORS **************** | |
| 454 | #Get noise |
|
464 | #Get noise | |
| 455 | noise, noise1 = self.__getNoise(powerNet, noise_timeStep, self.dataOut.timeInterval) |
|
465 | noise, noise1 = self.__getNoise(powerNet, noise_timeStep, self.dataOut.timeInterval) | |
| 456 | # noise = self.getNoise1(powerNet, noise_timeStep, self.dataOut.timeInterval) |
|
466 | # noise = self.getNoise1(powerNet, noise_timeStep, self.dataOut.timeInterval) | |
| 457 | #Get signal threshold |
|
467 | #Get signal threshold | |
| 458 | signalThresh = noise_multiple*noise |
|
468 | signalThresh = noise_multiple*noise | |
| 459 | #Meteor echoes detection |
|
469 | #Meteor echoes detection | |
| 460 | listMeteors = self.__findMeteors(powerNet, signalThresh) |
|
470 | listMeteors = self.__findMeteors(powerNet, signalThresh) | |
| 461 | #******* END OF NOISE LEVEL AND POSSIBLE METEORS CACULATION ********** |
|
471 | #******* END OF NOISE LEVEL AND POSSIBLE METEORS CACULATION ********** | |
| 462 |
|
472 | |||
| 463 | #************** REMOVE MULTIPLE DETECTIONS (3.5) *************************** |
|
473 | #************** REMOVE MULTIPLE DETECTIONS (3.5) *************************** | |
| 464 | #Parameters |
|
474 | #Parameters | |
| 465 | heiRange = self.dataOut.getHeiRange() |
|
475 | heiRange = self.dataOut.getHeiRange() | |
| 466 | rangeInterval = heiRange[1] - heiRange[0] |
|
476 | rangeInterval = heiRange[1] - heiRange[0] | |
| 467 | rangeLimit = multDet_rangeLimit/rangeInterval |
|
477 | rangeLimit = multDet_rangeLimit/rangeInterval | |
| 468 | timeLimit = multDet_timeLimit/self.dataOut.timeInterval |
|
478 | timeLimit = multDet_timeLimit/self.dataOut.timeInterval | |
| 469 | #Multiple detection removals |
|
479 | #Multiple detection removals | |
| 470 | listMeteors1 = self.__removeMultipleDetections(listMeteors, rangeLimit, timeLimit) |
|
480 | listMeteors1 = self.__removeMultipleDetections(listMeteors, rangeLimit, timeLimit) | |
| 471 | #************ END OF REMOVE MULTIPLE DETECTIONS ********************** |
|
481 | #************ END OF REMOVE MULTIPLE DETECTIONS ********************** | |
| 472 |
|
482 | |||
| 473 | #********************* METEOR REESTIMATION (3.7, 3.8, 3.9, 3.10) ******************** |
|
483 | #********************* METEOR REESTIMATION (3.7, 3.8, 3.9, 3.10) ******************** | |
| 474 | #Parameters |
|
484 | #Parameters | |
| 475 | phaseThresh = phaseThresh*numpy.pi/180 |
|
485 | phaseThresh = phaseThresh*numpy.pi/180 | |
| 476 | thresh = [phaseThresh, noise_multiple, SNRThresh] |
|
486 | thresh = [phaseThresh, noise_multiple, SNRThresh] | |
| 477 | #Meteor reestimation (Errors N 1, 6, 12, 17) |
|
487 | #Meteor reestimation (Errors N 1, 6, 12, 17) | |
| 478 | listMeteors2, listMeteorsPower, listMeteorsVolts = self.__meteorReestimation(listMeteors1, voltsPShift, pairslist, thresh, noise, self.dataOut.timeInterval, self.dataOut.frequency) |
|
488 | listMeteors2, listMeteorsPower, listMeteorsVolts = self.__meteorReestimation(listMeteors1, voltsPShift, pairslist, thresh, noise, self.dataOut.timeInterval, self.dataOut.frequency) | |
| 479 | # listMeteors2, listMeteorsPower, listMeteorsVolts = self.meteorReestimation3(listMeteors2, listMeteorsPower, listMeteorsVolts, voltsPShift, pairslist, thresh, noise) |
|
489 | # listMeteors2, listMeteorsPower, listMeteorsVolts = self.meteorReestimation3(listMeteors2, listMeteorsPower, listMeteorsVolts, voltsPShift, pairslist, thresh, noise) | |
| 480 | #Estimation of decay times (Errors N 7, 8, 11) |
|
490 | #Estimation of decay times (Errors N 7, 8, 11) | |
| 481 | listMeteors3 = self.__estimateDecayTime(listMeteors2, listMeteorsPower, self.dataOut.timeInterval, self.dataOut.frequency) |
|
491 | listMeteors3 = self.__estimateDecayTime(listMeteors2, listMeteorsPower, self.dataOut.timeInterval, self.dataOut.frequency) | |
| 482 | #******************* END OF METEOR REESTIMATION ******************* |
|
492 | #******************* END OF METEOR REESTIMATION ******************* | |
| 483 |
|
493 | |||
| 484 | #********************* METEOR PARAMETERS CALCULATION (3.11, 3.12, 3.13) ************************** |
|
494 | #********************* METEOR PARAMETERS CALCULATION (3.11, 3.12, 3.13) ************************** | |
| 485 | #Calculating Radial Velocity (Error N 15) |
|
495 | #Calculating Radial Velocity (Error N 15) | |
| 486 | radialStdThresh = 10 |
|
496 | radialStdThresh = 10 | |
| 487 | listMeteors4 = self.__getRadialVelocity(listMeteors3, listMeteorsVolts, radialStdThresh, pairslist, self.dataOut.timeInterval) |
|
497 | listMeteors4 = self.__getRadialVelocity(listMeteors3, listMeteorsVolts, radialStdThresh, pairslist, self.dataOut.timeInterval) | |
| 488 |
|
498 | |||
| 489 | if len(listMeteors4) > 0: |
|
499 | if len(listMeteors4) > 0: | |
| 490 | #Setting New Array |
|
500 | #Setting New Array | |
| 491 | date = repr(self.dataOut.datatime) |
|
501 | date = repr(self.dataOut.datatime) | |
| 492 | arrayMeteors4, arrayParameters = self.__setNewArrays(listMeteors4, date, heiRang) |
|
502 | arrayMeteors4, arrayParameters = self.__setNewArrays(listMeteors4, date, heiRang) | |
| 493 |
|
503 | |||
| 494 | #Calculate AOA (Error N 3, 4) |
|
504 | #Calculate AOA (Error N 3, 4) | |
| 495 | #JONES ET AL. 1998 |
|
505 | #JONES ET AL. 1998 | |
| 496 | AOAthresh = numpy.pi/8 |
|
506 | AOAthresh = numpy.pi/8 | |
| 497 | error = arrayParameters[:,-1] |
|
507 | error = arrayParameters[:,-1] | |
| 498 | phases = -arrayMeteors4[:,9:13] |
|
508 | phases = -arrayMeteors4[:,9:13] | |
| 499 | pairsList = [] |
|
509 | pairsList = [] | |
| 500 | pairsList.append((0,3)) |
|
510 | pairsList.append((0,3)) | |
| 501 | pairsList.append((1,2)) |
|
511 | pairsList.append((1,2)) | |
| 502 | arrayParameters[:,4:7], arrayParameters[:,-1] = self.__getAOA(phases, pairsList, error, AOAthresh, azimuth) |
|
512 | arrayParameters[:,4:7], arrayParameters[:,-1] = self.__getAOA(phases, pairsList, error, AOAthresh, azimuth) | |
| 503 |
|
513 | |||
| 504 | #Calculate Heights (Error N 13 and 14) |
|
514 | #Calculate Heights (Error N 13 and 14) | |
| 505 | error = arrayParameters[:,-1] |
|
515 | error = arrayParameters[:,-1] | |
| 506 | Ranges = arrayParameters[:,2] |
|
516 | Ranges = arrayParameters[:,2] | |
| 507 | zenith = arrayParameters[:,5] |
|
517 | zenith = arrayParameters[:,5] | |
| 508 | arrayParameters[:,3], arrayParameters[:,-1] = self.__getHeights(Ranges, zenith, error, hmin, hmax) |
|
518 | arrayParameters[:,3], arrayParameters[:,-1] = self.__getHeights(Ranges, zenith, error, hmin, hmax) | |
| 509 | #********************* END OF PARAMETERS CALCULATION ************************** |
|
519 | #********************* END OF PARAMETERS CALCULATION ************************** | |
| 510 |
|
520 | |||
| 511 | #***************************+ SAVE DATA IN HDF5 FORMAT ********************** |
|
521 | #***************************+ SAVE DATA IN HDF5 FORMAT ********************** | |
| 512 | self.dataOut.data_param = arrayParameters |
|
522 | self.dataOut.data_param = arrayParameters | |
| 513 |
|
523 | |||
| 514 | return |
|
524 | return | |
| 515 |
|
525 | |||
| 516 | def __getHardwarePhaseDiff(self, voltage0, pairslist, newheis, n): |
|
526 | def __getHardwarePhaseDiff(self, voltage0, pairslist, newheis, n): | |
| 517 |
|
527 | |||
| 518 | minIndex = min(newheis[0]) |
|
528 | minIndex = min(newheis[0]) | |
| 519 | maxIndex = max(newheis[0]) |
|
529 | maxIndex = max(newheis[0]) | |
| 520 |
|
530 | |||
| 521 | voltage = voltage0[:,:,minIndex:maxIndex+1] |
|
531 | voltage = voltage0[:,:,minIndex:maxIndex+1] | |
| 522 | nLength = voltage.shape[1]/n |
|
532 | nLength = voltage.shape[1]/n | |
| 523 | nMin = 0 |
|
533 | nMin = 0 | |
| 524 | nMax = 0 |
|
534 | nMax = 0 | |
| 525 | phaseOffset = numpy.zeros((len(pairslist),n)) |
|
535 | phaseOffset = numpy.zeros((len(pairslist),n)) | |
| 526 |
|
536 | |||
| 527 | for i in range(n): |
|
537 | for i in range(n): | |
| 528 | nMax += nLength |
|
538 | nMax += nLength | |
| 529 | phaseCCF = -numpy.angle(self.__calculateCCF(voltage[:,nMin:nMax,:], pairslist, [0])) |
|
539 | phaseCCF = -numpy.angle(self.__calculateCCF(voltage[:,nMin:nMax,:], pairslist, [0])) | |
| 530 | phaseCCF = numpy.mean(phaseCCF, axis = 2) |
|
540 | phaseCCF = numpy.mean(phaseCCF, axis = 2) | |
| 531 | phaseOffset[:,i] = phaseCCF.transpose() |
|
541 | phaseOffset[:,i] = phaseCCF.transpose() | |
| 532 | nMin = nMax |
|
542 | nMin = nMax | |
| 533 | # phaseDiff, phaseArrival = self.estimatePhaseDifference(voltage, pairslist) |
|
543 | # phaseDiff, phaseArrival = self.estimatePhaseDifference(voltage, pairslist) | |
| 534 |
|
544 | |||
| 535 | #Remove Outliers |
|
545 | #Remove Outliers | |
| 536 | factor = 2 |
|
546 | factor = 2 | |
| 537 | wt = phaseOffset - signal.medfilt(phaseOffset,(1,5)) |
|
547 | wt = phaseOffset - signal.medfilt(phaseOffset,(1,5)) | |
| 538 | dw = numpy.std(wt,axis = 1) |
|
548 | dw = numpy.std(wt,axis = 1) | |
| 539 | dw = dw.reshape((dw.size,1)) |
|
549 | dw = dw.reshape((dw.size,1)) | |
| 540 | ind = numpy.where(numpy.logical_or(wt>dw*factor,wt<-dw*factor)) |
|
550 | ind = numpy.where(numpy.logical_or(wt>dw*factor,wt<-dw*factor)) | |
| 541 | phaseOffset[ind] = numpy.nan |
|
551 | phaseOffset[ind] = numpy.nan | |
| 542 | phaseOffset = stats.nanmean(phaseOffset, axis=1) |
|
552 | phaseOffset = stats.nanmean(phaseOffset, axis=1) | |
| 543 |
|
553 | |||
| 544 | return phaseOffset |
|
554 | return phaseOffset | |
| 545 |
|
555 | |||
| 546 | def __shiftPhase(self, data, phaseShift): |
|
556 | def __shiftPhase(self, data, phaseShift): | |
| 547 | #this will shift the phase of a complex number |
|
557 | #this will shift the phase of a complex number | |
| 548 | dataShifted = numpy.abs(data) * numpy.exp((numpy.angle(data)+phaseShift)*1j) |
|
558 | dataShifted = numpy.abs(data) * numpy.exp((numpy.angle(data)+phaseShift)*1j) | |
| 549 | return dataShifted |
|
559 | return dataShifted | |
| 550 |
|
560 | |||
| 551 | def __estimatePhaseDifference(self, array, pairslist): |
|
561 | def __estimatePhaseDifference(self, array, pairslist): | |
| 552 | nChannel = array.shape[0] |
|
562 | nChannel = array.shape[0] | |
| 553 | nHeights = array.shape[2] |
|
563 | nHeights = array.shape[2] | |
| 554 | numPairs = len(pairslist) |
|
564 | numPairs = len(pairslist) | |
| 555 | # phaseCCF = numpy.zeros((nChannel, 5, nHeights)) |
|
565 | # phaseCCF = numpy.zeros((nChannel, 5, nHeights)) | |
| 556 | phaseCCF = numpy.angle(self.__calculateCCF(array, pairslist, [-2,-1,0,1,2])) |
|
566 | phaseCCF = numpy.angle(self.__calculateCCF(array, pairslist, [-2,-1,0,1,2])) | |
| 557 |
|
567 | |||
| 558 | #Correct phases |
|
568 | #Correct phases | |
| 559 | derPhaseCCF = phaseCCF[:,1:,:] - phaseCCF[:,0:-1,:] |
|
569 | derPhaseCCF = phaseCCF[:,1:,:] - phaseCCF[:,0:-1,:] | |
| 560 | indDer = numpy.where(numpy.abs(derPhaseCCF) > numpy.pi) |
|
570 | indDer = numpy.where(numpy.abs(derPhaseCCF) > numpy.pi) | |
| 561 |
|
571 | |||
| 562 | if indDer[0].shape[0] > 0: |
|
572 | if indDer[0].shape[0] > 0: | |
| 563 | for i in range(indDer[0].shape[0]): |
|
573 | for i in range(indDer[0].shape[0]): | |
| 564 | signo = -numpy.sign(derPhaseCCF[indDer[0][i],indDer[1][i],indDer[2][i]]) |
|
574 | signo = -numpy.sign(derPhaseCCF[indDer[0][i],indDer[1][i],indDer[2][i]]) | |
| 565 | phaseCCF[indDer[0][i],indDer[1][i]+1:,:] += signo*2*numpy.pi |
|
575 | phaseCCF[indDer[0][i],indDer[1][i]+1:,:] += signo*2*numpy.pi | |
| 566 |
|
576 | |||
| 567 | # for j in range(numSides): |
|
577 | # for j in range(numSides): | |
| 568 | # phaseCCFAux = self.calculateCCF(arrayCenter, arraySides[j,:,:], [-2,1,0,1,2]) |
|
578 | # phaseCCFAux = self.calculateCCF(arrayCenter, arraySides[j,:,:], [-2,1,0,1,2]) | |
| 569 | # phaseCCF[j,:,:] = numpy.angle(phaseCCFAux) |
|
579 | # phaseCCF[j,:,:] = numpy.angle(phaseCCFAux) | |
| 570 | # |
|
580 | # | |
| 571 | #Linear |
|
581 | #Linear | |
| 572 | phaseInt = numpy.zeros((numPairs,1)) |
|
582 | phaseInt = numpy.zeros((numPairs,1)) | |
| 573 | angAllCCF = phaseCCF[:,[0,1,3,4],0] |
|
583 | angAllCCF = phaseCCF[:,[0,1,3,4],0] | |
| 574 | for j in range(numPairs): |
|
584 | for j in range(numPairs): | |
| 575 | fit = stats.linregress([-2,-1,1,2],angAllCCF[j,:]) |
|
585 | fit = stats.linregress([-2,-1,1,2],angAllCCF[j,:]) | |
| 576 | phaseInt[j] = fit[1] |
|
586 | phaseInt[j] = fit[1] | |
| 577 | #Phase Differences |
|
587 | #Phase Differences | |
| 578 | phaseDiff = phaseInt - phaseCCF[:,2,:] |
|
588 | phaseDiff = phaseInt - phaseCCF[:,2,:] | |
| 579 | phaseArrival = phaseInt.reshape(phaseInt.size) |
|
589 | phaseArrival = phaseInt.reshape(phaseInt.size) | |
| 580 |
|
590 | |||
| 581 | #Dealias |
|
591 | #Dealias | |
| 582 | indAlias = numpy.where(phaseArrival > numpy.pi) |
|
592 | indAlias = numpy.where(phaseArrival > numpy.pi) | |
| 583 | phaseArrival[indAlias] -= 2*numpy.pi |
|
593 | phaseArrival[indAlias] -= 2*numpy.pi | |
| 584 | indAlias = numpy.where(phaseArrival < -numpy.pi) |
|
594 | indAlias = numpy.where(phaseArrival < -numpy.pi) | |
| 585 | phaseArrival[indAlias] += 2*numpy.pi |
|
595 | phaseArrival[indAlias] += 2*numpy.pi | |
| 586 |
|
596 | |||
| 587 | return phaseDiff, phaseArrival |
|
597 | return phaseDiff, phaseArrival | |
| 588 |
|
598 | |||
| 589 | def __coherentDetection(self, volts, timeSegment, timeInterval, pairslist, thresh): |
|
599 | def __coherentDetection(self, volts, timeSegment, timeInterval, pairslist, thresh): | |
| 590 | #this function will run the coherent detection used in Holdworth et al. 2004 and return the net power |
|
600 | #this function will run the coherent detection used in Holdworth et al. 2004 and return the net power | |
| 591 | #find the phase shifts of each channel over 1 second intervals |
|
601 | #find the phase shifts of each channel over 1 second intervals | |
| 592 | #only look at ranges below the beacon signal |
|
602 | #only look at ranges below the beacon signal | |
| 593 | numProfPerBlock = numpy.ceil(timeSegment/timeInterval) |
|
603 | numProfPerBlock = numpy.ceil(timeSegment/timeInterval) | |
| 594 | numBlocks = int(volts.shape[1]/numProfPerBlock) |
|
604 | numBlocks = int(volts.shape[1]/numProfPerBlock) | |
| 595 | numHeights = volts.shape[2] |
|
605 | numHeights = volts.shape[2] | |
| 596 | nChannel = volts.shape[0] |
|
606 | nChannel = volts.shape[0] | |
| 597 | voltsCohDet = volts.copy() |
|
607 | voltsCohDet = volts.copy() | |
| 598 |
|
608 | |||
| 599 | pairsarray = numpy.array(pairslist) |
|
609 | pairsarray = numpy.array(pairslist) | |
| 600 | indSides = pairsarray[:,1] |
|
610 | indSides = pairsarray[:,1] | |
| 601 | # indSides = numpy.array(range(nChannel)) |
|
611 | # indSides = numpy.array(range(nChannel)) | |
| 602 | # indSides = numpy.delete(indSides, indCenter) |
|
612 | # indSides = numpy.delete(indSides, indCenter) | |
| 603 | # |
|
613 | # | |
| 604 | # listCenter = numpy.array_split(volts[indCenter,:,:], numBlocks, 0) |
|
614 | # listCenter = numpy.array_split(volts[indCenter,:,:], numBlocks, 0) | |
| 605 | listBlocks = numpy.array_split(volts, numBlocks, 1) |
|
615 | listBlocks = numpy.array_split(volts, numBlocks, 1) | |
| 606 |
|
616 | |||
| 607 | startInd = 0 |
|
617 | startInd = 0 | |
| 608 | endInd = 0 |
|
618 | endInd = 0 | |
| 609 |
|
619 | |||
| 610 | for i in range(numBlocks): |
|
620 | for i in range(numBlocks): | |
| 611 | startInd = endInd |
|
621 | startInd = endInd | |
| 612 | endInd = endInd + listBlocks[i].shape[1] |
|
622 | endInd = endInd + listBlocks[i].shape[1] | |
| 613 |
|
623 | |||
| 614 | arrayBlock = listBlocks[i] |
|
624 | arrayBlock = listBlocks[i] | |
| 615 | # arrayBlockCenter = listCenter[i] |
|
625 | # arrayBlockCenter = listCenter[i] | |
| 616 |
|
626 | |||
| 617 | #Estimate the Phase Difference |
|
627 | #Estimate the Phase Difference | |
| 618 | phaseDiff, aux = self.__estimatePhaseDifference(arrayBlock, pairslist) |
|
628 | phaseDiff, aux = self.__estimatePhaseDifference(arrayBlock, pairslist) | |
| 619 | #Phase Difference RMS |
|
629 | #Phase Difference RMS | |
| 620 | arrayPhaseRMS = numpy.abs(phaseDiff) |
|
630 | arrayPhaseRMS = numpy.abs(phaseDiff) | |
| 621 | phaseRMSaux = numpy.sum(arrayPhaseRMS < thresh,0) |
|
631 | phaseRMSaux = numpy.sum(arrayPhaseRMS < thresh,0) | |
| 622 | indPhase = numpy.where(phaseRMSaux==4) |
|
632 | indPhase = numpy.where(phaseRMSaux==4) | |
| 623 | #Shifting |
|
633 | #Shifting | |
| 624 | if indPhase[0].shape[0] > 0: |
|
634 | if indPhase[0].shape[0] > 0: | |
| 625 | for j in range(indSides.size): |
|
635 | for j in range(indSides.size): | |
| 626 | arrayBlock[indSides[j],:,indPhase] = self.__shiftPhase(arrayBlock[indSides[j],:,indPhase], phaseDiff[j,indPhase].transpose()) |
|
636 | arrayBlock[indSides[j],:,indPhase] = self.__shiftPhase(arrayBlock[indSides[j],:,indPhase], phaseDiff[j,indPhase].transpose()) | |
| 627 | voltsCohDet[:,startInd:endInd,:] = arrayBlock |
|
637 | voltsCohDet[:,startInd:endInd,:] = arrayBlock | |
| 628 |
|
638 | |||
| 629 | return voltsCohDet |
|
639 | return voltsCohDet | |
| 630 |
|
640 | |||
| 631 | def __calculateCCF(self, volts, pairslist ,laglist): |
|
641 | def __calculateCCF(self, volts, pairslist ,laglist): | |
| 632 |
|
642 | |||
| 633 | nHeights = volts.shape[2] |
|
643 | nHeights = volts.shape[2] | |
| 634 | nPoints = volts.shape[1] |
|
644 | nPoints = volts.shape[1] | |
| 635 | voltsCCF = numpy.zeros((len(pairslist), len(laglist), nHeights),dtype = 'complex') |
|
645 | voltsCCF = numpy.zeros((len(pairslist), len(laglist), nHeights),dtype = 'complex') | |
| 636 |
|
646 | |||
| 637 | for i in range(len(pairslist)): |
|
647 | for i in range(len(pairslist)): | |
| 638 | volts1 = volts[pairslist[i][0]] |
|
648 | volts1 = volts[pairslist[i][0]] | |
| 639 | volts2 = volts[pairslist[i][1]] |
|
649 | volts2 = volts[pairslist[i][1]] | |
| 640 |
|
650 | |||
| 641 | for t in range(len(laglist)): |
|
651 | for t in range(len(laglist)): | |
| 642 | idxT = laglist[t] |
|
652 | idxT = laglist[t] | |
| 643 | if idxT >= 0: |
|
653 | if idxT >= 0: | |
| 644 | vStacked = numpy.vstack((volts2[idxT:,:], |
|
654 | vStacked = numpy.vstack((volts2[idxT:,:], | |
| 645 | numpy.zeros((idxT, nHeights),dtype='complex'))) |
|
655 | numpy.zeros((idxT, nHeights),dtype='complex'))) | |
| 646 | else: |
|
656 | else: | |
| 647 | vStacked = numpy.vstack((numpy.zeros((-idxT, nHeights),dtype='complex'), |
|
657 | vStacked = numpy.vstack((numpy.zeros((-idxT, nHeights),dtype='complex'), | |
| 648 | volts2[:(nPoints + idxT),:])) |
|
658 | volts2[:(nPoints + idxT),:])) | |
| 649 | voltsCCF[i,t,:] = numpy.sum((numpy.conjugate(volts1)*vStacked),axis=0) |
|
659 | voltsCCF[i,t,:] = numpy.sum((numpy.conjugate(volts1)*vStacked),axis=0) | |
| 650 |
|
660 | |||
| 651 | vStacked = None |
|
661 | vStacked = None | |
| 652 | return voltsCCF |
|
662 | return voltsCCF | |
| 653 |
|
663 | |||
| 654 | def __getNoise(self, power, timeSegment, timeInterval): |
|
664 | def __getNoise(self, power, timeSegment, timeInterval): | |
| 655 | numProfPerBlock = numpy.ceil(timeSegment/timeInterval) |
|
665 | numProfPerBlock = numpy.ceil(timeSegment/timeInterval) | |
| 656 | numBlocks = int(power.shape[0]/numProfPerBlock) |
|
666 | numBlocks = int(power.shape[0]/numProfPerBlock) | |
| 657 | numHeights = power.shape[1] |
|
667 | numHeights = power.shape[1] | |
| 658 |
|
668 | |||
| 659 | listPower = numpy.array_split(power, numBlocks, 0) |
|
669 | listPower = numpy.array_split(power, numBlocks, 0) | |
| 660 | noise = numpy.zeros((power.shape[0], power.shape[1])) |
|
670 | noise = numpy.zeros((power.shape[0], power.shape[1])) | |
| 661 | noise1 = numpy.zeros((power.shape[0], power.shape[1])) |
|
671 | noise1 = numpy.zeros((power.shape[0], power.shape[1])) | |
| 662 |
|
672 | |||
| 663 | startInd = 0 |
|
673 | startInd = 0 | |
| 664 | endInd = 0 |
|
674 | endInd = 0 | |
| 665 |
|
675 | |||
| 666 | for i in range(numBlocks): #split por canal |
|
676 | for i in range(numBlocks): #split por canal | |
| 667 | startInd = endInd |
|
677 | startInd = endInd | |
| 668 | endInd = endInd + listPower[i].shape[0] |
|
678 | endInd = endInd + listPower[i].shape[0] | |
| 669 |
|
679 | |||
| 670 | arrayBlock = listPower[i] |
|
680 | arrayBlock = listPower[i] | |
| 671 | noiseAux = numpy.mean(arrayBlock, 0) |
|
681 | noiseAux = numpy.mean(arrayBlock, 0) | |
| 672 | # noiseAux = numpy.median(noiseAux) |
|
682 | # noiseAux = numpy.median(noiseAux) | |
| 673 | # noiseAux = numpy.mean(arrayBlock) |
|
683 | # noiseAux = numpy.mean(arrayBlock) | |
| 674 | noise[startInd:endInd,:] = noise[startInd:endInd,:] + noiseAux |
|
684 | noise[startInd:endInd,:] = noise[startInd:endInd,:] + noiseAux | |
| 675 |
|
685 | |||
| 676 | noiseAux1 = numpy.mean(arrayBlock) |
|
686 | noiseAux1 = numpy.mean(arrayBlock) | |
| 677 | noise1[startInd:endInd,:] = noise1[startInd:endInd,:] + noiseAux1 |
|
687 | noise1[startInd:endInd,:] = noise1[startInd:endInd,:] + noiseAux1 | |
| 678 |
|
688 | |||
| 679 | return noise, noise1 |
|
689 | return noise, noise1 | |
| 680 |
|
690 | |||
| 681 | def __findMeteors(self, power, thresh): |
|
691 | def __findMeteors(self, power, thresh): | |
| 682 | nProf = power.shape[0] |
|
692 | nProf = power.shape[0] | |
| 683 | nHeights = power.shape[1] |
|
693 | nHeights = power.shape[1] | |
| 684 | listMeteors = [] |
|
694 | listMeteors = [] | |
| 685 |
|
695 | |||
| 686 | for i in range(nHeights): |
|
696 | for i in range(nHeights): | |
| 687 | powerAux = power[:,i] |
|
697 | powerAux = power[:,i] | |
| 688 | threshAux = thresh[:,i] |
|
698 | threshAux = thresh[:,i] | |
| 689 |
|
699 | |||
| 690 | indUPthresh = numpy.where(powerAux > threshAux)[0] |
|
700 | indUPthresh = numpy.where(powerAux > threshAux)[0] | |
| 691 | indDNthresh = numpy.where(powerAux <= threshAux)[0] |
|
701 | indDNthresh = numpy.where(powerAux <= threshAux)[0] | |
| 692 |
|
702 | |||
| 693 | j = 0 |
|
703 | j = 0 | |
| 694 |
|
704 | |||
| 695 | while (j < indUPthresh.size - 2): |
|
705 | while (j < indUPthresh.size - 2): | |
| 696 | if (indUPthresh[j + 2] == indUPthresh[j] + 2): |
|
706 | if (indUPthresh[j + 2] == indUPthresh[j] + 2): | |
| 697 | indDNAux = numpy.where(indDNthresh > indUPthresh[j]) |
|
707 | indDNAux = numpy.where(indDNthresh > indUPthresh[j]) | |
| 698 | indDNthresh = indDNthresh[indDNAux] |
|
708 | indDNthresh = indDNthresh[indDNAux] | |
| 699 |
|
709 | |||
| 700 | if (indDNthresh.size > 0): |
|
710 | if (indDNthresh.size > 0): | |
| 701 | indEnd = indDNthresh[0] - 1 |
|
711 | indEnd = indDNthresh[0] - 1 | |
| 702 | indInit = indUPthresh[j] |
|
712 | indInit = indUPthresh[j] | |
| 703 |
|
713 | |||
| 704 | meteor = powerAux[indInit:indEnd + 1] |
|
714 | meteor = powerAux[indInit:indEnd + 1] | |
| 705 | indPeak = meteor.argmax() + indInit |
|
715 | indPeak = meteor.argmax() + indInit | |
| 706 | FLA = sum(numpy.conj(meteor)*numpy.hstack((meteor[1:],0))) |
|
716 | FLA = sum(numpy.conj(meteor)*numpy.hstack((meteor[1:],0))) | |
| 707 |
|
717 | |||
| 708 | listMeteors.append(numpy.array([i,indInit,indPeak,indEnd,FLA])) #CHEQUEAR!!!!! |
|
718 | listMeteors.append(numpy.array([i,indInit,indPeak,indEnd,FLA])) #CHEQUEAR!!!!! | |
| 709 | j = numpy.where(indUPthresh == indEnd)[0] + 1 |
|
719 | j = numpy.where(indUPthresh == indEnd)[0] + 1 | |
| 710 | else: j+=1 |
|
720 | else: j+=1 | |
| 711 | else: j+=1 |
|
721 | else: j+=1 | |
| 712 |
|
722 | |||
| 713 | return listMeteors |
|
723 | return listMeteors | |
| 714 |
|
724 | |||
| 715 | def __removeMultipleDetections(self,listMeteors, rangeLimit, timeLimit): |
|
725 | def __removeMultipleDetections(self,listMeteors, rangeLimit, timeLimit): | |
| 716 |
|
726 | |||
| 717 | arrayMeteors = numpy.asarray(listMeteors) |
|
727 | arrayMeteors = numpy.asarray(listMeteors) | |
| 718 | listMeteors1 = [] |
|
728 | listMeteors1 = [] | |
| 719 |
|
729 | |||
| 720 | while arrayMeteors.shape[0] > 0: |
|
730 | while arrayMeteors.shape[0] > 0: | |
| 721 | FLAs = arrayMeteors[:,4] |
|
731 | FLAs = arrayMeteors[:,4] | |
| 722 | maxFLA = FLAs.argmax() |
|
732 | maxFLA = FLAs.argmax() | |
| 723 | listMeteors1.append(arrayMeteors[maxFLA,:]) |
|
733 | listMeteors1.append(arrayMeteors[maxFLA,:]) | |
| 724 |
|
734 | |||
| 725 | MeteorInitTime = arrayMeteors[maxFLA,1] |
|
735 | MeteorInitTime = arrayMeteors[maxFLA,1] | |
| 726 | MeteorEndTime = arrayMeteors[maxFLA,3] |
|
736 | MeteorEndTime = arrayMeteors[maxFLA,3] | |
| 727 | MeteorHeight = arrayMeteors[maxFLA,0] |
|
737 | MeteorHeight = arrayMeteors[maxFLA,0] | |
| 728 |
|
738 | |||
| 729 | #Check neighborhood |
|
739 | #Check neighborhood | |
| 730 | maxHeightIndex = MeteorHeight + rangeLimit |
|
740 | maxHeightIndex = MeteorHeight + rangeLimit | |
| 731 | minHeightIndex = MeteorHeight - rangeLimit |
|
741 | minHeightIndex = MeteorHeight - rangeLimit | |
| 732 | minTimeIndex = MeteorInitTime - timeLimit |
|
742 | minTimeIndex = MeteorInitTime - timeLimit | |
| 733 | maxTimeIndex = MeteorEndTime + timeLimit |
|
743 | maxTimeIndex = MeteorEndTime + timeLimit | |
| 734 |
|
744 | |||
| 735 | #Check Heights |
|
745 | #Check Heights | |
| 736 | indHeight = numpy.logical_and(arrayMeteors[:,0] >= minHeightIndex, arrayMeteors[:,0] <= maxHeightIndex) |
|
746 | indHeight = numpy.logical_and(arrayMeteors[:,0] >= minHeightIndex, arrayMeteors[:,0] <= maxHeightIndex) | |
| 737 | indTime = numpy.logical_and(arrayMeteors[:,3] >= minTimeIndex, arrayMeteors[:,1] <= maxTimeIndex) |
|
747 | indTime = numpy.logical_and(arrayMeteors[:,3] >= minTimeIndex, arrayMeteors[:,1] <= maxTimeIndex) | |
| 738 | indBoth = numpy.where(numpy.logical_and(indTime,indHeight)) |
|
748 | indBoth = numpy.where(numpy.logical_and(indTime,indHeight)) | |
| 739 |
|
749 | |||
| 740 | arrayMeteors = numpy.delete(arrayMeteors, indBoth, axis = 0) |
|
750 | arrayMeteors = numpy.delete(arrayMeteors, indBoth, axis = 0) | |
| 741 |
|
751 | |||
| 742 | return listMeteors1 |
|
752 | return listMeteors1 | |
| 743 |
|
753 | |||
| 744 | def __meteorReestimation(self, listMeteors, volts, pairslist, thresh, noise, timeInterval,frequency): |
|
754 | def __meteorReestimation(self, listMeteors, volts, pairslist, thresh, noise, timeInterval,frequency): | |
| 745 | numHeights = volts.shape[2] |
|
755 | numHeights = volts.shape[2] | |
| 746 | nChannel = volts.shape[0] |
|
756 | nChannel = volts.shape[0] | |
| 747 |
|
757 | |||
| 748 | thresholdPhase = thresh[0] |
|
758 | thresholdPhase = thresh[0] | |
| 749 | thresholdNoise = thresh[1] |
|
759 | thresholdNoise = thresh[1] | |
| 750 | thresholdDB = float(thresh[2]) |
|
760 | thresholdDB = float(thresh[2]) | |
| 751 |
|
761 | |||
| 752 | thresholdDB1 = 10**(thresholdDB/10) |
|
762 | thresholdDB1 = 10**(thresholdDB/10) | |
| 753 | pairsarray = numpy.array(pairslist) |
|
763 | pairsarray = numpy.array(pairslist) | |
| 754 | indSides = pairsarray[:,1] |
|
764 | indSides = pairsarray[:,1] | |
| 755 |
|
765 | |||
| 756 | pairslist1 = list(pairslist) |
|
766 | pairslist1 = list(pairslist) | |
| 757 | pairslist1.append((0,1)) |
|
767 | pairslist1.append((0,1)) | |
| 758 | pairslist1.append((3,4)) |
|
768 | pairslist1.append((3,4)) | |
| 759 |
|
769 | |||
| 760 | listMeteors1 = [] |
|
770 | listMeteors1 = [] | |
| 761 | listPowerSeries = [] |
|
771 | listPowerSeries = [] | |
| 762 | listVoltageSeries = [] |
|
772 | listVoltageSeries = [] | |
| 763 | #volts has the war data |
|
773 | #volts has the war data | |
| 764 |
|
774 | |||
| 765 | if frequency == 30e6: |
|
775 | if frequency == 30e6: | |
| 766 | timeLag = 45*10**-3 |
|
776 | timeLag = 45*10**-3 | |
| 767 | else: |
|
777 | else: | |
| 768 | timeLag = 15*10**-3 |
|
778 | timeLag = 15*10**-3 | |
| 769 | lag = numpy.ceil(timeLag/timeInterval) |
|
779 | lag = numpy.ceil(timeLag/timeInterval) | |
| 770 |
|
780 | |||
| 771 | for i in range(len(listMeteors)): |
|
781 | for i in range(len(listMeteors)): | |
| 772 |
|
782 | |||
| 773 | ###################### 3.6 - 3.7 PARAMETERS REESTIMATION ######################### |
|
783 | ###################### 3.6 - 3.7 PARAMETERS REESTIMATION ######################### | |
| 774 | meteorAux = numpy.zeros(16) |
|
784 | meteorAux = numpy.zeros(16) | |
| 775 |
|
785 | |||
| 776 | #Loading meteor Data (mHeight, mStart, mPeak, mEnd) |
|
786 | #Loading meteor Data (mHeight, mStart, mPeak, mEnd) | |
| 777 | mHeight = listMeteors[i][0] |
|
787 | mHeight = listMeteors[i][0] | |
| 778 | mStart = listMeteors[i][1] |
|
788 | mStart = listMeteors[i][1] | |
| 779 | mPeak = listMeteors[i][2] |
|
789 | mPeak = listMeteors[i][2] | |
| 780 | mEnd = listMeteors[i][3] |
|
790 | mEnd = listMeteors[i][3] | |
| 781 |
|
791 | |||
| 782 | #get the volt data between the start and end times of the meteor |
|
792 | #get the volt data between the start and end times of the meteor | |
| 783 | meteorVolts = volts[:,mStart:mEnd+1,mHeight] |
|
793 | meteorVolts = volts[:,mStart:mEnd+1,mHeight] | |
| 784 | meteorVolts = meteorVolts.reshape(meteorVolts.shape[0], meteorVolts.shape[1], 1) |
|
794 | meteorVolts = meteorVolts.reshape(meteorVolts.shape[0], meteorVolts.shape[1], 1) | |
| 785 |
|
795 | |||
| 786 | #3.6. Phase Difference estimation |
|
796 | #3.6. Phase Difference estimation | |
| 787 | phaseDiff, aux = self.__estimatePhaseDifference(meteorVolts, pairslist) |
|
797 | phaseDiff, aux = self.__estimatePhaseDifference(meteorVolts, pairslist) | |
| 788 |
|
798 | |||
| 789 | #3.7. Phase difference removal & meteor start, peak and end times reestimated |
|
799 | #3.7. Phase difference removal & meteor start, peak and end times reestimated | |
| 790 | #meteorVolts0.- all Channels, all Profiles |
|
800 | #meteorVolts0.- all Channels, all Profiles | |
| 791 | meteorVolts0 = volts[:,:,mHeight] |
|
801 | meteorVolts0 = volts[:,:,mHeight] | |
| 792 | meteorThresh = noise[:,mHeight]*thresholdNoise |
|
802 | meteorThresh = noise[:,mHeight]*thresholdNoise | |
| 793 | meteorNoise = noise[:,mHeight] |
|
803 | meteorNoise = noise[:,mHeight] | |
| 794 | meteorVolts0[indSides,:] = self.__shiftPhase(meteorVolts0[indSides,:], phaseDiff) #Phase Shifting |
|
804 | meteorVolts0[indSides,:] = self.__shiftPhase(meteorVolts0[indSides,:], phaseDiff) #Phase Shifting | |
| 795 | powerNet0 = numpy.nansum(numpy.abs(meteorVolts0)**2, axis = 0) #Power |
|
805 | powerNet0 = numpy.nansum(numpy.abs(meteorVolts0)**2, axis = 0) #Power | |
| 796 |
|
806 | |||
| 797 | #Times reestimation |
|
807 | #Times reestimation | |
| 798 | mStart1 = numpy.where(powerNet0[:mPeak] < meteorThresh[:mPeak])[0] |
|
808 | mStart1 = numpy.where(powerNet0[:mPeak] < meteorThresh[:mPeak])[0] | |
| 799 | if mStart1.size > 0: |
|
809 | if mStart1.size > 0: | |
| 800 | mStart1 = mStart1[-1] + 1 |
|
810 | mStart1 = mStart1[-1] + 1 | |
| 801 |
|
811 | |||
| 802 | else: |
|
812 | else: | |
| 803 | mStart1 = mPeak |
|
813 | mStart1 = mPeak | |
| 804 |
|
814 | |||
| 805 | mEnd1 = numpy.where(powerNet0[mPeak:] < meteorThresh[mPeak:])[0][0] + mPeak - 1 |
|
815 | mEnd1 = numpy.where(powerNet0[mPeak:] < meteorThresh[mPeak:])[0][0] + mPeak - 1 | |
| 806 | mEndDecayTime1 = numpy.where(powerNet0[mPeak:] < meteorNoise[mPeak:])[0] |
|
816 | mEndDecayTime1 = numpy.where(powerNet0[mPeak:] < meteorNoise[mPeak:])[0] | |
| 807 | if mEndDecayTime1.size == 0: |
|
817 | if mEndDecayTime1.size == 0: | |
| 808 | mEndDecayTime1 = powerNet0.size |
|
818 | mEndDecayTime1 = powerNet0.size | |
| 809 | else: |
|
819 | else: | |
| 810 | mEndDecayTime1 = mEndDecayTime1[0] + mPeak - 1 |
|
820 | mEndDecayTime1 = mEndDecayTime1[0] + mPeak - 1 | |
| 811 | # mPeak1 = meteorVolts0[mStart1:mEnd1 + 1].argmax() |
|
821 | # mPeak1 = meteorVolts0[mStart1:mEnd1 + 1].argmax() | |
| 812 |
|
822 | |||
| 813 | #meteorVolts1.- all Channels, from start to end |
|
823 | #meteorVolts1.- all Channels, from start to end | |
| 814 | meteorVolts1 = meteorVolts0[:,mStart1:mEnd1 + 1] |
|
824 | meteorVolts1 = meteorVolts0[:,mStart1:mEnd1 + 1] | |
| 815 | meteorVolts2 = meteorVolts0[:,mPeak + lag:mEnd1 + 1] |
|
825 | meteorVolts2 = meteorVolts0[:,mPeak + lag:mEnd1 + 1] | |
| 816 | if meteorVolts2.shape[1] == 0: |
|
826 | if meteorVolts2.shape[1] == 0: | |
| 817 | meteorVolts2 = meteorVolts0[:,mPeak:mEnd1 + 1] |
|
827 | meteorVolts2 = meteorVolts0[:,mPeak:mEnd1 + 1] | |
| 818 | meteorVolts1 = meteorVolts1.reshape(meteorVolts1.shape[0], meteorVolts1.shape[1], 1) |
|
828 | meteorVolts1 = meteorVolts1.reshape(meteorVolts1.shape[0], meteorVolts1.shape[1], 1) | |
| 819 | meteorVolts2 = meteorVolts2.reshape(meteorVolts2.shape[0], meteorVolts2.shape[1], 1) |
|
829 | meteorVolts2 = meteorVolts2.reshape(meteorVolts2.shape[0], meteorVolts2.shape[1], 1) | |
| 820 | ##################### END PARAMETERS REESTIMATION ######################### |
|
830 | ##################### END PARAMETERS REESTIMATION ######################### | |
| 821 |
|
831 | |||
| 822 | ##################### 3.8 PHASE DIFFERENCE REESTIMATION ######################## |
|
832 | ##################### 3.8 PHASE DIFFERENCE REESTIMATION ######################## | |
| 823 | # if mEnd1 - mStart1 > 4: #Error Number 6: echo less than 5 samples long; too short for analysis |
|
833 | # if mEnd1 - mStart1 > 4: #Error Number 6: echo less than 5 samples long; too short for analysis | |
| 824 | if meteorVolts2.shape[1] > 0: |
|
834 | if meteorVolts2.shape[1] > 0: | |
| 825 | #Phase Difference re-estimation |
|
835 | #Phase Difference re-estimation | |
| 826 | phaseDiff1, phaseDiffint = self.__estimatePhaseDifference(meteorVolts2, pairslist1) #Phase Difference Estimation |
|
836 | phaseDiff1, phaseDiffint = self.__estimatePhaseDifference(meteorVolts2, pairslist1) #Phase Difference Estimation | |
| 827 | # phaseDiff1, phaseDiffint = self.estimatePhaseDifference(meteorVolts2, pairslist) |
|
837 | # phaseDiff1, phaseDiffint = self.estimatePhaseDifference(meteorVolts2, pairslist) | |
| 828 | meteorVolts2 = meteorVolts2.reshape(meteorVolts2.shape[0], meteorVolts2.shape[1]) |
|
838 | meteorVolts2 = meteorVolts2.reshape(meteorVolts2.shape[0], meteorVolts2.shape[1]) | |
| 829 | phaseDiff11 = numpy.reshape(phaseDiff1, (phaseDiff1.shape[0],1)) |
|
839 | phaseDiff11 = numpy.reshape(phaseDiff1, (phaseDiff1.shape[0],1)) | |
| 830 | meteorVolts2[indSides,:] = self.__shiftPhase(meteorVolts2[indSides,:], phaseDiff11[0:4]) #Phase Shifting |
|
840 | meteorVolts2[indSides,:] = self.__shiftPhase(meteorVolts2[indSides,:], phaseDiff11[0:4]) #Phase Shifting | |
| 831 |
|
841 | |||
| 832 | #Phase Difference RMS |
|
842 | #Phase Difference RMS | |
| 833 | phaseRMS1 = numpy.sqrt(numpy.mean(numpy.square(phaseDiff1))) |
|
843 | phaseRMS1 = numpy.sqrt(numpy.mean(numpy.square(phaseDiff1))) | |
| 834 | powerNet1 = numpy.nansum(numpy.abs(meteorVolts1[:,:])**2,0) |
|
844 | powerNet1 = numpy.nansum(numpy.abs(meteorVolts1[:,:])**2,0) | |
| 835 | #Data from Meteor |
|
845 | #Data from Meteor | |
| 836 | mPeak1 = powerNet1.argmax() + mStart1 |
|
846 | mPeak1 = powerNet1.argmax() + mStart1 | |
| 837 | mPeakPower1 = powerNet1.max() |
|
847 | mPeakPower1 = powerNet1.max() | |
| 838 | noiseAux = sum(noise[mStart1:mEnd1 + 1,mHeight]) |
|
848 | noiseAux = sum(noise[mStart1:mEnd1 + 1,mHeight]) | |
| 839 | mSNR1 = (sum(powerNet1)-noiseAux)/noiseAux |
|
849 | mSNR1 = (sum(powerNet1)-noiseAux)/noiseAux | |
| 840 | Meteor1 = numpy.array([mHeight, mStart1, mPeak1, mEnd1, mPeakPower1, mSNR1, phaseRMS1]) |
|
850 | Meteor1 = numpy.array([mHeight, mStart1, mPeak1, mEnd1, mPeakPower1, mSNR1, phaseRMS1]) | |
| 841 | Meteor1 = numpy.hstack((Meteor1,phaseDiffint)) |
|
851 | Meteor1 = numpy.hstack((Meteor1,phaseDiffint)) | |
| 842 | PowerSeries = powerNet0[mStart1:mEndDecayTime1 + 1] |
|
852 | PowerSeries = powerNet0[mStart1:mEndDecayTime1 + 1] | |
| 843 | #Vectorize |
|
853 | #Vectorize | |
| 844 | meteorAux[0:7] = [mHeight, mStart1, mPeak1, mEnd1, mPeakPower1, mSNR1, phaseRMS1] |
|
854 | meteorAux[0:7] = [mHeight, mStart1, mPeak1, mEnd1, mPeakPower1, mSNR1, phaseRMS1] | |
| 845 | meteorAux[7:11] = phaseDiffint[0:4] |
|
855 | meteorAux[7:11] = phaseDiffint[0:4] | |
| 846 |
|
856 | |||
| 847 | #Rejection Criterions |
|
857 | #Rejection Criterions | |
| 848 | if phaseRMS1 > thresholdPhase: #Error Number 17: Phase variation |
|
858 | if phaseRMS1 > thresholdPhase: #Error Number 17: Phase variation | |
| 849 | meteorAux[-1] = 17 |
|
859 | meteorAux[-1] = 17 | |
| 850 | elif mSNR1 < thresholdDB1: #Error Number 1: SNR < threshold dB |
|
860 | elif mSNR1 < thresholdDB1: #Error Number 1: SNR < threshold dB | |
| 851 | meteorAux[-1] = 1 |
|
861 | meteorAux[-1] = 1 | |
| 852 |
|
862 | |||
| 853 |
|
863 | |||
| 854 | else: |
|
864 | else: | |
| 855 | meteorAux[0:4] = [mHeight, mStart, mPeak, mEnd] |
|
865 | meteorAux[0:4] = [mHeight, mStart, mPeak, mEnd] | |
| 856 | meteorAux[-1] = 6 #Error Number 6: echo less than 5 samples long; too short for analysis |
|
866 | meteorAux[-1] = 6 #Error Number 6: echo less than 5 samples long; too short for analysis | |
| 857 | PowerSeries = 0 |
|
867 | PowerSeries = 0 | |
| 858 |
|
868 | |||
| 859 | listMeteors1.append(meteorAux) |
|
869 | listMeteors1.append(meteorAux) | |
| 860 | listPowerSeries.append(PowerSeries) |
|
870 | listPowerSeries.append(PowerSeries) | |
| 861 | listVoltageSeries.append(meteorVolts1) |
|
871 | listVoltageSeries.append(meteorVolts1) | |
| 862 |
|
872 | |||
| 863 | return listMeteors1, listPowerSeries, listVoltageSeries |
|
873 | return listMeteors1, listPowerSeries, listVoltageSeries | |
| 864 |
|
874 | |||
| 865 | def __estimateDecayTime(self, listMeteors, listPower, timeInterval, frequency): |
|
875 | def __estimateDecayTime(self, listMeteors, listPower, timeInterval, frequency): | |
| 866 |
|
876 | |||
| 867 | threshError = 10 |
|
877 | threshError = 10 | |
| 868 | #Depending if it is 30 or 50 MHz |
|
878 | #Depending if it is 30 or 50 MHz | |
| 869 | if frequency == 30e6: |
|
879 | if frequency == 30e6: | |
| 870 | timeLag = 45*10**-3 |
|
880 | timeLag = 45*10**-3 | |
| 871 | else: |
|
881 | else: | |
| 872 | timeLag = 15*10**-3 |
|
882 | timeLag = 15*10**-3 | |
| 873 | lag = numpy.ceil(timeLag/timeInterval) |
|
883 | lag = numpy.ceil(timeLag/timeInterval) | |
| 874 |
|
884 | |||
| 875 | listMeteors1 = [] |
|
885 | listMeteors1 = [] | |
| 876 |
|
886 | |||
| 877 | for i in range(len(listMeteors)): |
|
887 | for i in range(len(listMeteors)): | |
| 878 | meteorPower = listPower[i] |
|
888 | meteorPower = listPower[i] | |
| 879 | meteorAux = listMeteors[i] |
|
889 | meteorAux = listMeteors[i] | |
| 880 |
|
890 | |||
| 881 | if meteorAux[-1] == 0: |
|
891 | if meteorAux[-1] == 0: | |
| 882 |
|
892 | |||
| 883 | try: |
|
893 | try: | |
| 884 | indmax = meteorPower.argmax() |
|
894 | indmax = meteorPower.argmax() | |
| 885 | indlag = indmax + lag |
|
895 | indlag = indmax + lag | |
| 886 |
|
896 | |||
| 887 | y = meteorPower[indlag:] |
|
897 | y = meteorPower[indlag:] | |
| 888 | x = numpy.arange(0, y.size)*timeLag |
|
898 | x = numpy.arange(0, y.size)*timeLag | |
| 889 |
|
899 | |||
| 890 | #first guess |
|
900 | #first guess | |
| 891 | a = y[0] |
|
901 | a = y[0] | |
| 892 | tau = timeLag |
|
902 | tau = timeLag | |
| 893 | #exponential fit |
|
903 | #exponential fit | |
| 894 | popt, pcov = optimize.curve_fit(self.__exponential_function, x, y, p0 = [a, tau]) |
|
904 | popt, pcov = optimize.curve_fit(self.__exponential_function, x, y, p0 = [a, tau]) | |
| 895 | y1 = self.__exponential_function(x, *popt) |
|
905 | y1 = self.__exponential_function(x, *popt) | |
| 896 | #error estimation |
|
906 | #error estimation | |
| 897 | error = sum((y - y1)**2)/(numpy.var(y)*(y.size - popt.size)) |
|
907 | error = sum((y - y1)**2)/(numpy.var(y)*(y.size - popt.size)) | |
| 898 |
|
908 | |||
| 899 | decayTime = popt[1] |
|
909 | decayTime = popt[1] | |
| 900 | riseTime = indmax*timeInterval |
|
910 | riseTime = indmax*timeInterval | |
| 901 | meteorAux[11:13] = [decayTime, error] |
|
911 | meteorAux[11:13] = [decayTime, error] | |
| 902 |
|
912 | |||
| 903 | #Table items 7, 8 and 11 |
|
913 | #Table items 7, 8 and 11 | |
| 904 | if (riseTime > 0.3): #Number 7: Echo rise exceeds 0.3s |
|
914 | if (riseTime > 0.3): #Number 7: Echo rise exceeds 0.3s | |
| 905 | meteorAux[-1] = 7 |
|
915 | meteorAux[-1] = 7 | |
| 906 | elif (decayTime < 2*riseTime) : #Number 8: Echo decay time less than than twice rise time |
|
916 | elif (decayTime < 2*riseTime) : #Number 8: Echo decay time less than than twice rise time | |
| 907 | meteorAux[-1] = 8 |
|
917 | meteorAux[-1] = 8 | |
| 908 | if (error > threshError): #Number 11: Poor fit to amplitude for estimation of decay time |
|
918 | if (error > threshError): #Number 11: Poor fit to amplitude for estimation of decay time | |
| 909 | meteorAux[-1] = 11 |
|
919 | meteorAux[-1] = 11 | |
| 910 |
|
920 | |||
| 911 |
|
921 | |||
| 912 | except: |
|
922 | except: | |
| 913 | meteorAux[-1] = 11 |
|
923 | meteorAux[-1] = 11 | |
| 914 |
|
924 | |||
| 915 |
|
925 | |||
| 916 | listMeteors1.append(meteorAux) |
|
926 | listMeteors1.append(meteorAux) | |
| 917 |
|
927 | |||
| 918 | return listMeteors1 |
|
928 | return listMeteors1 | |
| 919 |
|
929 | |||
| 920 | #Exponential Function |
|
930 | #Exponential Function | |
| 921 |
|
931 | |||
| 922 | def __exponential_function(self, x, a, tau): |
|
932 | def __exponential_function(self, x, a, tau): | |
| 923 | y = a*numpy.exp(-x/tau) |
|
933 | y = a*numpy.exp(-x/tau) | |
| 924 | return y |
|
934 | return y | |
| 925 |
|
935 | |||
| 926 | def __getRadialVelocity(self, listMeteors, listVolts, radialStdThresh, pairslist, timeInterval): |
|
936 | def __getRadialVelocity(self, listMeteors, listVolts, radialStdThresh, pairslist, timeInterval): | |
| 927 |
|
937 | |||
| 928 | pairslist1 = list(pairslist) |
|
938 | pairslist1 = list(pairslist) | |
| 929 | pairslist1.append((0,1)) |
|
939 | pairslist1.append((0,1)) | |
| 930 | pairslist1.append((3,4)) |
|
940 | pairslist1.append((3,4)) | |
| 931 | numPairs = len(pairslist1) |
|
941 | numPairs = len(pairslist1) | |
| 932 | #Time Lag |
|
942 | #Time Lag | |
| 933 | timeLag = 45*10**-3 |
|
943 | timeLag = 45*10**-3 | |
| 934 | c = 3e8 |
|
944 | c = 3e8 | |
| 935 | lag = numpy.ceil(timeLag/timeInterval) |
|
945 | lag = numpy.ceil(timeLag/timeInterval) | |
| 936 | freq = 30e6 |
|
946 | freq = 30e6 | |
| 937 |
|
947 | |||
| 938 | listMeteors1 = [] |
|
948 | listMeteors1 = [] | |
| 939 |
|
949 | |||
| 940 | for i in range(len(listMeteors)): |
|
950 | for i in range(len(listMeteors)): | |
| 941 | meteor = listMeteors[i] |
|
951 | meteor = listMeteors[i] | |
| 942 | meteorAux = numpy.hstack((meteor[:-1], 0, 0, meteor[-1])) |
|
952 | meteorAux = numpy.hstack((meteor[:-1], 0, 0, meteor[-1])) | |
| 943 | if meteor[-1] == 0: |
|
953 | if meteor[-1] == 0: | |
| 944 | mStart = listMeteors[i][1] |
|
954 | mStart = listMeteors[i][1] | |
| 945 | mPeak = listMeteors[i][2] |
|
955 | mPeak = listMeteors[i][2] | |
| 946 | mLag = mPeak - mStart + lag |
|
956 | mLag = mPeak - mStart + lag | |
| 947 |
|
957 | |||
| 948 | #get the volt data between the start and end times of the meteor |
|
958 | #get the volt data between the start and end times of the meteor | |
| 949 | meteorVolts = listVolts[i] |
|
959 | meteorVolts = listVolts[i] | |
| 950 | meteorVolts = meteorVolts.reshape(meteorVolts.shape[0], meteorVolts.shape[1], 1) |
|
960 | meteorVolts = meteorVolts.reshape(meteorVolts.shape[0], meteorVolts.shape[1], 1) | |
| 951 |
|
961 | |||
| 952 | #Get CCF |
|
962 | #Get CCF | |
| 953 | allCCFs = self.__calculateCCF(meteorVolts, pairslist1, [-2,-1,0,1,2]) |
|
963 | allCCFs = self.__calculateCCF(meteorVolts, pairslist1, [-2,-1,0,1,2]) | |
| 954 |
|
964 | |||
| 955 | #Method 2 |
|
965 | #Method 2 | |
| 956 | slopes = numpy.zeros(numPairs) |
|
966 | slopes = numpy.zeros(numPairs) | |
| 957 | time = numpy.array([-2,-1,1,2])*timeInterval |
|
967 | time = numpy.array([-2,-1,1,2])*timeInterval | |
| 958 | angAllCCF = numpy.angle(allCCFs[:,[0,1,3,4],0]) |
|
968 | angAllCCF = numpy.angle(allCCFs[:,[0,1,3,4],0]) | |
| 959 |
|
969 | |||
| 960 | #Correct phases |
|
970 | #Correct phases | |
| 961 | derPhaseCCF = angAllCCF[:,1:] - angAllCCF[:,0:-1] |
|
971 | derPhaseCCF = angAllCCF[:,1:] - angAllCCF[:,0:-1] | |
| 962 | indDer = numpy.where(numpy.abs(derPhaseCCF) > numpy.pi) |
|
972 | indDer = numpy.where(numpy.abs(derPhaseCCF) > numpy.pi) | |
| 963 |
|
973 | |||
| 964 | if indDer[0].shape[0] > 0: |
|
974 | if indDer[0].shape[0] > 0: | |
| 965 | for i in range(indDer[0].shape[0]): |
|
975 | for i in range(indDer[0].shape[0]): | |
| 966 | signo = -numpy.sign(derPhaseCCF[indDer[0][i],indDer[1][i]]) |
|
976 | signo = -numpy.sign(derPhaseCCF[indDer[0][i],indDer[1][i]]) | |
| 967 | angAllCCF[indDer[0][i],indDer[1][i]+1:] += signo*2*numpy.pi |
|
977 | angAllCCF[indDer[0][i],indDer[1][i]+1:] += signo*2*numpy.pi | |
| 968 |
|
978 | |||
| 969 | # fit = scipy.stats.linregress(numpy.array([-2,-1,1,2])*timeInterval, numpy.array([phaseLagN2s[i],phaseLagN1s[i],phaseLag1s[i],phaseLag2s[i]])) |
|
979 | # fit = scipy.stats.linregress(numpy.array([-2,-1,1,2])*timeInterval, numpy.array([phaseLagN2s[i],phaseLagN1s[i],phaseLag1s[i],phaseLag2s[i]])) | |
| 970 | for j in range(numPairs): |
|
980 | for j in range(numPairs): | |
| 971 | fit = stats.linregress(time, angAllCCF[j,:]) |
|
981 | fit = stats.linregress(time, angAllCCF[j,:]) | |
| 972 | slopes[j] = fit[0] |
|
982 | slopes[j] = fit[0] | |
| 973 |
|
983 | |||
| 974 | #Remove Outlier |
|
984 | #Remove Outlier | |
| 975 | # indOut = numpy.argmax(numpy.abs(slopes - numpy.mean(slopes))) |
|
985 | # indOut = numpy.argmax(numpy.abs(slopes - numpy.mean(slopes))) | |
| 976 | # slopes = numpy.delete(slopes,indOut) |
|
986 | # slopes = numpy.delete(slopes,indOut) | |
| 977 | # indOut = numpy.argmax(numpy.abs(slopes - numpy.mean(slopes))) |
|
987 | # indOut = numpy.argmax(numpy.abs(slopes - numpy.mean(slopes))) | |
| 978 | # slopes = numpy.delete(slopes,indOut) |
|
988 | # slopes = numpy.delete(slopes,indOut) | |
| 979 |
|
989 | |||
| 980 | radialVelocity = -numpy.mean(slopes)*(0.25/numpy.pi)*(c/freq) |
|
990 | radialVelocity = -numpy.mean(slopes)*(0.25/numpy.pi)*(c/freq) | |
| 981 | radialError = numpy.std(slopes)*(0.25/numpy.pi)*(c/freq) |
|
991 | radialError = numpy.std(slopes)*(0.25/numpy.pi)*(c/freq) | |
| 982 | meteorAux[-2] = radialError |
|
992 | meteorAux[-2] = radialError | |
| 983 | meteorAux[-3] = radialVelocity |
|
993 | meteorAux[-3] = radialVelocity | |
| 984 |
|
994 | |||
| 985 | #Setting Error |
|
995 | #Setting Error | |
| 986 | #Number 15: Radial Drift velocity or projected horizontal velocity exceeds 200 m/s |
|
996 | #Number 15: Radial Drift velocity or projected horizontal velocity exceeds 200 m/s | |
| 987 | if numpy.abs(radialVelocity) > 200: |
|
997 | if numpy.abs(radialVelocity) > 200: | |
| 988 | meteorAux[-1] = 15 |
|
998 | meteorAux[-1] = 15 | |
| 989 | #Number 12: Poor fit to CCF variation for estimation of radial drift velocity |
|
999 | #Number 12: Poor fit to CCF variation for estimation of radial drift velocity | |
| 990 | elif radialError > radialStdThresh: |
|
1000 | elif radialError > radialStdThresh: | |
| 991 | meteorAux[-1] = 12 |
|
1001 | meteorAux[-1] = 12 | |
| 992 |
|
1002 | |||
| 993 | listMeteors1.append(meteorAux) |
|
1003 | listMeteors1.append(meteorAux) | |
| 994 | return listMeteors1 |
|
1004 | return listMeteors1 | |
| 995 |
|
1005 | |||
| 996 | def __setNewArrays(self, listMeteors, date, heiRang): |
|
1006 | def __setNewArrays(self, listMeteors, date, heiRang): | |
| 997 |
|
1007 | |||
| 998 | #New arrays |
|
1008 | #New arrays | |
| 999 | arrayMeteors = numpy.array(listMeteors) |
|
1009 | arrayMeteors = numpy.array(listMeteors) | |
| 1000 | arrayParameters = numpy.zeros((len(listMeteors),10)) |
|
1010 | arrayParameters = numpy.zeros((len(listMeteors),10)) | |
| 1001 |
|
1011 | |||
| 1002 | #Date inclusion |
|
1012 | #Date inclusion | |
| 1003 | date = re.findall(r'\((.*?)\)', date) |
|
1013 | date = re.findall(r'\((.*?)\)', date) | |
| 1004 | date = date[0].split(',') |
|
1014 | date = date[0].split(',') | |
| 1005 | date = map(int, date) |
|
1015 | date = map(int, date) | |
| 1006 | date = [date[0]*10000 + date[1]*100 + date[2], date[3]*10000 + date[4]*100 + date[5]] |
|
1016 | date = [date[0]*10000 + date[1]*100 + date[2], date[3]*10000 + date[4]*100 + date[5]] | |
| 1007 | arrayDate = numpy.tile(date, (len(listMeteors), 1)) |
|
1017 | arrayDate = numpy.tile(date, (len(listMeteors), 1)) | |
| 1008 |
|
1018 | |||
| 1009 | #Meteor array |
|
1019 | #Meteor array | |
| 1010 | arrayMeteors[:,0] = heiRang[arrayMeteors[:,0].astype(int)] |
|
1020 | arrayMeteors[:,0] = heiRang[arrayMeteors[:,0].astype(int)] | |
| 1011 | arrayMeteors = numpy.hstack((arrayDate, arrayMeteors)) |
|
1021 | arrayMeteors = numpy.hstack((arrayDate, arrayMeteors)) | |
| 1012 |
|
1022 | |||
| 1013 | #Parameters Array |
|
1023 | #Parameters Array | |
| 1014 | arrayParameters[:,0:3] = arrayMeteors[:,0:3] |
|
1024 | arrayParameters[:,0:3] = arrayMeteors[:,0:3] | |
| 1015 | arrayParameters[:,-3:] = arrayMeteors[:,-3:] |
|
1025 | arrayParameters[:,-3:] = arrayMeteors[:,-3:] | |
| 1016 |
|
1026 | |||
| 1017 | return arrayMeteors, arrayParameters |
|
1027 | return arrayMeteors, arrayParameters | |
| 1018 |
|
1028 | |||
| 1019 | def __getAOA(self, phases, pairsList, error, AOAthresh, azimuth): |
|
1029 | def __getAOA(self, phases, pairsList, error, AOAthresh, azimuth): | |
| 1020 |
|
1030 | |||
| 1021 | arrayAOA = numpy.zeros((phases.shape[0],3)) |
|
1031 | arrayAOA = numpy.zeros((phases.shape[0],3)) | |
| 1022 | cosdir0, cosdir = self.__getDirectionCosines(phases, pairsList) |
|
1032 | cosdir0, cosdir = self.__getDirectionCosines(phases, pairsList) | |
| 1023 |
|
1033 | |||
| 1024 | arrayAOA[:,:2] = self.__calculateAOA(cosdir, azimuth) |
|
1034 | arrayAOA[:,:2] = self.__calculateAOA(cosdir, azimuth) | |
| 1025 | cosDirError = numpy.sum(numpy.abs(cosdir0 - cosdir), axis = 1) |
|
1035 | cosDirError = numpy.sum(numpy.abs(cosdir0 - cosdir), axis = 1) | |
| 1026 | arrayAOA[:,2] = cosDirError |
|
1036 | arrayAOA[:,2] = cosDirError | |
| 1027 |
|
1037 | |||
| 1028 | azimuthAngle = arrayAOA[:,0] |
|
1038 | azimuthAngle = arrayAOA[:,0] | |
| 1029 | zenithAngle = arrayAOA[:,1] |
|
1039 | zenithAngle = arrayAOA[:,1] | |
| 1030 |
|
1040 | |||
| 1031 | #Setting Error |
|
1041 | #Setting Error | |
| 1032 | #Number 3: AOA not fesible |
|
1042 | #Number 3: AOA not fesible | |
| 1033 | indInvalid = numpy.where(numpy.logical_and((numpy.logical_or(numpy.isnan(zenithAngle), numpy.isnan(azimuthAngle))),error == 0))[0] |
|
1043 | indInvalid = numpy.where(numpy.logical_and((numpy.logical_or(numpy.isnan(zenithAngle), numpy.isnan(azimuthAngle))),error == 0))[0] | |
| 1034 | error[indInvalid] = 3 |
|
1044 | error[indInvalid] = 3 | |
| 1035 | #Number 4: Large difference in AOAs obtained from different antenna baselines |
|
1045 | #Number 4: Large difference in AOAs obtained from different antenna baselines | |
| 1036 | indInvalid = numpy.where(numpy.logical_and(cosDirError > AOAthresh,error == 0))[0] |
|
1046 | indInvalid = numpy.where(numpy.logical_and(cosDirError > AOAthresh,error == 0))[0] | |
| 1037 | error[indInvalid] = 4 |
|
1047 | error[indInvalid] = 4 | |
| 1038 | return arrayAOA, error |
|
1048 | return arrayAOA, error | |
| 1039 |
|
1049 | |||
| 1040 | def __getDirectionCosines(self, arrayPhase, pairsList): |
|
1050 | def __getDirectionCosines(self, arrayPhase, pairsList): | |
| 1041 |
|
1051 | |||
| 1042 | #Initializing some variables |
|
1052 | #Initializing some variables | |
| 1043 | ang_aux = numpy.array([-8,-7,-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6,7,8])*2*numpy.pi |
|
1053 | ang_aux = numpy.array([-8,-7,-6,-5,-4,-3,-2,-1,0,1,2,3,4,5,6,7,8])*2*numpy.pi | |
| 1044 | ang_aux = ang_aux.reshape(1,ang_aux.size) |
|
1054 | ang_aux = ang_aux.reshape(1,ang_aux.size) | |
| 1045 |
|
1055 | |||
| 1046 | cosdir = numpy.zeros((arrayPhase.shape[0],2)) |
|
1056 | cosdir = numpy.zeros((arrayPhase.shape[0],2)) | |
| 1047 | cosdir0 = numpy.zeros((arrayPhase.shape[0],2)) |
|
1057 | cosdir0 = numpy.zeros((arrayPhase.shape[0],2)) | |
| 1048 |
|
1058 | |||
| 1049 |
|
1059 | |||
| 1050 | for i in range(2): |
|
1060 | for i in range(2): | |
| 1051 | #First Estimation |
|
1061 | #First Estimation | |
| 1052 | phi0_aux = arrayPhase[:,pairsList[i][0]] + arrayPhase[:,pairsList[i][1]] |
|
1062 | phi0_aux = arrayPhase[:,pairsList[i][0]] + arrayPhase[:,pairsList[i][1]] | |
| 1053 | #Dealias |
|
1063 | #Dealias | |
| 1054 | indcsi = numpy.where(phi0_aux > numpy.pi) |
|
1064 | indcsi = numpy.where(phi0_aux > numpy.pi) | |
| 1055 | phi0_aux[indcsi] -= 2*numpy.pi |
|
1065 | phi0_aux[indcsi] -= 2*numpy.pi | |
| 1056 | indcsi = numpy.where(phi0_aux < -numpy.pi) |
|
1066 | indcsi = numpy.where(phi0_aux < -numpy.pi) | |
| 1057 | phi0_aux[indcsi] += 2*numpy.pi |
|
1067 | phi0_aux[indcsi] += 2*numpy.pi | |
| 1058 | #Direction Cosine 0 |
|
1068 | #Direction Cosine 0 | |
| 1059 | cosdir0[:,i] = -(phi0_aux)/(2*numpy.pi*0.5) |
|
1069 | cosdir0[:,i] = -(phi0_aux)/(2*numpy.pi*0.5) | |
| 1060 |
|
1070 | |||
| 1061 | #Most-Accurate Second Estimation |
|
1071 | #Most-Accurate Second Estimation | |
| 1062 | phi1_aux = arrayPhase[:,pairsList[i][0]] - arrayPhase[:,pairsList[i][1]] |
|
1072 | phi1_aux = arrayPhase[:,pairsList[i][0]] - arrayPhase[:,pairsList[i][1]] | |
| 1063 | phi1_aux = phi1_aux.reshape(phi1_aux.size,1) |
|
1073 | phi1_aux = phi1_aux.reshape(phi1_aux.size,1) | |
| 1064 | #Direction Cosine 1 |
|
1074 | #Direction Cosine 1 | |
| 1065 | cosdir1 = -(phi1_aux + ang_aux)/(2*numpy.pi*4.5) |
|
1075 | cosdir1 = -(phi1_aux + ang_aux)/(2*numpy.pi*4.5) | |
| 1066 |
|
1076 | |||
| 1067 | #Searching the correct Direction Cosine |
|
1077 | #Searching the correct Direction Cosine | |
| 1068 | cosdir0_aux = cosdir0[:,i] |
|
1078 | cosdir0_aux = cosdir0[:,i] | |
| 1069 | cosdir0_aux = cosdir0_aux.reshape(cosdir0_aux.size,1) |
|
1079 | cosdir0_aux = cosdir0_aux.reshape(cosdir0_aux.size,1) | |
| 1070 | #Minimum Distance |
|
1080 | #Minimum Distance | |
| 1071 | cosDiff = (cosdir1 - cosdir0_aux)**2 |
|
1081 | cosDiff = (cosdir1 - cosdir0_aux)**2 | |
| 1072 | indcos = cosDiff.argmin(axis = 1) |
|
1082 | indcos = cosDiff.argmin(axis = 1) | |
| 1073 | #Saving Value obtained |
|
1083 | #Saving Value obtained | |
| 1074 | cosdir[:,i] = cosdir1[numpy.arange(len(indcos)),indcos] |
|
1084 | cosdir[:,i] = cosdir1[numpy.arange(len(indcos)),indcos] | |
| 1075 |
|
1085 | |||
| 1076 | return cosdir0, cosdir |
|
1086 | return cosdir0, cosdir | |
| 1077 |
|
1087 | |||
| 1078 | def __calculateAOA(self, cosdir, azimuth): |
|
1088 | def __calculateAOA(self, cosdir, azimuth): | |
| 1079 | cosdirX = cosdir[:,0] |
|
1089 | cosdirX = cosdir[:,0] | |
| 1080 | cosdirY = cosdir[:,1] |
|
1090 | cosdirY = cosdir[:,1] | |
| 1081 |
|
1091 | |||
| 1082 | zenithAngle = numpy.arccos(numpy.sqrt(1 - cosdirX**2 - cosdirY**2))*180/numpy.pi |
|
1092 | zenithAngle = numpy.arccos(numpy.sqrt(1 - cosdirX**2 - cosdirY**2))*180/numpy.pi | |
| 1083 | azimuthAngle = numpy.arctan2(cosdirX,cosdirY)*180/numpy.pi + azimuth #0 deg north, 90 deg east |
|
1093 | azimuthAngle = numpy.arctan2(cosdirX,cosdirY)*180/numpy.pi + azimuth #0 deg north, 90 deg east | |
| 1084 | angles = numpy.vstack((azimuthAngle, zenithAngle)).transpose() |
|
1094 | angles = numpy.vstack((azimuthAngle, zenithAngle)).transpose() | |
| 1085 |
|
1095 | |||
| 1086 | return angles |
|
1096 | return angles | |
| 1087 |
|
1097 | |||
| 1088 | def __getHeights(self, Ranges, zenith, error, minHeight, maxHeight): |
|
1098 | def __getHeights(self, Ranges, zenith, error, minHeight, maxHeight): | |
| 1089 |
|
1099 | |||
| 1090 | Ramb = 375 #Ramb = c/(2*PRF) |
|
1100 | Ramb = 375 #Ramb = c/(2*PRF) | |
| 1091 | Re = 6371 #Earth Radius |
|
1101 | Re = 6371 #Earth Radius | |
| 1092 | heights = numpy.zeros(Ranges.shape) |
|
1102 | heights = numpy.zeros(Ranges.shape) | |
| 1093 |
|
1103 | |||
| 1094 | R_aux = numpy.array([0,1,2])*Ramb |
|
1104 | R_aux = numpy.array([0,1,2])*Ramb | |
| 1095 | R_aux = R_aux.reshape(1,R_aux.size) |
|
1105 | R_aux = R_aux.reshape(1,R_aux.size) | |
| 1096 |
|
1106 | |||
| 1097 | Ranges = Ranges.reshape(Ranges.size,1) |
|
1107 | Ranges = Ranges.reshape(Ranges.size,1) | |
| 1098 |
|
1108 | |||
| 1099 | Ri = Ranges + R_aux |
|
1109 | Ri = Ranges + R_aux | |
| 1100 | hi = numpy.sqrt(Re**2 + Ri**2 + (2*Re*numpy.cos(zenith*numpy.pi/180)*Ri.transpose()).transpose()) - Re |
|
1110 | hi = numpy.sqrt(Re**2 + Ri**2 + (2*Re*numpy.cos(zenith*numpy.pi/180)*Ri.transpose()).transpose()) - Re | |
| 1101 |
|
1111 | |||
| 1102 | #Check if there is a height between 70 and 110 km |
|
1112 | #Check if there is a height between 70 and 110 km | |
| 1103 | h_bool = numpy.sum(numpy.logical_and(hi > minHeight, hi < maxHeight), axis = 1) |
|
1113 | h_bool = numpy.sum(numpy.logical_and(hi > minHeight, hi < maxHeight), axis = 1) | |
| 1104 | ind_h = numpy.where(h_bool == 1)[0] |
|
1114 | ind_h = numpy.where(h_bool == 1)[0] | |
| 1105 |
|
1115 | |||
| 1106 | hCorr = hi[ind_h, :] |
|
1116 | hCorr = hi[ind_h, :] | |
| 1107 | ind_hCorr = numpy.where(numpy.logical_and(hi > minHeight, hi < maxHeight)) |
|
1117 | ind_hCorr = numpy.where(numpy.logical_and(hi > minHeight, hi < maxHeight)) | |
| 1108 |
|
1118 | |||
| 1109 | hCorr = hi[ind_hCorr] |
|
1119 | hCorr = hi[ind_hCorr] | |
| 1110 | heights[ind_h] = hCorr |
|
1120 | heights[ind_h] = hCorr | |
| 1111 |
|
1121 | |||
| 1112 | #Setting Error |
|
1122 | #Setting Error | |
| 1113 | #Number 13: Height unresolvable echo: not valid height within 70 to 110 km |
|
1123 | #Number 13: Height unresolvable echo: not valid height within 70 to 110 km | |
| 1114 | #Number 14: Height ambiguous echo: more than one possible height within 70 to 110 km |
|
1124 | #Number 14: Height ambiguous echo: more than one possible height within 70 to 110 km | |
| 1115 |
|
1125 | |||
| 1116 | indInvalid2 = numpy.where(numpy.logical_and(h_bool > 1, error == 0))[0] |
|
1126 | indInvalid2 = numpy.where(numpy.logical_and(h_bool > 1, error == 0))[0] | |
| 1117 | error[indInvalid2] = 14 |
|
1127 | error[indInvalid2] = 14 | |
| 1118 | indInvalid1 = numpy.where(numpy.logical_and(h_bool == 0, error == 0))[0] |
|
1128 | indInvalid1 = numpy.where(numpy.logical_and(h_bool == 0, error == 0))[0] | |
| 1119 | error[indInvalid1] = 13 |
|
1129 | error[indInvalid1] = 13 | |
| 1120 |
|
1130 | |||
| 1121 | return heights, error |
|
1131 | return heights, error | |
| 1122 |
|
1132 | |||
| 1123 | def SpectralFitting(self, getSNR = True, path=None, file=None, groupList=None): |
|
1133 | def SpectralFitting(self, getSNR = True, path=None, file=None, groupList=None): | |
| 1124 |
|
1134 | |||
| 1125 | ''' |
|
1135 | ''' | |
| 1126 | Function GetMoments() |
|
1136 | Function GetMoments() | |
| 1127 |
|
1137 | |||
| 1128 | Input: |
|
1138 | Input: | |
| 1129 | Output: |
|
1139 | Output: | |
| 1130 | Variables modified: |
|
1140 | Variables modified: | |
| 1131 | ''' |
|
1141 | ''' | |
| 1132 | if path != None: |
|
1142 | if path != None: | |
| 1133 | sys.path.append(path) |
|
1143 | sys.path.append(path) | |
| 1134 | self.dataOut.library = importlib.import_module(file) |
|
1144 | self.dataOut.library = importlib.import_module(file) | |
| 1135 |
|
1145 | |||
| 1136 | #To be inserted as a parameter |
|
1146 | #To be inserted as a parameter | |
| 1137 | groupArray = numpy.array(groupList) |
|
1147 | groupArray = numpy.array(groupList) | |
| 1138 | # groupArray = numpy.array([[0,1],[2,3]]) |
|
1148 | # groupArray = numpy.array([[0,1],[2,3]]) | |
| 1139 | self.dataOut.groupList = groupArray |
|
1149 | self.dataOut.groupList = groupArray | |
| 1140 |
|
1150 | |||
| 1141 | nGroups = groupArray.shape[0] |
|
1151 | nGroups = groupArray.shape[0] | |
| 1142 | nChannels = self.dataIn.nChannels |
|
1152 | nChannels = self.dataIn.nChannels | |
| 1143 | nHeights=self.dataIn.heightList.size |
|
1153 | nHeights=self.dataIn.heightList.size | |
| 1144 |
|
1154 | |||
| 1145 | #Parameters Array |
|
1155 | #Parameters Array | |
| 1146 | self.dataOut.data_param = None |
|
1156 | self.dataOut.data_param = None | |
| 1147 |
|
1157 | |||
| 1148 | #Set constants |
|
1158 | #Set constants | |
| 1149 | constants = self.dataOut.library.setConstants(self.dataIn) |
|
1159 | constants = self.dataOut.library.setConstants(self.dataIn) | |
| 1150 | self.dataOut.constants = constants |
|
1160 | self.dataOut.constants = constants | |
| 1151 | M = self.dataIn.normFactor |
|
1161 | M = self.dataIn.normFactor | |
| 1152 | N = self.dataIn.nFFTPoints |
|
1162 | N = self.dataIn.nFFTPoints | |
| 1153 | ippSeconds = self.dataIn.ippSeconds |
|
1163 | ippSeconds = self.dataIn.ippSeconds | |
| 1154 | K = self.dataIn.nIncohInt |
|
1164 | K = self.dataIn.nIncohInt | |
| 1155 | pairsArray = numpy.array(self.dataIn.pairsList) |
|
1165 | pairsArray = numpy.array(self.dataIn.pairsList) | |
| 1156 |
|
1166 | |||
| 1157 | #List of possible combinations |
|
1167 | #List of possible combinations | |
| 1158 | listComb = itertools.combinations(numpy.arange(groupArray.shape[1]),2) |
|
1168 | listComb = itertools.combinations(numpy.arange(groupArray.shape[1]),2) | |
| 1159 | indCross = numpy.zeros(len(list(listComb)), dtype = 'int') |
|
1169 | indCross = numpy.zeros(len(list(listComb)), dtype = 'int') | |
| 1160 |
|
1170 | |||
| 1161 | if getSNR: |
|
1171 | if getSNR: | |
| 1162 | listChannels = groupArray.reshape((groupArray.size)) |
|
1172 | listChannels = groupArray.reshape((groupArray.size)) | |
| 1163 | listChannels.sort() |
|
1173 | listChannels.sort() | |
| 1164 | noise = self.dataIn.getNoise() |
|
1174 | noise = self.dataIn.getNoise() | |
| 1165 | self.dataOut.data_SNR = self.__getSNR(self.dataIn.data_spc[listChannels,:,:], noise[listChannels]) |
|
1175 | self.dataOut.data_SNR = self.__getSNR(self.dataIn.data_spc[listChannels,:,:], noise[listChannels]) | |
| 1166 |
|
1176 | |||
| 1167 | for i in range(nGroups): |
|
1177 | for i in range(nGroups): | |
| 1168 | coord = groupArray[i,:] |
|
1178 | coord = groupArray[i,:] | |
| 1169 |
|
1179 | |||
| 1170 | #Input data array |
|
1180 | #Input data array | |
| 1171 | data = self.dataIn.data_spc[coord,:,:]/(M*N) |
|
1181 | data = self.dataIn.data_spc[coord,:,:]/(M*N) | |
| 1172 | data = data.reshape((data.shape[0]*data.shape[1],data.shape[2])) |
|
1182 | data = data.reshape((data.shape[0]*data.shape[1],data.shape[2])) | |
| 1173 |
|
1183 | |||
| 1174 | #Cross Spectra data array for Covariance Matrixes |
|
1184 | #Cross Spectra data array for Covariance Matrixes | |
| 1175 | ind = 0 |
|
1185 | ind = 0 | |
| 1176 | for pairs in listComb: |
|
1186 | for pairs in listComb: | |
| 1177 | pairsSel = numpy.array([coord[x],coord[y]]) |
|
1187 | pairsSel = numpy.array([coord[x],coord[y]]) | |
| 1178 | indCross[ind] = int(numpy.where(numpy.all(pairsArray == pairsSel, axis = 1))[0][0]) |
|
1188 | indCross[ind] = int(numpy.where(numpy.all(pairsArray == pairsSel, axis = 1))[0][0]) | |
| 1179 | ind += 1 |
|
1189 | ind += 1 | |
| 1180 | dataCross = self.dataIn.data_cspc[indCross,:,:]/(M*N) |
|
1190 | dataCross = self.dataIn.data_cspc[indCross,:,:]/(M*N) | |
| 1181 | dataCross = dataCross**2/K |
|
1191 | dataCross = dataCross**2/K | |
| 1182 |
|
1192 | |||
| 1183 | for h in range(nHeights): |
|
1193 | for h in range(nHeights): | |
| 1184 | # print self.dataOut.heightList[h] |
|
1194 | # print self.dataOut.heightList[h] | |
| 1185 |
|
1195 | |||
| 1186 | #Input |
|
1196 | #Input | |
| 1187 | d = data[:,h] |
|
1197 | d = data[:,h] | |
| 1188 |
|
1198 | |||
| 1189 | #Covariance Matrix |
|
1199 | #Covariance Matrix | |
| 1190 | D = numpy.diag(d**2/K) |
|
1200 | D = numpy.diag(d**2/K) | |
| 1191 | ind = 0 |
|
1201 | ind = 0 | |
| 1192 | for pairs in listComb: |
|
1202 | for pairs in listComb: | |
| 1193 | #Coordinates in Covariance Matrix |
|
1203 | #Coordinates in Covariance Matrix | |
| 1194 | x = pairs[0] |
|
1204 | x = pairs[0] | |
| 1195 | y = pairs[1] |
|
1205 | y = pairs[1] | |
| 1196 | #Channel Index |
|
1206 | #Channel Index | |
| 1197 | S12 = dataCross[ind,:,h] |
|
1207 | S12 = dataCross[ind,:,h] | |
| 1198 | D12 = numpy.diag(S12) |
|
1208 | D12 = numpy.diag(S12) | |
| 1199 | #Completing Covariance Matrix with Cross Spectras |
|
1209 | #Completing Covariance Matrix with Cross Spectras | |
| 1200 | D[x*N:(x+1)*N,y*N:(y+1)*N] = D12 |
|
1210 | D[x*N:(x+1)*N,y*N:(y+1)*N] = D12 | |
| 1201 | D[y*N:(y+1)*N,x*N:(x+1)*N] = D12 |
|
1211 | D[y*N:(y+1)*N,x*N:(x+1)*N] = D12 | |
| 1202 | ind += 1 |
|
1212 | ind += 1 | |
| 1203 | Dinv=numpy.linalg.inv(D) |
|
1213 | Dinv=numpy.linalg.inv(D) | |
| 1204 | L=numpy.linalg.cholesky(Dinv) |
|
1214 | L=numpy.linalg.cholesky(Dinv) | |
| 1205 | LT=L.T |
|
1215 | LT=L.T | |
| 1206 |
|
1216 | |||
| 1207 | dp = numpy.dot(LT,d) |
|
1217 | dp = numpy.dot(LT,d) | |
| 1208 |
|
1218 | |||
| 1209 | #Initial values |
|
1219 | #Initial values | |
| 1210 | data_spc = self.dataIn.data_spc[coord,:,h] |
|
1220 | data_spc = self.dataIn.data_spc[coord,:,h] | |
| 1211 |
|
1221 | |||
| 1212 | if (h>0)and(error1[3]<5): |
|
1222 | if (h>0)and(error1[3]<5): | |
| 1213 | p0 = self.dataOut.data_param[i,:,h-1] |
|
1223 | p0 = self.dataOut.data_param[i,:,h-1] | |
| 1214 | else: |
|
1224 | else: | |
| 1215 | p0 = numpy.array(self.dataOut.library.initialValuesFunction(data_spc, constants, i)) |
|
1225 | p0 = numpy.array(self.dataOut.library.initialValuesFunction(data_spc, constants, i)) | |
| 1216 |
|
1226 | |||
| 1217 | try: |
|
1227 | try: | |
| 1218 | #Least Squares |
|
1228 | #Least Squares | |
| 1219 | minp,covp,infodict,mesg,ier = optimize.leastsq(self.__residFunction,p0,args=(dp,LT,constants),full_output=True) |
|
1229 | minp,covp,infodict,mesg,ier = optimize.leastsq(self.__residFunction,p0,args=(dp,LT,constants),full_output=True) | |
| 1220 | # minp,covp = optimize.leastsq(self.__residFunction,p0,args=(dp,LT,constants)) |
|
1230 | # minp,covp = optimize.leastsq(self.__residFunction,p0,args=(dp,LT,constants)) | |
| 1221 | #Chi square error |
|
1231 | #Chi square error | |
| 1222 | error0 = numpy.sum(infodict['fvec']**2)/(2*N) |
|
1232 | error0 = numpy.sum(infodict['fvec']**2)/(2*N) | |
| 1223 | #Error with Jacobian |
|
1233 | #Error with Jacobian | |
| 1224 | error1 = self.dataOut.library.errorFunction(minp,constants,LT) |
|
1234 | error1 = self.dataOut.library.errorFunction(minp,constants,LT) | |
| 1225 | except: |
|
1235 | except: | |
| 1226 | minp = p0*numpy.nan |
|
1236 | minp = p0*numpy.nan | |
| 1227 | error0 = numpy.nan |
|
1237 | error0 = numpy.nan | |
| 1228 | error1 = p0*numpy.nan |
|
1238 | error1 = p0*numpy.nan | |
| 1229 |
|
1239 | |||
| 1230 | #Save |
|
1240 | #Save | |
| 1231 | if self.dataOut.data_param == None: |
|
1241 | if self.dataOut.data_param == None: | |
| 1232 | self.dataOut.data_param = numpy.zeros((nGroups, p0.size, nHeights))*numpy.nan |
|
1242 | self.dataOut.data_param = numpy.zeros((nGroups, p0.size, nHeights))*numpy.nan | |
| 1233 | self.dataOut.data_error = numpy.zeros((nGroups, p0.size + 1, nHeights))*numpy.nan |
|
1243 | self.dataOut.data_error = numpy.zeros((nGroups, p0.size + 1, nHeights))*numpy.nan | |
| 1234 |
|
1244 | |||
| 1235 | self.dataOut.data_error[i,:,h] = numpy.hstack((error0,error1)) |
|
1245 | self.dataOut.data_error[i,:,h] = numpy.hstack((error0,error1)) | |
| 1236 | self.dataOut.data_param[i,:,h] = minp |
|
1246 | self.dataOut.data_param[i,:,h] = minp | |
| 1237 | return |
|
1247 | return | |
| 1238 |
|
1248 | |||
| 1239 |
|
1249 | |||
| 1240 | def __residFunction(self, p, dp, LT, constants): |
|
1250 | def __residFunction(self, p, dp, LT, constants): | |
| 1241 |
|
1251 | |||
| 1242 | fm = self.dataOut.library.modelFunction(p, constants) |
|
1252 | fm = self.dataOut.library.modelFunction(p, constants) | |
| 1243 | fmp=numpy.dot(LT,fm) |
|
1253 | fmp=numpy.dot(LT,fm) | |
| 1244 |
|
1254 | |||
| 1245 | return dp-fmp |
|
1255 | return dp-fmp | |
| 1246 |
|
1256 | |||
| 1247 | def __getSNR(self, z, noise): |
|
1257 | def __getSNR(self, z, noise): | |
| 1248 |
|
1258 | |||
| 1249 | avg = numpy.average(z, axis=1) |
|
1259 | avg = numpy.average(z, axis=1) | |
| 1250 | SNR = (avg.T-noise)/noise |
|
1260 | SNR = (avg.T-noise)/noise | |
| 1251 | SNR = SNR.T |
|
1261 | SNR = SNR.T | |
| 1252 | return SNR |
|
1262 | return SNR | |
| 1253 |
|
1263 | |||
| 1254 | def __chisq(p,chindex,hindex): |
|
1264 | def __chisq(p,chindex,hindex): | |
| 1255 | #similar to Resid but calculates CHI**2 |
|
1265 | #similar to Resid but calculates CHI**2 | |
| 1256 | [LT,d,fm]=setupLTdfm(p,chindex,hindex) |
|
1266 | [LT,d,fm]=setupLTdfm(p,chindex,hindex) | |
| 1257 | dp=numpy.dot(LT,d) |
|
1267 | dp=numpy.dot(LT,d) | |
| 1258 | fmp=numpy.dot(LT,fm) |
|
1268 | fmp=numpy.dot(LT,fm) | |
| 1259 | chisq=numpy.dot((dp-fmp).T,(dp-fmp)) |
|
1269 | chisq=numpy.dot((dp-fmp).T,(dp-fmp)) | |
| 1260 | return chisq |
|
1270 | return chisq | |
| 1261 |
|
1271 | |||
| 1262 |
|
1272 | |||
| 1263 |
|
1273 | |||
| 1264 | class WindProfiler(Operation): |
|
1274 | class WindProfiler(Operation): | |
| 1265 |
|
1275 | |||
| 1266 | __isConfig = False |
|
1276 | __isConfig = False | |
| 1267 |
|
1277 | |||
| 1268 | __initime = None |
|
1278 | __initime = None | |
| 1269 | __lastdatatime = None |
|
1279 | __lastdatatime = None | |
| 1270 | __integrationtime = None |
|
1280 | __integrationtime = None | |
| 1271 |
|
1281 | |||
| 1272 | __buffer = None |
|
1282 | __buffer = None | |
| 1273 |
|
1283 | |||
| 1274 | __dataReady = False |
|
1284 | __dataReady = False | |
| 1275 |
|
1285 | |||
| 1276 | __firstdata = None |
|
1286 | __firstdata = None | |
| 1277 |
|
1287 | |||
| 1278 | n = None |
|
1288 | n = None | |
| 1279 |
|
1289 | |||
| 1280 | def __init__(self): |
|
1290 | def __init__(self): | |
| 1281 | Operation.__init__(self) |
|
1291 | Operation.__init__(self) | |
| 1282 |
|
1292 | |||
| 1283 | def __calculateCosDir(self, elev, azim): |
|
1293 | def __calculateCosDir(self, elev, azim): | |
| 1284 | zen = (90 - elev)*numpy.pi/180 |
|
1294 | zen = (90 - elev)*numpy.pi/180 | |
| 1285 | azim = azim*numpy.pi/180 |
|
1295 | azim = azim*numpy.pi/180 | |
| 1286 | cosDirX = numpy.sqrt((1-numpy.cos(zen)**2)/((1+numpy.tan(azim)**2))) |
|
1296 | cosDirX = numpy.sqrt((1-numpy.cos(zen)**2)/((1+numpy.tan(azim)**2))) | |
| 1287 | cosDirY = numpy.sqrt(1-numpy.cos(zen)**2-cosDirX**2) |
|
1297 | cosDirY = numpy.sqrt(1-numpy.cos(zen)**2-cosDirX**2) | |
| 1288 |
|
1298 | |||
| 1289 | signX = numpy.sign(numpy.cos(azim)) |
|
1299 | signX = numpy.sign(numpy.cos(azim)) | |
| 1290 | signY = numpy.sign(numpy.sin(azim)) |
|
1300 | signY = numpy.sign(numpy.sin(azim)) | |
| 1291 |
|
1301 | |||
| 1292 | cosDirX = numpy.copysign(cosDirX, signX) |
|
1302 | cosDirX = numpy.copysign(cosDirX, signX) | |
| 1293 | cosDirY = numpy.copysign(cosDirY, signY) |
|
1303 | cosDirY = numpy.copysign(cosDirY, signY) | |
| 1294 | return cosDirX, cosDirY |
|
1304 | return cosDirX, cosDirY | |
| 1295 |
|
1305 | |||
| 1296 | def __calculateAngles(self, theta_x, theta_y, azimuth): |
|
1306 | def __calculateAngles(self, theta_x, theta_y, azimuth): | |
| 1297 |
|
1307 | |||
| 1298 | dir_cosw = numpy.sqrt(1-theta_x**2-theta_y**2) |
|
1308 | dir_cosw = numpy.sqrt(1-theta_x**2-theta_y**2) | |
| 1299 | zenith_arr = numpy.arccos(dir_cosw) |
|
1309 | zenith_arr = numpy.arccos(dir_cosw) | |
| 1300 | azimuth_arr = numpy.arctan2(theta_x,theta_y) + azimuth*math.pi/180 |
|
1310 | azimuth_arr = numpy.arctan2(theta_x,theta_y) + azimuth*math.pi/180 | |
| 1301 |
|
1311 | |||
| 1302 | dir_cosu = numpy.sin(azimuth_arr)*numpy.sin(zenith_arr) |
|
1312 | dir_cosu = numpy.sin(azimuth_arr)*numpy.sin(zenith_arr) | |
| 1303 | dir_cosv = numpy.cos(azimuth_arr)*numpy.sin(zenith_arr) |
|
1313 | dir_cosv = numpy.cos(azimuth_arr)*numpy.sin(zenith_arr) | |
| 1304 |
|
1314 | |||
| 1305 | return azimuth_arr, zenith_arr, dir_cosu, dir_cosv, dir_cosw |
|
1315 | return azimuth_arr, zenith_arr, dir_cosu, dir_cosv, dir_cosw | |
| 1306 |
|
1316 | |||
| 1307 | def __calculateMatA(self, dir_cosu, dir_cosv, dir_cosw, horOnly): |
|
1317 | def __calculateMatA(self, dir_cosu, dir_cosv, dir_cosw, horOnly): | |
| 1308 |
|
1318 | |||
| 1309 | # |
|
1319 | # | |
| 1310 | if horOnly: |
|
1320 | if horOnly: | |
| 1311 | A = numpy.c_[dir_cosu,dir_cosv] |
|
1321 | A = numpy.c_[dir_cosu,dir_cosv] | |
| 1312 | else: |
|
1322 | else: | |
| 1313 | A = numpy.c_[dir_cosu,dir_cosv,dir_cosw] |
|
1323 | A = numpy.c_[dir_cosu,dir_cosv,dir_cosw] | |
| 1314 | A = numpy.asmatrix(A) |
|
1324 | A = numpy.asmatrix(A) | |
| 1315 | A1 = numpy.linalg.inv(A.transpose()*A)*A.transpose() |
|
1325 | A1 = numpy.linalg.inv(A.transpose()*A)*A.transpose() | |
| 1316 |
|
1326 | |||
| 1317 | return A1 |
|
1327 | return A1 | |
| 1318 |
|
1328 | |||
| 1319 | def __correctValues(self, heiRang, phi, velRadial, SNR): |
|
1329 | def __correctValues(self, heiRang, phi, velRadial, SNR): | |
| 1320 | listPhi = phi.tolist() |
|
1330 | listPhi = phi.tolist() | |
| 1321 | maxid = listPhi.index(max(listPhi)) |
|
1331 | maxid = listPhi.index(max(listPhi)) | |
| 1322 | minid = listPhi.index(min(listPhi)) |
|
1332 | minid = listPhi.index(min(listPhi)) | |
| 1323 |
|
1333 | |||
| 1324 | rango = range(len(phi)) |
|
1334 | rango = range(len(phi)) | |
| 1325 | # rango = numpy.delete(rango,maxid) |
|
1335 | # rango = numpy.delete(rango,maxid) | |
| 1326 |
|
1336 | |||
| 1327 | heiRang1 = heiRang*math.cos(phi[maxid]) |
|
1337 | heiRang1 = heiRang*math.cos(phi[maxid]) | |
| 1328 | heiRangAux = heiRang*math.cos(phi[minid]) |
|
1338 | heiRangAux = heiRang*math.cos(phi[minid]) | |
| 1329 | indOut = (heiRang1 < heiRangAux[0]).nonzero() |
|
1339 | indOut = (heiRang1 < heiRangAux[0]).nonzero() | |
| 1330 | heiRang1 = numpy.delete(heiRang1,indOut) |
|
1340 | heiRang1 = numpy.delete(heiRang1,indOut) | |
| 1331 |
|
1341 | |||
| 1332 | velRadial1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
1342 | velRadial1 = numpy.zeros([len(phi),len(heiRang1)]) | |
| 1333 | SNR1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
1343 | SNR1 = numpy.zeros([len(phi),len(heiRang1)]) | |
| 1334 |
|
1344 | |||
| 1335 | for i in rango: |
|
1345 | for i in rango: | |
| 1336 | x = heiRang*math.cos(phi[i]) |
|
1346 | x = heiRang*math.cos(phi[i]) | |
| 1337 | y1 = velRadial[i,:] |
|
1347 | y1 = velRadial[i,:] | |
| 1338 | f1 = interpolate.interp1d(x,y1,kind = 'cubic') |
|
1348 | f1 = interpolate.interp1d(x,y1,kind = 'cubic') | |
| 1339 |
|
1349 | |||
| 1340 | x1 = heiRang1 |
|
1350 | x1 = heiRang1 | |
| 1341 | y11 = f1(x1) |
|
1351 | y11 = f1(x1) | |
| 1342 |
|
1352 | |||
| 1343 | y2 = SNR[i,:] |
|
1353 | y2 = SNR[i,:] | |
| 1344 | f2 = interpolate.interp1d(x,y2,kind = 'cubic') |
|
1354 | f2 = interpolate.interp1d(x,y2,kind = 'cubic') | |
| 1345 | y21 = f2(x1) |
|
1355 | y21 = f2(x1) | |
| 1346 |
|
1356 | |||
| 1347 | velRadial1[i,:] = y11 |
|
1357 | velRadial1[i,:] = y11 | |
| 1348 | SNR1[i,:] = y21 |
|
1358 | SNR1[i,:] = y21 | |
| 1349 |
|
1359 | |||
| 1350 | return heiRang1, velRadial1, SNR1 |
|
1360 | return heiRang1, velRadial1, SNR1 | |
| 1351 |
|
1361 | |||
| 1352 | def __calculateVelUVW(self, A, velRadial): |
|
1362 | def __calculateVelUVW(self, A, velRadial): | |
| 1353 |
|
1363 | |||
| 1354 | #Operacion Matricial |
|
1364 | #Operacion Matricial | |
| 1355 | # velUVW = numpy.zeros((velRadial.shape[1],3)) |
|
1365 | # velUVW = numpy.zeros((velRadial.shape[1],3)) | |
| 1356 | # for ind in range(velRadial.shape[1]): |
|
1366 | # for ind in range(velRadial.shape[1]): | |
| 1357 | # velUVW[ind,:] = numpy.dot(A,velRadial[:,ind]) |
|
1367 | # velUVW[ind,:] = numpy.dot(A,velRadial[:,ind]) | |
| 1358 | # velUVW = velUVW.transpose() |
|
1368 | # velUVW = velUVW.transpose() | |
| 1359 | velUVW = numpy.zeros((A.shape[0],velRadial.shape[1])) |
|
1369 | velUVW = numpy.zeros((A.shape[0],velRadial.shape[1])) | |
| 1360 | velUVW[:,:] = numpy.dot(A,velRadial) |
|
1370 | velUVW[:,:] = numpy.dot(A,velRadial) | |
| 1361 |
|
1371 | |||
| 1362 |
|
1372 | |||
| 1363 | return velUVW |
|
1373 | return velUVW | |
| 1364 |
|
1374 | |||
| 1365 | def techniqueDBS(self, velRadial0, dirCosx, disrCosy, azimuth, correct, horizontalOnly, heiRang, SNR0): |
|
1375 | def techniqueDBS(self, velRadial0, dirCosx, disrCosy, azimuth, correct, horizontalOnly, heiRang, SNR0): | |
| 1366 | """ |
|
1376 | """ | |
| 1367 | Function that implements Doppler Beam Swinging (DBS) technique. |
|
1377 | Function that implements Doppler Beam Swinging (DBS) technique. | |
| 1368 |
|
1378 | |||
| 1369 | Input: Radial velocities, Direction cosines (x and y) of the Beam, Antenna azimuth, |
|
1379 | Input: Radial velocities, Direction cosines (x and y) of the Beam, Antenna azimuth, | |
| 1370 | Direction correction (if necessary), Ranges and SNR |
|
1380 | Direction correction (if necessary), Ranges and SNR | |
| 1371 |
|
1381 | |||
| 1372 | Output: Winds estimation (Zonal, Meridional and Vertical) |
|
1382 | Output: Winds estimation (Zonal, Meridional and Vertical) | |
| 1373 |
|
1383 | |||
| 1374 | Parameters affected: Winds, height range, SNR |
|
1384 | Parameters affected: Winds, height range, SNR | |
| 1375 | """ |
|
1385 | """ | |
| 1376 | azimuth_arr, zenith_arr, dir_cosu, dir_cosv, dir_cosw = self.__calculateAngles(dirCosx, disrCosy, azimuth) |
|
1386 | azimuth_arr, zenith_arr, dir_cosu, dir_cosv, dir_cosw = self.__calculateAngles(dirCosx, disrCosy, azimuth) | |
| 1377 | heiRang1, velRadial1, SNR1 = self.__correctValues(heiRang, zenith_arr, correct*velRadial0, SNR0) |
|
1387 | heiRang1, velRadial1, SNR1 = self.__correctValues(heiRang, zenith_arr, correct*velRadial0, SNR0) | |
| 1378 | A = self.__calculateMatA(dir_cosu, dir_cosv, dir_cosw, horizontalOnly) |
|
1388 | A = self.__calculateMatA(dir_cosu, dir_cosv, dir_cosw, horizontalOnly) | |
| 1379 |
|
1389 | |||
| 1380 | #Calculo de Componentes de la velocidad con DBS |
|
1390 | #Calculo de Componentes de la velocidad con DBS | |
| 1381 | winds = self.__calculateVelUVW(A,velRadial1) |
|
1391 | winds = self.__calculateVelUVW(A,velRadial1) | |
| 1382 |
|
1392 | |||
| 1383 | return winds, heiRang1, SNR1 |
|
1393 | return winds, heiRang1, SNR1 | |
| 1384 |
|
1394 | |||
| 1385 | def __calculateDistance(self, posx, posy, pairsCrossCorr, pairsList, pairs, azimuth = None): |
|
1395 | def __calculateDistance(self, posx, posy, pairsCrossCorr, pairsList, pairs, azimuth = None): | |
| 1386 |
|
1396 | |||
| 1387 | posx = numpy.asarray(posx) |
|
1397 | posx = numpy.asarray(posx) | |
| 1388 | posy = numpy.asarray(posy) |
|
1398 | posy = numpy.asarray(posy) | |
| 1389 |
|
1399 | |||
| 1390 | #Rotacion Inversa para alinear con el azimuth |
|
1400 | #Rotacion Inversa para alinear con el azimuth | |
| 1391 | if azimuth!= None: |
|
1401 | if azimuth!= None: | |
| 1392 | azimuth = azimuth*math.pi/180 |
|
1402 | azimuth = azimuth*math.pi/180 | |
| 1393 | posx1 = posx*math.cos(azimuth) + posy*math.sin(azimuth) |
|
1403 | posx1 = posx*math.cos(azimuth) + posy*math.sin(azimuth) | |
| 1394 | posy1 = -posx*math.sin(azimuth) + posy*math.cos(azimuth) |
|
1404 | posy1 = -posx*math.sin(azimuth) + posy*math.cos(azimuth) | |
| 1395 | else: |
|
1405 | else: | |
| 1396 | posx1 = posx |
|
1406 | posx1 = posx | |
| 1397 | posy1 = posy |
|
1407 | posy1 = posy | |
| 1398 |
|
1408 | |||
| 1399 | #Calculo de Distancias |
|
1409 | #Calculo de Distancias | |
| 1400 | distx = numpy.zeros(pairsCrossCorr.size) |
|
1410 | distx = numpy.zeros(pairsCrossCorr.size) | |
| 1401 | disty = numpy.zeros(pairsCrossCorr.size) |
|
1411 | disty = numpy.zeros(pairsCrossCorr.size) | |
| 1402 | dist = numpy.zeros(pairsCrossCorr.size) |
|
1412 | dist = numpy.zeros(pairsCrossCorr.size) | |
| 1403 | ang = numpy.zeros(pairsCrossCorr.size) |
|
1413 | ang = numpy.zeros(pairsCrossCorr.size) | |
| 1404 |
|
1414 | |||
| 1405 | for i in range(pairsCrossCorr.size): |
|
1415 | for i in range(pairsCrossCorr.size): | |
| 1406 | distx[i] = posx1[pairsList[pairsCrossCorr[i]][1]] - posx1[pairsList[pairsCrossCorr[i]][0]] |
|
1416 | distx[i] = posx1[pairsList[pairsCrossCorr[i]][1]] - posx1[pairsList[pairsCrossCorr[i]][0]] | |
| 1407 | disty[i] = posy1[pairsList[pairsCrossCorr[i]][1]] - posy1[pairsList[pairsCrossCorr[i]][0]] |
|
1417 | disty[i] = posy1[pairsList[pairsCrossCorr[i]][1]] - posy1[pairsList[pairsCrossCorr[i]][0]] | |
| 1408 | dist[i] = numpy.sqrt(distx[i]**2 + disty[i]**2) |
|
1418 | dist[i] = numpy.sqrt(distx[i]**2 + disty[i]**2) | |
| 1409 | ang[i] = numpy.arctan2(disty[i],distx[i]) |
|
1419 | ang[i] = numpy.arctan2(disty[i],distx[i]) | |
| 1410 | #Calculo de Matrices |
|
1420 | #Calculo de Matrices | |
| 1411 | nPairs = len(pairs) |
|
1421 | nPairs = len(pairs) | |
| 1412 | ang1 = numpy.zeros((nPairs, 2, 1)) |
|
1422 | ang1 = numpy.zeros((nPairs, 2, 1)) | |
| 1413 | dist1 = numpy.zeros((nPairs, 2, 1)) |
|
1423 | dist1 = numpy.zeros((nPairs, 2, 1)) | |
| 1414 |
|
1424 | |||
| 1415 | for j in range(nPairs): |
|
1425 | for j in range(nPairs): | |
| 1416 | dist1[j,0,0] = dist[pairs[j][0]] |
|
1426 | dist1[j,0,0] = dist[pairs[j][0]] | |
| 1417 | dist1[j,1,0] = dist[pairs[j][1]] |
|
1427 | dist1[j,1,0] = dist[pairs[j][1]] | |
| 1418 | ang1[j,0,0] = ang[pairs[j][0]] |
|
1428 | ang1[j,0,0] = ang[pairs[j][0]] | |
| 1419 | ang1[j,1,0] = ang[pairs[j][1]] |
|
1429 | ang1[j,1,0] = ang[pairs[j][1]] | |
| 1420 |
|
1430 | |||
| 1421 | return distx,disty, dist1,ang1 |
|
1431 | return distx,disty, dist1,ang1 | |
| 1422 |
|
1432 | |||
| 1423 | def __calculateVelVer(self, phase, lagTRange, _lambda): |
|
1433 | def __calculateVelVer(self, phase, lagTRange, _lambda): | |
| 1424 |
|
1434 | |||
| 1425 | Ts = lagTRange[1] - lagTRange[0] |
|
1435 | Ts = lagTRange[1] - lagTRange[0] | |
| 1426 | velW = -_lambda*phase/(4*math.pi*Ts) |
|
1436 | velW = -_lambda*phase/(4*math.pi*Ts) | |
| 1427 |
|
1437 | |||
| 1428 | return velW |
|
1438 | return velW | |
| 1429 |
|
1439 | |||
| 1430 | def __calculateVelHorDir(self, dist, tau1, tau2, ang): |
|
1440 | def __calculateVelHorDir(self, dist, tau1, tau2, ang): | |
| 1431 | nPairs = tau1.shape[0] |
|
1441 | nPairs = tau1.shape[0] | |
| 1432 | vel = numpy.zeros((nPairs,3,tau1.shape[2])) |
|
1442 | vel = numpy.zeros((nPairs,3,tau1.shape[2])) | |
| 1433 |
|
1443 | |||
| 1434 | angCos = numpy.cos(ang) |
|
1444 | angCos = numpy.cos(ang) | |
| 1435 | angSin = numpy.sin(ang) |
|
1445 | angSin = numpy.sin(ang) | |
| 1436 |
|
1446 | |||
| 1437 | vel0 = dist*tau1/(2*tau2**2) |
|
1447 | vel0 = dist*tau1/(2*tau2**2) | |
| 1438 | vel[:,0,:] = (vel0*angCos).sum(axis = 1) |
|
1448 | vel[:,0,:] = (vel0*angCos).sum(axis = 1) | |
| 1439 | vel[:,1,:] = (vel0*angSin).sum(axis = 1) |
|
1449 | vel[:,1,:] = (vel0*angSin).sum(axis = 1) | |
| 1440 |
|
1450 | |||
| 1441 | ind = numpy.where(numpy.isinf(vel)) |
|
1451 | ind = numpy.where(numpy.isinf(vel)) | |
| 1442 | vel[ind] = numpy.nan |
|
1452 | vel[ind] = numpy.nan | |
| 1443 |
|
1453 | |||
| 1444 | return vel |
|
1454 | return vel | |
| 1445 |
|
1455 | |||
| 1446 | def __getPairsAutoCorr(self, pairsList, nChannels): |
|
1456 | def __getPairsAutoCorr(self, pairsList, nChannels): | |
| 1447 |
|
1457 | |||
| 1448 | pairsAutoCorr = numpy.zeros(nChannels, dtype = 'int')*numpy.nan |
|
1458 | pairsAutoCorr = numpy.zeros(nChannels, dtype = 'int')*numpy.nan | |
| 1449 |
|
1459 | |||
| 1450 | for l in range(len(pairsList)): |
|
1460 | for l in range(len(pairsList)): | |
| 1451 | firstChannel = pairsList[l][0] |
|
1461 | firstChannel = pairsList[l][0] | |
| 1452 | secondChannel = pairsList[l][1] |
|
1462 | secondChannel = pairsList[l][1] | |
| 1453 |
|
1463 | |||
| 1454 | #Obteniendo pares de Autocorrelacion |
|
1464 | #Obteniendo pares de Autocorrelacion | |
| 1455 | if firstChannel == secondChannel: |
|
1465 | if firstChannel == secondChannel: | |
| 1456 | pairsAutoCorr[firstChannel] = int(l) |
|
1466 | pairsAutoCorr[firstChannel] = int(l) | |
| 1457 |
|
1467 | |||
| 1458 | pairsAutoCorr = pairsAutoCorr.astype(int) |
|
1468 | pairsAutoCorr = pairsAutoCorr.astype(int) | |
| 1459 |
|
1469 | |||
| 1460 | pairsCrossCorr = range(len(pairsList)) |
|
1470 | pairsCrossCorr = range(len(pairsList)) | |
| 1461 | pairsCrossCorr = numpy.delete(pairsCrossCorr,pairsAutoCorr) |
|
1471 | pairsCrossCorr = numpy.delete(pairsCrossCorr,pairsAutoCorr) | |
| 1462 |
|
1472 | |||
| 1463 | return pairsAutoCorr, pairsCrossCorr |
|
1473 | return pairsAutoCorr, pairsCrossCorr | |
| 1464 |
|
1474 | |||
| 1465 | def techniqueSA(self, pairsSelected, pairsList, nChannels, tau, azimuth, _lambda, position_x, position_y, lagTRange, correctFactor): |
|
1475 | def techniqueSA(self, pairsSelected, pairsList, nChannels, tau, azimuth, _lambda, position_x, position_y, lagTRange, correctFactor): | |
| 1466 | """ |
|
1476 | """ | |
| 1467 | Function that implements Spaced Antenna (SA) technique. |
|
1477 | Function that implements Spaced Antenna (SA) technique. | |
| 1468 |
|
1478 | |||
| 1469 | Input: Radial velocities, Direction cosines (x and y) of the Beam, Antenna azimuth, |
|
1479 | Input: Radial velocities, Direction cosines (x and y) of the Beam, Antenna azimuth, | |
| 1470 | Direction correction (if necessary), Ranges and SNR |
|
1480 | Direction correction (if necessary), Ranges and SNR | |
| 1471 |
|
1481 | |||
| 1472 | Output: Winds estimation (Zonal, Meridional and Vertical) |
|
1482 | Output: Winds estimation (Zonal, Meridional and Vertical) | |
| 1473 |
|
1483 | |||
| 1474 | Parameters affected: Winds |
|
1484 | Parameters affected: Winds | |
| 1475 | """ |
|
1485 | """ | |
| 1476 | #Cross Correlation pairs obtained |
|
1486 | #Cross Correlation pairs obtained | |
| 1477 | pairsAutoCorr, pairsCrossCorr = self.__getPairsAutoCorr(pairsList, nChannels) |
|
1487 | pairsAutoCorr, pairsCrossCorr = self.__getPairsAutoCorr(pairsList, nChannels) | |
| 1478 | pairsArray = numpy.array(pairsList)[pairsCrossCorr] |
|
1488 | pairsArray = numpy.array(pairsList)[pairsCrossCorr] | |
| 1479 | pairsSelArray = numpy.array(pairsSelected) |
|
1489 | pairsSelArray = numpy.array(pairsSelected) | |
| 1480 | pairs = [] |
|
1490 | pairs = [] | |
| 1481 |
|
1491 | |||
| 1482 | #Wind estimation pairs obtained |
|
1492 | #Wind estimation pairs obtained | |
| 1483 | for i in range(pairsSelArray.shape[0]/2): |
|
1493 | for i in range(pairsSelArray.shape[0]/2): | |
| 1484 | ind1 = numpy.where(numpy.all(pairsArray == pairsSelArray[2*i], axis = 1))[0][0] |
|
1494 | ind1 = numpy.where(numpy.all(pairsArray == pairsSelArray[2*i], axis = 1))[0][0] | |
| 1485 | ind2 = numpy.where(numpy.all(pairsArray == pairsSelArray[2*i + 1], axis = 1))[0][0] |
|
1495 | ind2 = numpy.where(numpy.all(pairsArray == pairsSelArray[2*i + 1], axis = 1))[0][0] | |
| 1486 | pairs.append((ind1,ind2)) |
|
1496 | pairs.append((ind1,ind2)) | |
| 1487 |
|
1497 | |||
| 1488 | indtau = tau.shape[0]/2 |
|
1498 | indtau = tau.shape[0]/2 | |
| 1489 | tau1 = tau[:indtau,:] |
|
1499 | tau1 = tau[:indtau,:] | |
| 1490 | tau2 = tau[indtau:-1,:] |
|
1500 | tau2 = tau[indtau:-1,:] | |
| 1491 | tau1 = tau1[pairs,:] |
|
1501 | tau1 = tau1[pairs,:] | |
| 1492 | tau2 = tau2[pairs,:] |
|
1502 | tau2 = tau2[pairs,:] | |
| 1493 | phase1 = tau[-1,:] |
|
1503 | phase1 = tau[-1,:] | |
| 1494 |
|
1504 | |||
| 1495 | #--------------------------------------------------------------------- |
|
1505 | #--------------------------------------------------------------------- | |
| 1496 | #Metodo Directo |
|
1506 | #Metodo Directo | |
| 1497 | distx, disty, dist, ang = self.__calculateDistance(position_x, position_y, pairsCrossCorr, pairsList, pairs,azimuth) |
|
1507 | distx, disty, dist, ang = self.__calculateDistance(position_x, position_y, pairsCrossCorr, pairsList, pairs,azimuth) | |
| 1498 | winds = self.__calculateVelHorDir(dist, tau1, tau2, ang) |
|
1508 | winds = self.__calculateVelHorDir(dist, tau1, tau2, ang) | |
| 1499 | winds = stats.nanmean(winds, axis=0) |
|
1509 | winds = stats.nanmean(winds, axis=0) | |
| 1500 | #--------------------------------------------------------------------- |
|
1510 | #--------------------------------------------------------------------- | |
| 1501 | #Metodo General |
|
1511 | #Metodo General | |
| 1502 | # distx, disty, dist = self.calculateDistance(position_x,position_y,pairsCrossCorr, pairsList, azimuth) |
|
1512 | # distx, disty, dist = self.calculateDistance(position_x,position_y,pairsCrossCorr, pairsList, azimuth) | |
| 1503 | # #Calculo Coeficientes de Funcion de Correlacion |
|
1513 | # #Calculo Coeficientes de Funcion de Correlacion | |
| 1504 | # F,G,A,B,H = self.calculateCoef(tau1,tau2,distx,disty,n) |
|
1514 | # F,G,A,B,H = self.calculateCoef(tau1,tau2,distx,disty,n) | |
| 1505 | # #Calculo de Velocidades |
|
1515 | # #Calculo de Velocidades | |
| 1506 | # winds = self.calculateVelUV(F,G,A,B,H) |
|
1516 | # winds = self.calculateVelUV(F,G,A,B,H) | |
| 1507 |
|
1517 | |||
| 1508 | #--------------------------------------------------------------------- |
|
1518 | #--------------------------------------------------------------------- | |
| 1509 | winds[2,:] = self.__calculateVelVer(phase1, lagTRange, _lambda) |
|
1519 | winds[2,:] = self.__calculateVelVer(phase1, lagTRange, _lambda) | |
| 1510 | winds = correctFactor*winds |
|
1520 | winds = correctFactor*winds | |
| 1511 | return winds |
|
1521 | return winds | |
| 1512 |
|
1522 | |||
| 1513 | def __checkTime(self, currentTime, paramInterval, outputInterval): |
|
1523 | def __checkTime(self, currentTime, paramInterval, outputInterval): | |
| 1514 |
|
1524 | |||
| 1515 | dataTime = currentTime + paramInterval |
|
1525 | dataTime = currentTime + paramInterval | |
| 1516 | deltaTime = dataTime - self.__initime |
|
1526 | deltaTime = dataTime - self.__initime | |
| 1517 |
|
1527 | |||
| 1518 | if deltaTime >= outputInterval or deltaTime < 0: |
|
1528 | if deltaTime >= outputInterval or deltaTime < 0: | |
| 1519 | self.__dataReady = True |
|
1529 | self.__dataReady = True | |
| 1520 | return |
|
1530 | return | |
| 1521 |
|
1531 | |||
| 1522 | def techniqueMeteors(self, arrayMeteor, meteorThresh, heightMin, heightMax): |
|
1532 | def techniqueMeteors(self, arrayMeteor, meteorThresh, heightMin, heightMax): | |
| 1523 | ''' |
|
1533 | ''' | |
| 1524 | Function that implements winds estimation technique with detected meteors. |
|
1534 | Function that implements winds estimation technique with detected meteors. | |
| 1525 |
|
1535 | |||
| 1526 | Input: Detected meteors, Minimum meteor quantity to wind estimation |
|
1536 | Input: Detected meteors, Minimum meteor quantity to wind estimation | |
| 1527 |
|
1537 | |||
| 1528 | Output: Winds estimation (Zonal and Meridional) |
|
1538 | Output: Winds estimation (Zonal and Meridional) | |
| 1529 |
|
1539 | |||
| 1530 | Parameters affected: Winds |
|
1540 | Parameters affected: Winds | |
| 1531 | ''' |
|
1541 | ''' | |
| 1532 | #Settings |
|
1542 | #Settings | |
| 1533 | nInt = (heightMax - heightMin)/2 |
|
1543 | nInt = (heightMax - heightMin)/2 | |
| 1534 | winds = numpy.zeros((2,nInt))*numpy.nan |
|
1544 | winds = numpy.zeros((2,nInt))*numpy.nan | |
| 1535 |
|
1545 | |||
| 1536 | #Filter errors |
|
1546 | #Filter errors | |
| 1537 | error = numpy.where(arrayMeteor[:,-1] == 0)[0] |
|
1547 | error = numpy.where(arrayMeteor[:,-1] == 0)[0] | |
| 1538 | finalMeteor = arrayMeteor[error,:] |
|
1548 | finalMeteor = arrayMeteor[error,:] | |
| 1539 |
|
1549 | |||
| 1540 | #Meteor Histogram |
|
1550 | #Meteor Histogram | |
| 1541 | finalHeights = finalMeteor[:,3] |
|
1551 | finalHeights = finalMeteor[:,3] | |
| 1542 | hist = numpy.histogram(finalHeights, bins = nInt, range = (heightMin,heightMax)) |
|
1552 | hist = numpy.histogram(finalHeights, bins = nInt, range = (heightMin,heightMax)) | |
| 1543 | nMeteorsPerI = hist[0] |
|
1553 | nMeteorsPerI = hist[0] | |
| 1544 | heightPerI = hist[1] |
|
1554 | heightPerI = hist[1] | |
| 1545 |
|
1555 | |||
| 1546 | #Sort of meteors |
|
1556 | #Sort of meteors | |
| 1547 | indSort = finalHeights.argsort() |
|
1557 | indSort = finalHeights.argsort() | |
| 1548 | finalMeteor2 = finalMeteor[indSort,:] |
|
1558 | finalMeteor2 = finalMeteor[indSort,:] | |
| 1549 |
|
1559 | |||
| 1550 | # Calculating winds |
|
1560 | # Calculating winds | |
| 1551 | ind1 = 0 |
|
1561 | ind1 = 0 | |
| 1552 | ind2 = 0 |
|
1562 | ind2 = 0 | |
| 1553 |
|
1563 | |||
| 1554 | for i in range(nInt): |
|
1564 | for i in range(nInt): | |
| 1555 | nMet = nMeteorsPerI[i] |
|
1565 | nMet = nMeteorsPerI[i] | |
| 1556 | ind1 = ind2 |
|
1566 | ind1 = ind2 | |
| 1557 | ind2 = ind1 + nMet |
|
1567 | ind2 = ind1 + nMet | |
| 1558 |
|
1568 | |||
| 1559 | meteorAux = finalMeteor2[ind1:ind2,:] |
|
1569 | meteorAux = finalMeteor2[ind1:ind2,:] | |
| 1560 |
|
1570 | |||
| 1561 | if meteorAux.shape[0] >= meteorThresh: |
|
1571 | if meteorAux.shape[0] >= meteorThresh: | |
| 1562 | vel = meteorAux[:, 7] |
|
1572 | vel = meteorAux[:, 7] | |
| 1563 | zen = meteorAux[:, 5]*numpy.pi/180 |
|
1573 | zen = meteorAux[:, 5]*numpy.pi/180 | |
| 1564 | azim = meteorAux[:, 4]*numpy.pi/180 |
|
1574 | azim = meteorAux[:, 4]*numpy.pi/180 | |
| 1565 |
|
1575 | |||
| 1566 | n = numpy.cos(zen) |
|
1576 | n = numpy.cos(zen) | |
| 1567 | # m = (1 - n**2)/(1 - numpy.tan(azim)**2) |
|
1577 | # m = (1 - n**2)/(1 - numpy.tan(azim)**2) | |
| 1568 | # l = m*numpy.tan(azim) |
|
1578 | # l = m*numpy.tan(azim) | |
| 1569 | l = numpy.sin(zen)*numpy.sin(azim) |
|
1579 | l = numpy.sin(zen)*numpy.sin(azim) | |
| 1570 | m = numpy.sin(zen)*numpy.cos(azim) |
|
1580 | m = numpy.sin(zen)*numpy.cos(azim) | |
| 1571 |
|
1581 | |||
| 1572 | A = numpy.vstack((l, m)).transpose() |
|
1582 | A = numpy.vstack((l, m)).transpose() | |
| 1573 | A1 = numpy.dot(numpy.linalg.inv( numpy.dot(A.transpose(),A) ),A.transpose()) |
|
1583 | A1 = numpy.dot(numpy.linalg.inv( numpy.dot(A.transpose(),A) ),A.transpose()) | |
| 1574 | windsAux = numpy.dot(A1, vel) |
|
1584 | windsAux = numpy.dot(A1, vel) | |
| 1575 |
|
1585 | |||
| 1576 | winds[0,i] = windsAux[0] |
|
1586 | winds[0,i] = windsAux[0] | |
| 1577 | winds[1,i] = windsAux[1] |
|
1587 | winds[1,i] = windsAux[1] | |
| 1578 |
|
1588 | |||
| 1579 | return winds, heightPerI[:-1] |
|
1589 | return winds, heightPerI[:-1] | |
| 1580 |
|
1590 | |||
| 1581 | def run(self, dataOut, technique, **kwargs): |
|
1591 | def run(self, dataOut, technique, **kwargs): | |
| 1582 |
|
1592 | |||
| 1583 | param = dataOut.data_param |
|
1593 | param = dataOut.data_param | |
| 1584 | if dataOut.abscissaList != None: |
|
1594 | if dataOut.abscissaList != None: | |
| 1585 | absc = dataOut.abscissaList[:-1] |
|
1595 | absc = dataOut.abscissaList[:-1] | |
| 1586 | noise = dataOut.noise |
|
1596 | noise = dataOut.noise | |
| 1587 | heightList = dataOut.getHeiRange() |
|
1597 | heightList = dataOut.getHeiRange() | |
| 1588 | SNR = dataOut.data_SNR |
|
1598 | SNR = dataOut.data_SNR | |
| 1589 |
|
1599 | |||
| 1590 | if technique == 'DBS': |
|
1600 | if technique == 'DBS': | |
| 1591 |
|
1601 | |||
| 1592 | if kwargs.has_key('dirCosx') and kwargs.has_key('dirCosy'): |
|
1602 | if kwargs.has_key('dirCosx') and kwargs.has_key('dirCosy'): | |
| 1593 | theta_x = numpy.array(kwargs['dirCosx']) |
|
1603 | theta_x = numpy.array(kwargs['dirCosx']) | |
| 1594 | theta_y = numpy.array(kwargs['dirCosy']) |
|
1604 | theta_y = numpy.array(kwargs['dirCosy']) | |
| 1595 | else: |
|
1605 | else: | |
| 1596 | elev = numpy.array(kwargs['elevation']) |
|
1606 | elev = numpy.array(kwargs['elevation']) | |
| 1597 | azim = numpy.array(kwargs['azimuth']) |
|
1607 | azim = numpy.array(kwargs['azimuth']) | |
| 1598 | theta_x, theta_y = self.__calculateCosDir(elev, azim) |
|
1608 | theta_x, theta_y = self.__calculateCosDir(elev, azim) | |
| 1599 | azimuth = kwargs['correctAzimuth'] |
|
1609 | azimuth = kwargs['correctAzimuth'] | |
| 1600 | if kwargs.has_key('horizontalOnly'): |
|
1610 | if kwargs.has_key('horizontalOnly'): | |
| 1601 | horizontalOnly = kwargs['horizontalOnly'] |
|
1611 | horizontalOnly = kwargs['horizontalOnly'] | |
| 1602 | else: horizontalOnly = False |
|
1612 | else: horizontalOnly = False | |
| 1603 | if kwargs.has_key('correctFactor'): |
|
1613 | if kwargs.has_key('correctFactor'): | |
| 1604 | correctFactor = kwargs['correctFactor'] |
|
1614 | correctFactor = kwargs['correctFactor'] | |
| 1605 | else: correctFactor = 1 |
|
1615 | else: correctFactor = 1 | |
| 1606 | if kwargs.has_key('channelList'): |
|
1616 | if kwargs.has_key('channelList'): | |
| 1607 | channelList = kwargs['channelList'] |
|
1617 | channelList = kwargs['channelList'] | |
| 1608 | if len(channelList) == 2: |
|
1618 | if len(channelList) == 2: | |
| 1609 | horizontalOnly = True |
|
1619 | horizontalOnly = True | |
| 1610 | arrayChannel = numpy.array(channelList) |
|
1620 | arrayChannel = numpy.array(channelList) | |
| 1611 | param = param[arrayChannel,:,:] |
|
1621 | param = param[arrayChannel,:,:] | |
| 1612 | theta_x = theta_x[arrayChannel] |
|
1622 | theta_x = theta_x[arrayChannel] | |
| 1613 | theta_y = theta_y[arrayChannel] |
|
1623 | theta_y = theta_y[arrayChannel] | |
| 1614 |
|
1624 | |||
| 1615 | velRadial0 = param[:,1,:] #Radial velocity |
|
1625 | velRadial0 = param[:,1,:] #Radial velocity | |
| 1616 | dataOut.data_output, dataOut.heightList, dataOut.data_SNR = self.techniqueDBS(velRadial0, theta_x, theta_y, azimuth, correctFactor, horizontalOnly, heightList, SNR) #DBS Function |
|
1626 | dataOut.data_output, dataOut.heightList, dataOut.data_SNR = self.techniqueDBS(velRadial0, theta_x, theta_y, azimuth, correctFactor, horizontalOnly, heightList, SNR) #DBS Function | |
| 1617 | dataOut.utctimeInit = dataOut.utctime |
|
1627 | dataOut.utctimeInit = dataOut.utctime | |
| 1618 | dataOut.outputInterval = dataOut.timeInterval |
|
1628 | dataOut.outputInterval = dataOut.timeInterval | |
| 1619 |
|
1629 | |||
| 1620 | elif technique == 'SA': |
|
1630 | elif technique == 'SA': | |
| 1621 |
|
1631 | |||
| 1622 | #Parameters |
|
1632 | #Parameters | |
| 1623 | position_x = kwargs['positionX'] |
|
1633 | position_x = kwargs['positionX'] | |
| 1624 | position_y = kwargs['positionY'] |
|
1634 | position_y = kwargs['positionY'] | |
| 1625 | azimuth = kwargs['azimuth'] |
|
1635 | azimuth = kwargs['azimuth'] | |
| 1626 |
|
1636 | |||
| 1627 | if kwargs.has_key('crosspairsList'): |
|
1637 | if kwargs.has_key('crosspairsList'): | |
| 1628 | pairs = kwargs['crosspairsList'] |
|
1638 | pairs = kwargs['crosspairsList'] | |
| 1629 | else: |
|
1639 | else: | |
| 1630 | pairs = None |
|
1640 | pairs = None | |
| 1631 |
|
1641 | |||
| 1632 | if kwargs.has_key('correctFactor'): |
|
1642 | if kwargs.has_key('correctFactor'): | |
| 1633 | correctFactor = kwargs['correctFactor'] |
|
1643 | correctFactor = kwargs['correctFactor'] | |
| 1634 | else: |
|
1644 | else: | |
| 1635 | correctFactor = 1 |
|
1645 | correctFactor = 1 | |
| 1636 |
|
1646 | |||
| 1637 | tau = dataOut.data_param |
|
1647 | tau = dataOut.data_param | |
| 1638 | _lambda = dataOut.C/dataOut.frequency |
|
1648 | _lambda = dataOut.C/dataOut.frequency | |
| 1639 | pairsList = dataOut.groupList |
|
1649 | pairsList = dataOut.groupList | |
| 1640 | nChannels = dataOut.nChannels |
|
1650 | nChannels = dataOut.nChannels | |
| 1641 |
|
1651 | |||
| 1642 | dataOut.data_output = self.techniqueSA(pairs, pairsList, nChannels, tau, azimuth, _lambda, position_x, position_y, absc, correctFactor) |
|
1652 | dataOut.data_output = self.techniqueSA(pairs, pairsList, nChannels, tau, azimuth, _lambda, position_x, position_y, absc, correctFactor) | |
| 1643 | dataOut.utctimeInit = dataOut.utctime |
|
1653 | dataOut.utctimeInit = dataOut.utctime | |
| 1644 | dataOut.outputInterval = dataOut.timeInterval |
|
1654 | dataOut.outputInterval = dataOut.timeInterval | |
| 1645 |
|
1655 | |||
| 1646 | elif technique == 'Meteors': |
|
1656 | elif technique == 'Meteors': | |
| 1647 | dataOut.flagNoData = True |
|
1657 | dataOut.flagNoData = True | |
| 1648 | self.__dataReady = False |
|
1658 | self.__dataReady = False | |
| 1649 |
|
1659 | |||
| 1650 | if kwargs.has_key('nHours'): |
|
1660 | if kwargs.has_key('nHours'): | |
| 1651 | nHours = kwargs['nHours'] |
|
1661 | nHours = kwargs['nHours'] | |
| 1652 | else: |
|
1662 | else: | |
| 1653 | nHours = 1 |
|
1663 | nHours = 1 | |
| 1654 |
|
1664 | |||
| 1655 | if kwargs.has_key('meteorsPerBin'): |
|
1665 | if kwargs.has_key('meteorsPerBin'): | |
| 1656 | meteorThresh = kwargs['meteorsPerBin'] |
|
1666 | meteorThresh = kwargs['meteorsPerBin'] | |
| 1657 | else: |
|
1667 | else: | |
| 1658 | meteorThresh = 6 |
|
1668 | meteorThresh = 6 | |
| 1659 |
|
1669 | |||
| 1660 | if kwargs.has_key('hmin'): |
|
1670 | if kwargs.has_key('hmin'): | |
| 1661 | hmin = kwargs['hmin'] |
|
1671 | hmin = kwargs['hmin'] | |
| 1662 | else: hmin = 70 |
|
1672 | else: hmin = 70 | |
| 1663 | if kwargs.has_key('hmax'): |
|
1673 | if kwargs.has_key('hmax'): | |
| 1664 | hmax = kwargs['hmax'] |
|
1674 | hmax = kwargs['hmax'] | |
| 1665 | else: hmax = 110 |
|
1675 | else: hmax = 110 | |
| 1666 |
|
1676 | |||
| 1667 | dataOut.outputInterval = nHours*3600 |
|
1677 | dataOut.outputInterval = nHours*3600 | |
| 1668 |
|
1678 | |||
| 1669 | if self.__isConfig == False: |
|
1679 | if self.__isConfig == False: | |
| 1670 | # self.__initime = dataOut.datatime.replace(minute = 0, second = 0, microsecond = 03) |
|
1680 | # self.__initime = dataOut.datatime.replace(minute = 0, second = 0, microsecond = 03) | |
| 1671 | #Get Initial LTC time |
|
1681 | #Get Initial LTC time | |
| 1672 | self.__initime = datetime.datetime.utcfromtimestamp(self.dataOut.utctime) |
|
1682 | self.__initime = datetime.datetime.utcfromtimestamp(self.dataOut.utctime) | |
| 1673 | self.__initime = (self.__initime.replace(minute = 0, second = 0, microsecond = 0) - datetime.datetime(1970, 1, 1)).total_seconds() |
|
1683 | self.__initime = (self.__initime.replace(minute = 0, second = 0, microsecond = 0) - datetime.datetime(1970, 1, 1)).total_seconds() | |
| 1674 |
|
1684 | |||
| 1675 | self.__isConfig = True |
|
1685 | self.__isConfig = True | |
| 1676 |
|
1686 | |||
| 1677 | if self.__buffer == None: |
|
1687 | if self.__buffer == None: | |
| 1678 | self.__buffer = dataOut.data_param |
|
1688 | self.__buffer = dataOut.data_param | |
| 1679 | self.__firstdata = copy.copy(dataOut) |
|
1689 | self.__firstdata = copy.copy(dataOut) | |
| 1680 |
|
1690 | |||
| 1681 | else: |
|
1691 | else: | |
| 1682 | self.__buffer = numpy.vstack((self.__buffer, dataOut.data_param)) |
|
1692 | self.__buffer = numpy.vstack((self.__buffer, dataOut.data_param)) | |
| 1683 |
|
1693 | |||
| 1684 | self.__checkTime(dataOut.utctime, dataOut.paramInterval, dataOut.outputInterval) #Check if the buffer is ready |
|
1694 | self.__checkTime(dataOut.utctime, dataOut.paramInterval, dataOut.outputInterval) #Check if the buffer is ready | |
| 1685 |
|
1695 | |||
| 1686 | if self.__dataReady: |
|
1696 | if self.__dataReady: | |
| 1687 | dataOut.utctimeInit = self.__initime |
|
1697 | dataOut.utctimeInit = self.__initime | |
| 1688 |
|
1698 | |||
| 1689 | self.__initime += dataOut.outputInterval #to erase time offset |
|
1699 | self.__initime += dataOut.outputInterval #to erase time offset | |
| 1690 |
|
1700 | |||
| 1691 | dataOut.data_output, dataOut.heightList = self.techniqueMeteors(self.__buffer, meteorThresh, hmin, hmax) |
|
1701 | dataOut.data_output, dataOut.heightList = self.techniqueMeteors(self.__buffer, meteorThresh, hmin, hmax) | |
| 1692 | dataOut.flagNoData = False |
|
1702 | dataOut.flagNoData = False | |
| 1693 | self.__buffer = None |
|
1703 | self.__buffer = None | |
| 1694 |
|
1704 | |||
| 1695 | return |
|
1705 | return | |
| 1696 |
|
1706 | |||
| 1697 | class EWDriftsEstimation(Operation): |
|
1707 | class EWDriftsEstimation(Operation): | |
| 1698 |
|
1708 | |||
| 1699 |
|
1709 | |||
| 1700 | def __init__(self): |
|
1710 | def __init__(self): | |
| 1701 | Operation.__init__(self) |
|
1711 | Operation.__init__(self) | |
| 1702 |
|
1712 | |||
| 1703 | def __correctValues(self, heiRang, phi, velRadial, SNR): |
|
1713 | def __correctValues(self, heiRang, phi, velRadial, SNR): | |
| 1704 | listPhi = phi.tolist() |
|
1714 | listPhi = phi.tolist() | |
| 1705 | maxid = listPhi.index(max(listPhi)) |
|
1715 | maxid = listPhi.index(max(listPhi)) | |
| 1706 | minid = listPhi.index(min(listPhi)) |
|
1716 | minid = listPhi.index(min(listPhi)) | |
| 1707 |
|
1717 | |||
| 1708 | rango = range(len(phi)) |
|
1718 | rango = range(len(phi)) | |
| 1709 | # rango = numpy.delete(rango,maxid) |
|
1719 | # rango = numpy.delete(rango,maxid) | |
| 1710 |
|
1720 | |||
| 1711 | heiRang1 = heiRang*math.cos(phi[maxid]) |
|
1721 | heiRang1 = heiRang*math.cos(phi[maxid]) | |
| 1712 | heiRangAux = heiRang*math.cos(phi[minid]) |
|
1722 | heiRangAux = heiRang*math.cos(phi[minid]) | |
| 1713 | indOut = (heiRang1 < heiRangAux[0]).nonzero() |
|
1723 | indOut = (heiRang1 < heiRangAux[0]).nonzero() | |
| 1714 | heiRang1 = numpy.delete(heiRang1,indOut) |
|
1724 | heiRang1 = numpy.delete(heiRang1,indOut) | |
| 1715 |
|
1725 | |||
| 1716 | velRadial1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
1726 | velRadial1 = numpy.zeros([len(phi),len(heiRang1)]) | |
| 1717 | SNR1 = numpy.zeros([len(phi),len(heiRang1)]) |
|
1727 | SNR1 = numpy.zeros([len(phi),len(heiRang1)]) | |
| 1718 |
|
1728 | |||
| 1719 | for i in rango: |
|
1729 | for i in rango: | |
| 1720 | x = heiRang*math.cos(phi[i]) |
|
1730 | x = heiRang*math.cos(phi[i]) | |
| 1721 | y1 = velRadial[i,:] |
|
1731 | y1 = velRadial[i,:] | |
| 1722 | f1 = interpolate.interp1d(x,y1,kind = 'cubic') |
|
1732 | f1 = interpolate.interp1d(x,y1,kind = 'cubic') | |
| 1723 |
|
1733 | |||
| 1724 | x1 = heiRang1 |
|
1734 | x1 = heiRang1 | |
| 1725 | y11 = f1(x1) |
|
1735 | y11 = f1(x1) | |
| 1726 |
|
1736 | |||
| 1727 | y2 = SNR[i,:] |
|
1737 | y2 = SNR[i,:] | |
| 1728 | f2 = interpolate.interp1d(x,y2,kind = 'cubic') |
|
1738 | f2 = interpolate.interp1d(x,y2,kind = 'cubic') | |
| 1729 | y21 = f2(x1) |
|
1739 | y21 = f2(x1) | |
| 1730 |
|
1740 | |||
| 1731 | velRadial1[i,:] = y11 |
|
1741 | velRadial1[i,:] = y11 | |
| 1732 | SNR1[i,:] = y21 |
|
1742 | SNR1[i,:] = y21 | |
| 1733 |
|
1743 | |||
| 1734 | return heiRang1, velRadial1, SNR1 |
|
1744 | return heiRang1, velRadial1, SNR1 | |
| 1735 |
|
1745 | |||
| 1736 | def run(self, dataOut, zenith, zenithCorrection): |
|
1746 | def run(self, dataOut, zenith, zenithCorrection): | |
| 1737 | heiRang = dataOut.heightList |
|
1747 | heiRang = dataOut.heightList | |
| 1738 | velRadial = dataOut.data_param[:,3,:] |
|
1748 | velRadial = dataOut.data_param[:,3,:] | |
| 1739 | SNR = dataOut.data_SNR |
|
1749 | SNR = dataOut.data_SNR | |
| 1740 |
|
1750 | |||
| 1741 | zenith = numpy.array(zenith) |
|
1751 | zenith = numpy.array(zenith) | |
| 1742 | zenith -= zenithCorrection |
|
1752 | zenith -= zenithCorrection | |
| 1743 | zenith *= numpy.pi/180 |
|
1753 | zenith *= numpy.pi/180 | |
| 1744 |
|
1754 | |||
| 1745 | heiRang1, velRadial1, SNR1 = self.__correctValues(heiRang, numpy.abs(zenith), velRadial, SNR) |
|
1755 | heiRang1, velRadial1, SNR1 = self.__correctValues(heiRang, numpy.abs(zenith), velRadial, SNR) | |
| 1746 |
|
1756 | |||
| 1747 | alp = zenith[0] |
|
1757 | alp = zenith[0] | |
| 1748 | bet = zenith[1] |
|
1758 | bet = zenith[1] | |
| 1749 |
|
1759 | |||
| 1750 | w_w = velRadial1[0,:] |
|
1760 | w_w = velRadial1[0,:] | |
| 1751 | w_e = velRadial1[1,:] |
|
1761 | w_e = velRadial1[1,:] | |
| 1752 |
|
1762 | |||
| 1753 | w = (w_w*numpy.sin(bet) - w_e*numpy.sin(alp))/(numpy.cos(alp)*numpy.sin(bet) - numpy.cos(bet)*numpy.sin(alp)) |
|
1763 | w = (w_w*numpy.sin(bet) - w_e*numpy.sin(alp))/(numpy.cos(alp)*numpy.sin(bet) - numpy.cos(bet)*numpy.sin(alp)) | |
| 1754 | u = (w_w*numpy.cos(bet) - w_e*numpy.cos(alp))/(numpy.sin(alp)*numpy.cos(bet) - numpy.sin(bet)*numpy.cos(alp)) |
|
1764 | u = (w_w*numpy.cos(bet) - w_e*numpy.cos(alp))/(numpy.sin(alp)*numpy.cos(bet) - numpy.sin(bet)*numpy.cos(alp)) | |
| 1755 |
|
1765 | |||
| 1756 | winds = numpy.vstack((u,w)) |
|
1766 | winds = numpy.vstack((u,w)) | |
| 1757 |
|
1767 | |||
| 1758 | dataOut.heightList = heiRang1 |
|
1768 | dataOut.heightList = heiRang1 | |
| 1759 | dataOut.data_output = winds |
|
1769 | dataOut.data_output = winds | |
| 1760 | dataOut.data_SNR = SNR1 |
|
1770 | dataOut.data_SNR = SNR1 | |
| 1761 |
|
1771 | |||
| 1762 | dataOut.utctimeInit = dataOut.utctime |
|
1772 | dataOut.utctimeInit = dataOut.utctime | |
| 1763 | dataOut.outputInterval = dataOut.timeInterval |
|
1773 | dataOut.outputInterval = dataOut.timeInterval | |
| 1764 | return |
|
1774 | return | |
| 1765 |
|
1775 | |||
| 1766 |
|
1776 | |||
| 1767 |
|
1777 | |||
| 1768 |
|
1778 | |||
| 1769 |
|
1779 | |||
| 1770 | No newline at end of file |
|
1780 | ||
General Comments 0
You need to be logged in to leave comments.
Login now