1 | #include <fstream> |
---|
2 | #include <iostream> |
---|
3 | #include <iomanip> |
---|
4 | #include <vector> |
---|
5 | #include <Cubes/cubes.hh> |
---|
6 | #include <ATrous/atrous.hh> |
---|
7 | #include <Utils/utils.hh> |
---|
8 | |
---|
9 | using std::setw; |
---|
10 | |
---|
11 | ////////////////////////////////////////////////////////////////////////////////////// |
---|
12 | /** |
---|
13 | * Cube::ReconSearch1D() |
---|
14 | * This reconstructs a cube by performing a 1D a trous reconstruction |
---|
15 | * in the spectrum of each spatial pixel. |
---|
16 | * It then searches the cube using reconSearch (below). |
---|
17 | * |
---|
18 | * The resulting object list is stored in the Cube. |
---|
19 | */ |
---|
20 | void Cube::ReconSearch1D() |
---|
21 | { |
---|
22 | long xySize = axisDim[0] * axisDim[1]; |
---|
23 | long zdim = axisDim[2]; |
---|
24 | |
---|
25 | // Reconstruct the cube by 1d atrous transform in each spatial pixel |
---|
26 | std::cout<<"Reconstructing... "; |
---|
27 | for(int npix=0; npix<xySize; npix++){ |
---|
28 | if((100*npix/xySize)%5==0) |
---|
29 | std::cout<<setw(3)<<100*npix/xySize<<"% done"<<"\b\b\b\b\b\b\b\b\b"<<std::flush; |
---|
30 | float *spec = new float[zdim]; |
---|
31 | float *newSpec = new float[zdim]; |
---|
32 | for(int z=0;z<zdim;z++){ |
---|
33 | int cubepos = z*xySize + npix; |
---|
34 | spec[z] = this->array[cubepos]; |
---|
35 | } |
---|
36 | atrous1DReconstruct(axisDim[2],spec,newSpec,this->par); |
---|
37 | for(int z=0;z<zdim;z++){ |
---|
38 | int cubepos = z*xySize + npix; |
---|
39 | this->recon[cubepos] = newSpec[z]; |
---|
40 | } |
---|
41 | delete spec; |
---|
42 | delete newSpec; |
---|
43 | } |
---|
44 | this->reconExists = true; |
---|
45 | std::cout<<"All Done. Searching... "; |
---|
46 | |
---|
47 | this->objectList = reconSearch(this->axisDim,this->array,this->recon,this->par); |
---|
48 | this->updateDetectMap(); |
---|
49 | if(this->par.getFlagLog()) this->logDetectionList(); |
---|
50 | |
---|
51 | } |
---|
52 | |
---|
53 | ////////////////////////////////////////////////////////////////////////////////////// |
---|
54 | /** |
---|
55 | * Cube::ReconSearch2D() |
---|
56 | * This reconstructs a cube by performing a 2D a trous reconstruction |
---|
57 | * in each spatial image of the cube. |
---|
58 | * It then searches the cube using reconSearch (below). |
---|
59 | * |
---|
60 | * The resulting object list is stored in the Cube. |
---|
61 | */ |
---|
62 | void Cube::ReconSearch2D() |
---|
63 | { |
---|
64 | long xySize = axisDim[0] * axisDim[1]; |
---|
65 | |
---|
66 | // RECONSTRUCT THE CUBE BY 2D ATROUS TRANSFORM IN EACH CHANNEL |
---|
67 | bool *doChannel = new bool[axisDim[2]]; |
---|
68 | for(int z=0;z<axisDim[2];z++) |
---|
69 | doChannel[z] = !( this->par.getFlagMW() && (z>=this->par.getMinMW()) && (z<=this->par.getMaxMW()) ); |
---|
70 | std::cout<<"Reconstructing... "; |
---|
71 | for(int z=0;z<axisDim[2];z++){ |
---|
72 | if((100*z/axisDim[2])%5==0) |
---|
73 | std::cout<<setw(3)<<100*z/axisDim[2]<<"% done"<<"\b\b\b\b\b\b\b\b\b"<<std::flush; |
---|
74 | if(doChannel[z]){ |
---|
75 | float *im = new float[xySize]; |
---|
76 | float *newIm = new float[xySize]; |
---|
77 | for(int npix=0; npix<xySize; npix++){ |
---|
78 | int cubepos = z*xySize + npix; |
---|
79 | im[npix] = this->array[cubepos]; |
---|
80 | } |
---|
81 | atrous2DReconstruct(axisDim[0],axisDim[1],im,newIm,this->par); |
---|
82 | for(int npix=0; npix<xySize; npix++){ |
---|
83 | int cubepos = z*xySize + npix; |
---|
84 | this->recon[cubepos] = newIm[npix]; |
---|
85 | } |
---|
86 | delete im; |
---|
87 | delete newIm; |
---|
88 | } |
---|
89 | else |
---|
90 | for(int i=0; i<xySize; i++) this->recon[z*xySize+i] = this->array[z*xySize+i]; |
---|
91 | } |
---|
92 | this->reconExists = true; |
---|
93 | std::cout<<"All Done. \nSearching... "; |
---|
94 | |
---|
95 | this->objectList = reconSearch(this->axisDim,this->array,this->recon,this->par); |
---|
96 | |
---|
97 | this->updateDetectMap(); |
---|
98 | if(this->par.getFlagLog()) this->logDetectionList(); |
---|
99 | |
---|
100 | delete [] doChannel; |
---|
101 | |
---|
102 | } |
---|
103 | |
---|
104 | ////////////////////////////////////////////////////////////////////////////////////// |
---|
105 | /** |
---|
106 | * Cube::ReconSearch3D() |
---|
107 | * This reconstructs a cube by performing a full 3D a trous |
---|
108 | * reconstruction of the cube |
---|
109 | * It then searches the cube using reconSearch (below). |
---|
110 | * |
---|
111 | * The resulting object list is stored in the Cube. |
---|
112 | */ |
---|
113 | void Cube::ReconSearch3D() |
---|
114 | { |
---|
115 | if(this->axisDim[2]==1) this->ReconSearch2D(); |
---|
116 | else { |
---|
117 | |
---|
118 | std::cout<<" Reconstructing... "<<std::flush; |
---|
119 | atrous3DReconstruct(this->axisDim[0],this->axisDim[1],this->axisDim[2], |
---|
120 | this->array,this->recon,this->par); |
---|
121 | this->reconExists = true; |
---|
122 | std::cout<<"All Done. \n Searching... "<<std::flush; |
---|
123 | |
---|
124 | this->objectList = reconSearch(this->axisDim,this->array,this->recon,this->par); |
---|
125 | |
---|
126 | this->updateDetectMap(); |
---|
127 | if(this->par.getFlagLog()) this->logDetectionList(); |
---|
128 | |
---|
129 | } |
---|
130 | |
---|
131 | } |
---|
132 | |
---|
133 | |
---|
134 | ////////////////////////////////////////////////////////////////////////////////////// |
---|
135 | /** |
---|
136 | * reconSearch(long *dim, float *originalArray, float *reconArray, Param &par) |
---|
137 | * This searches for objects in a cube that has been reconstructed. |
---|
138 | * |
---|
139 | * Inputs: - dimension array |
---|
140 | * - original, un-reconstructed image array |
---|
141 | * - reconstructed image array |
---|
142 | * - parameters |
---|
143 | * |
---|
144 | * Searches first in each spatial pixel (1D search), |
---|
145 | * then in each channel image (2D search). |
---|
146 | * |
---|
147 | * Returns: vector of Detections resulting from the search. |
---|
148 | */ |
---|
149 | vector <Detection> reconSearch(long *dim, float *originalArray, float *reconArray, Param &par) |
---|
150 | { |
---|
151 | vector <Detection> outputList; |
---|
152 | int zdim = dim[2]; |
---|
153 | int xySize = dim[0] * dim[1]; |
---|
154 | int fullSize = zdim * xySize; |
---|
155 | int num, goodSize; |
---|
156 | |
---|
157 | // bool flagBlank=par.getFlagBlankPix(); |
---|
158 | float blankPixValue = par.getBlankPixVal(); |
---|
159 | bool *isGood = new bool[fullSize]; |
---|
160 | for(int pos=0;pos<fullSize;pos++) |
---|
161 | isGood[pos] = !par.isBlank(originalArray[pos]); |
---|
162 | // isGood[pos] = (!flagBlank) || (originalArray[pos]!=blankPixValue); |
---|
163 | |
---|
164 | float dud; |
---|
165 | |
---|
166 | // First search -- in each spectrum. |
---|
167 | // First, get stats |
---|
168 | if(zdim > 1){ |
---|
169 | std::cout << "1D: | |" << std::flush; |
---|
170 | // if(par.isVerbose()) std::cout << "Done 0%" << "\b\b\b\b\b\b\b\b" << std::flush; |
---|
171 | |
---|
172 | float *specMedian = new float[xySize]; |
---|
173 | float *specSigma = new float[xySize]; |
---|
174 | float *spec = new float[zdim]; |
---|
175 | |
---|
176 | for(int npix=0; npix<xySize; npix++){ |
---|
177 | goodSize=0; |
---|
178 | for(int z=0;z<zdim;z++) |
---|
179 | if(isGood[z*xySize+npix]) spec[goodSize++] = originalArray[z*xySize+npix]; |
---|
180 | if(goodSize>0) findMedianStats(spec,goodSize,specMedian[npix],dud); |
---|
181 | else specMedian[npix] = blankPixValue; |
---|
182 | goodSize=0; |
---|
183 | for(int z=0;z<zdim;z++) |
---|
184 | if(isGood[z*xySize+npix]) |
---|
185 | spec[goodSize++] = originalArray[z*xySize+npix]-reconArray[z*xySize+npix]; |
---|
186 | if(goodSize>0) findNormalStats(spec,goodSize,dud,specSigma[npix]); |
---|
187 | else specSigma[npix] = 1.; |
---|
188 | } |
---|
189 | |
---|
190 | // Next, do source finding. |
---|
191 | long *specdim = new long[2]; |
---|
192 | specdim[0] = zdim; specdim[1]=1; |
---|
193 | Image *spectrum = new Image(specdim); |
---|
194 | spectrum->saveParam(par); |
---|
195 | spectrum->pars().setBeamSize(2.); // for spectrum, only neighbouring channels correlated |
---|
196 | for(int npix=0; npix<xySize; npix++){ |
---|
197 | |
---|
198 | // if(par.isVerbose() && ((1000*npix/xySize)%10==0) ) |
---|
199 | // std::cout << "Done " << setw(2) << 100*npix/xySize << "%\b\b\b\b\b\b\b\b" << std::flush; |
---|
200 | if( par.isVerbose() && ((100*(npix+1)/xySize)%5 == 0) ){ |
---|
201 | std::cout << "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b|"; |
---|
202 | for(int i=0;i<(100*(npix+1)/xySize)/5;i++) std::cout << "#"; |
---|
203 | for(int i=(100*(npix+1)/xySize)/5;i<20;i++) std::cout << " "; |
---|
204 | std::cout << "|" << std::flush; |
---|
205 | } |
---|
206 | |
---|
207 | for(int z=0;z<zdim;z++) spec[z] = reconArray[z*xySize + npix]; |
---|
208 | spectrum->saveArray(spec,zdim); |
---|
209 | spectrum->setStats(specMedian[npix],specSigma[npix],par.getCut()); |
---|
210 | if(par.getFlagFDR()) spectrum->setupFDR(); |
---|
211 | spectrum->setMinSize(par.getMinChannels()); |
---|
212 | spectrum->lutz_detect(); |
---|
213 | for(int obj=0;obj<spectrum->getNumObj();obj++){ |
---|
214 | Detection *object = new Detection; |
---|
215 | *object = spectrum->getObject(obj); |
---|
216 | // if(par.getFlagGrowth()) growObject(*object,*spectrum); |
---|
217 | for(int pix=0;pix<object->getSize();pix++) { |
---|
218 | // Fix up coordinates of each pixel to match original array |
---|
219 | object->setZ(pix, object->getX(pix)); |
---|
220 | object->setX(pix, npix%dim[0]); |
---|
221 | object->setY(pix, npix/dim[0]); |
---|
222 | object->setF(pix, originalArray[object->getX(pix)+object->getY(pix)*dim[0]+object->getZ(pix)*xySize]); |
---|
223 | // NB: set F to the original value, not the recon value. |
---|
224 | } |
---|
225 | object->addOffsets(par); |
---|
226 | object->calcParams(); |
---|
227 | // outputList.push_back(*object); |
---|
228 | mergeIntoList(*object,outputList,par); |
---|
229 | delete object; |
---|
230 | } |
---|
231 | spectrum->clearDetectionList(); |
---|
232 | } |
---|
233 | delete [] spec; |
---|
234 | delete [] specdim; |
---|
235 | delete spectrum; |
---|
236 | delete [] specMedian; |
---|
237 | delete [] specSigma; |
---|
238 | |
---|
239 | num = outputList.size(); |
---|
240 | std::cout <<"\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\bFound " << num <<"; " << std::flush; |
---|
241 | |
---|
242 | } |
---|
243 | |
---|
244 | // Second search -- in each channel |
---|
245 | std::cout << "2D: | |" << std::flush; |
---|
246 | // if(par.isVerbose()) std::cout << "Done 0%" << "\b\b\b\b\b\b\b\b" << std::flush; |
---|
247 | float *imageMedian = new float[zdim]; |
---|
248 | float *imageSigma = new float[zdim]; |
---|
249 | float *image = new float[xySize]; |
---|
250 | // First, get stats |
---|
251 | for(int z=0; z<zdim; z++){ |
---|
252 | goodSize=0; |
---|
253 | for(int npix=0; npix<xySize; npix++) |
---|
254 | if(isGood[z*xySize + npix]) image[goodSize++] = originalArray[z*xySize + npix]; |
---|
255 | if(goodSize>0) findMedianStats(image,goodSize,imageMedian[z],dud); |
---|
256 | else imageMedian[z] = blankPixValue; |
---|
257 | goodSize=0; |
---|
258 | for(int npix=0; npix<xySize; npix++) |
---|
259 | if(isGood[z*xySize+npix]) |
---|
260 | image[goodSize++]=originalArray[z*xySize+npix]-reconArray[z*xySize+npix]; |
---|
261 | if(goodSize>0) findNormalStats(image,goodSize,dud,imageSigma[z]); |
---|
262 | else imageSigma[z] = 1.; |
---|
263 | } |
---|
264 | // Next, do source finding. |
---|
265 | long *imdim = new long[2]; |
---|
266 | imdim[0] = dim[0]; imdim[1] = dim[1]; |
---|
267 | Image *channelImage = new Image(imdim); |
---|
268 | channelImage->saveParam(par); |
---|
269 | |
---|
270 | bool *doChannel = new bool[zdim]; |
---|
271 | // purpose of this is to ignore the Milky Way channels -- if we are flagging them... |
---|
272 | for(int z=0;z<zdim;z++) |
---|
273 | doChannel[z] = !( par.getFlagMW() && (z>=par.getMinMW()) && (z<=par.getMaxMW()) ); |
---|
274 | |
---|
275 | for(int z=0; z<zdim; z++){ |
---|
276 | |
---|
277 | if( par.isVerbose() && ((100*(z+1)/zdim)%5 == 0) ){ |
---|
278 | std::cout << "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b|"; |
---|
279 | for(int i=0;i<(100*(z+1)/zdim)/5;i++) std::cout << "#"; |
---|
280 | for(int i=(100*(z+1)/zdim)/5;i<20;i++) std::cout << " "; |
---|
281 | std::cout << "|" << std::flush; |
---|
282 | } |
---|
283 | // if(par.isVerbose() && ((1000*z/zdim)%10==0) ) |
---|
284 | // std::cout << "Done " << setw(2) << 100*z/zdim << "%\b\b\b\b\b\b\b\b" << std::flush; |
---|
285 | |
---|
286 | if( doChannel[z] ){ |
---|
287 | for(int npix=0; npix<xySize; npix++) image[npix] = reconArray[z*xySize + npix]; |
---|
288 | channelImage->saveArray(image,xySize); |
---|
289 | channelImage->setStats(imageMedian[z],imageSigma[z],par.getCut()); |
---|
290 | if(par.getFlagFDR()) channelImage->setupFDR(); |
---|
291 | channelImage->setMinSize(par.getMinPix()); |
---|
292 | channelImage->lutz_detect(); |
---|
293 | for(int obj=0;obj<channelImage->getNumObj();obj++){ |
---|
294 | Detection *object = new Detection; |
---|
295 | *object = channelImage->getObject(obj); |
---|
296 | // if(par.getFlagGrowth()) growObject(*object,*channelImage); |
---|
297 | // Fix up z coordinates of each pixel to match original array (x & y are fine) |
---|
298 | for(int pix=0;pix<object->getSize();pix++){ |
---|
299 | object->setZ(pix, z); |
---|
300 | object->setF(pix, originalArray[object->getX(pix)+object->getY(pix)*dim[0]+z*xySize]); |
---|
301 | // NB: set F to the original value, not the recon value. |
---|
302 | } |
---|
303 | object->addOffsets(par); |
---|
304 | object->calcParams(); |
---|
305 | // outputList.push_back(*object); |
---|
306 | mergeIntoList(*object,outputList,par); |
---|
307 | delete object; |
---|
308 | } |
---|
309 | channelImage->clearDetectionList(); |
---|
310 | } |
---|
311 | |
---|
312 | } |
---|
313 | |
---|
314 | std::cout << "\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\b\bFound " << outputList.size() - num |
---|
315 | << ". " << std::endl << std::flush; |
---|
316 | |
---|
317 | delete [] image; |
---|
318 | delete [] imdim; |
---|
319 | delete channelImage; |
---|
320 | delete [] doChannel; |
---|
321 | delete [] imageMedian; |
---|
322 | delete [] imageSigma; |
---|
323 | |
---|
324 | delete [] isGood; |
---|
325 | |
---|
326 | return outputList; |
---|
327 | } |
---|