[d00f8ff] | 1 | """ |
---|
| 2 | This software was developed by the University of Tennessee as part of the |
---|
| 3 | Distributed Data Analysis of Neutron Scattering Experiments (DANSE) |
---|
| 4 | project funded by the US National Science Foundation. |
---|
| 5 | |
---|
| 6 | See the license text in license.txt |
---|
| 7 | |
---|
[a3f8d58] | 8 | copyright 2009, University of Tennessee |
---|
[d00f8ff] | 9 | """ |
---|
[4fe4394] | 10 | |
---|
[a3f8d58] | 11 | import DataLoader.extensions.smearer as smearer |
---|
[d00f8ff] | 12 | import numpy |
---|
| 13 | import math |
---|
[5859862] | 14 | import logging, sys |
---|
[d00f8ff] | 15 | import scipy.special |
---|
| 16 | |
---|
| 17 | def smear_selection(data1D): |
---|
| 18 | """ |
---|
| 19 | Creates the right type of smearer according |
---|
[4fe4394] | 20 | to the data. |
---|
| 21 | |
---|
| 22 | The canSAS format has a rule that either |
---|
| 23 | slit smearing data OR resolution smearing data |
---|
| 24 | is available. |
---|
| 25 | |
---|
| 26 | For the present purpose, we choose the one that |
---|
| 27 | has none-zero data. If both slit and resolution |
---|
| 28 | smearing arrays are filled with good data |
---|
| 29 | (which should not happen), then we choose the |
---|
| 30 | resolution smearing data. |
---|
| 31 | |
---|
| 32 | @param data1D: Data1D object |
---|
[d00f8ff] | 33 | """ |
---|
[4fe4394] | 34 | # Sanity check. If we are not dealing with a SANS Data1D |
---|
| 35 | # object, just return None |
---|
[21d2eb0] | 36 | if data1D.__class__.__name__ != 'Data1D': |
---|
| 37 | return None |
---|
| 38 | |
---|
| 39 | if not hasattr(data1D, "dx") and not hasattr(data1D, "dxl") and not hasattr(data1D, "dxw"): |
---|
[4fe4394] | 40 | return None |
---|
| 41 | |
---|
| 42 | # Look for resolution smearing data |
---|
| 43 | _found_resolution = False |
---|
| 44 | if data1D.dx is not None and len(data1D.dx)==len(data1D.x): |
---|
| 45 | |
---|
| 46 | # Check that we have non-zero data |
---|
| 47 | if data1D.dx[0]>0.0: |
---|
| 48 | _found_resolution = True |
---|
[c7ac15e] | 49 | #print "_found_resolution",_found_resolution |
---|
| 50 | #print "data1D.dx[0]",data1D.dx[0],data1D.dxl[0] |
---|
[4fe4394] | 51 | # If we found resolution smearing data, return a QSmearer |
---|
| 52 | if _found_resolution == True: |
---|
| 53 | return QSmearer(data1D) |
---|
| 54 | |
---|
| 55 | # Look for slit smearing data |
---|
| 56 | _found_slit = False |
---|
| 57 | if data1D.dxl is not None and len(data1D.dxl)==len(data1D.x) \ |
---|
| 58 | and data1D.dxw is not None and len(data1D.dxw)==len(data1D.x): |
---|
| 59 | |
---|
| 60 | # Check that we have non-zero data |
---|
| 61 | if data1D.dxl[0]>0.0 or data1D.dxw[0]>0.0: |
---|
| 62 | _found_slit = True |
---|
| 63 | |
---|
| 64 | # Sanity check: all data should be the same as a function of Q |
---|
| 65 | for item in data1D.dxl: |
---|
| 66 | if data1D.dxl[0] != item: |
---|
| 67 | _found_resolution = False |
---|
| 68 | break |
---|
| 69 | |
---|
| 70 | for item in data1D.dxw: |
---|
| 71 | if data1D.dxw[0] != item: |
---|
| 72 | _found_resolution = False |
---|
| 73 | break |
---|
| 74 | # If we found slit smearing data, return a slit smearer |
---|
| 75 | if _found_slit == True: |
---|
| 76 | return SlitSmearer(data1D) |
---|
| 77 | |
---|
| 78 | return None |
---|
| 79 | |
---|
[d00f8ff] | 80 | |
---|
| 81 | class _BaseSmearer(object): |
---|
| 82 | |
---|
| 83 | def __init__(self): |
---|
| 84 | self.nbins = 0 |
---|
| 85 | self._weights = None |
---|
[a3f8d58] | 86 | ## Internal flag to keep track of C++ smearer initialization |
---|
| 87 | self._init_complete = False |
---|
| 88 | self._smearer = None |
---|
| 89 | |
---|
| 90 | def __deepcopy__(self, memo={}): |
---|
| 91 | """ |
---|
| 92 | Return a valid copy of self. |
---|
| 93 | Avoid copying the _smearer C object and force a matrix recompute |
---|
| 94 | when the copy is used. |
---|
| 95 | """ |
---|
| 96 | result = _BaseSmearer() |
---|
| 97 | result.nbins = self.nbins |
---|
| 98 | return result |
---|
| 99 | |
---|
[d00f8ff] | 100 | |
---|
| 101 | def _compute_matrix(self): return NotImplemented |
---|
| 102 | |
---|
[5859862] | 103 | def get_bin_range(self, q_min=None, q_max=None): |
---|
| 104 | """ |
---|
| 105 | @param q_min: minimum q-value to smear |
---|
| 106 | @param q_max: maximum q-value to smear |
---|
| 107 | """ |
---|
[65883cf] | 108 | # If this is the first time we call for smearing, |
---|
| 109 | # initialize the C++ smearer object first |
---|
| 110 | if not self._init_complete: |
---|
| 111 | self._initialize_smearer() |
---|
| 112 | |
---|
[5859862] | 113 | if q_min == None: |
---|
| 114 | q_min = self.min |
---|
| 115 | |
---|
| 116 | if q_max == None: |
---|
| 117 | q_max = self.max |
---|
| 118 | |
---|
| 119 | _qmin_unsmeared, _qmax_unsmeared = self.get_unsmeared_range(q_min, q_max) |
---|
| 120 | |
---|
| 121 | _first_bin = None |
---|
| 122 | _last_bin = None |
---|
| 123 | |
---|
| 124 | step = (self.max-self.min)/(self.nbins-1.0) |
---|
[65883cf] | 125 | try: |
---|
| 126 | for i in range(self.nbins): |
---|
| 127 | q_i = smearer.get_q(self._smearer, i) |
---|
| 128 | if (q_i >= _qmin_unsmeared) and (q_i <= _qmax_unsmeared): |
---|
| 129 | # Identify first and last bin |
---|
| 130 | if _first_bin is None: |
---|
| 131 | _first_bin = i |
---|
| 132 | else: |
---|
| 133 | _last_bin = i |
---|
| 134 | except: |
---|
| 135 | raise RuntimeError, "_BaseSmearer.get_bin_range: error getting range\n %s" % sys.exc_value |
---|
[5859862] | 136 | |
---|
| 137 | return _first_bin, _last_bin |
---|
| 138 | |
---|
[a3f8d58] | 139 | def __call__(self, iq_in, first_bin=0, last_bin=None): |
---|
[d00f8ff] | 140 | """ |
---|
[a3f8d58] | 141 | Perform smearing |
---|
[d00f8ff] | 142 | """ |
---|
[a3f8d58] | 143 | # If this is the first time we call for smearing, |
---|
| 144 | # initialize the C++ smearer object first |
---|
| 145 | if not self._init_complete: |
---|
| 146 | self._initialize_smearer() |
---|
| 147 | |
---|
| 148 | # Get the max value for the last bin |
---|
| 149 | if last_bin is None or last_bin>=len(iq_in): |
---|
| 150 | last_bin = len(iq_in)-1 |
---|
| 151 | # Check that the first bin is positive |
---|
| 152 | if first_bin<0: |
---|
| 153 | first_bin = 0 |
---|
[d00f8ff] | 154 | |
---|
[a3f8d58] | 155 | # Sanity check |
---|
| 156 | if len(iq_in) != self.nbins: |
---|
| 157 | raise RuntimeError, "Invalid I(q) vector: inconsistent array length %d != %s" % (len(iq_in), str(self.nbins)) |
---|
| 158 | |
---|
| 159 | # Storage for smeared I(q) |
---|
| 160 | iq_out = numpy.zeros(self.nbins) |
---|
[65883cf] | 161 | smear_output = smearer.smear(self._smearer, iq_in, iq_out, first_bin, last_bin) |
---|
| 162 | if smear_output < 0: |
---|
| 163 | raise RuntimeError, "_BaseSmearer: could not smear, code = %g" % smear_output |
---|
[a3f8d58] | 164 | return iq_out |
---|
[d00f8ff] | 165 | |
---|
| 166 | class _SlitSmearer(_BaseSmearer): |
---|
| 167 | """ |
---|
| 168 | Slit smearing for I(q) array |
---|
| 169 | """ |
---|
| 170 | |
---|
| 171 | def __init__(self, nbins=None, width=None, height=None, min=None, max=None): |
---|
| 172 | """ |
---|
| 173 | Initialization |
---|
| 174 | |
---|
| 175 | @param iq: I(q) array [cm-1] |
---|
| 176 | @param width: slit width [A-1] |
---|
| 177 | @param height: slit height [A-1] |
---|
| 178 | @param min: Q_min [A-1] |
---|
| 179 | @param max: Q_max [A-1] |
---|
| 180 | """ |
---|
[a3f8d58] | 181 | _BaseSmearer.__init__(self) |
---|
[d00f8ff] | 182 | ## Slit width in Q units |
---|
| 183 | self.width = width |
---|
| 184 | ## Slit height in Q units |
---|
| 185 | self.height = height |
---|
| 186 | ## Q_min (Min Q-value for I(q)) |
---|
| 187 | self.min = min |
---|
| 188 | ## Q_max (Max Q_value for I(q)) |
---|
| 189 | self.max = max |
---|
| 190 | ## Number of Q bins |
---|
| 191 | self.nbins = nbins |
---|
| 192 | ## Number of points used in the smearing computation |
---|
[c7ac15e] | 193 | self.npts = 10000 |
---|
[d00f8ff] | 194 | ## Smearing matrix |
---|
| 195 | self._weights = None |
---|
[65883cf] | 196 | self.qvalues = None |
---|
[d00f8ff] | 197 | |
---|
[a3f8d58] | 198 | def _initialize_smearer(self): |
---|
[d00f8ff] | 199 | """ |
---|
[a3f8d58] | 200 | Initialize the C++ smearer object. |
---|
| 201 | This method HAS to be called before smearing |
---|
[d00f8ff] | 202 | """ |
---|
[5859862] | 203 | #self._smearer = smearer.new_slit_smearer(self.width, self.height, self.min, self.max, self.nbins) |
---|
| 204 | self._smearer = smearer.new_slit_smearer_with_q(self.width, self.height, self.qvalues) |
---|
[a3f8d58] | 205 | self._init_complete = True |
---|
[fe2ade9] | 206 | |
---|
[5859862] | 207 | def get_unsmeared_range(self, q_min, q_max): |
---|
| 208 | """ |
---|
| 209 | Determine the range needed in unsmeared-Q to cover |
---|
| 210 | the smeared Q range |
---|
| 211 | """ |
---|
| 212 | # Range used for input to smearing |
---|
| 213 | _qmin_unsmeared = q_min |
---|
| 214 | _qmax_unsmeared = q_max |
---|
| 215 | try: |
---|
| 216 | _qmin_unsmeared = self.min |
---|
| 217 | _qmax_unsmeared = self.max |
---|
| 218 | except: |
---|
| 219 | logging.error("_SlitSmearer.get_bin_range: %s" % sys.exc_value) |
---|
| 220 | return _qmin_unsmeared, _qmax_unsmeared |
---|
[d00f8ff] | 221 | |
---|
| 222 | class SlitSmearer(_SlitSmearer): |
---|
| 223 | """ |
---|
| 224 | Adaptor for slit smearing class and SANS data |
---|
| 225 | """ |
---|
| 226 | def __init__(self, data1D): |
---|
| 227 | """ |
---|
| 228 | Assumption: equally spaced bins of increasing q-values. |
---|
| 229 | |
---|
| 230 | @param data1D: data used to set the smearing parameters |
---|
| 231 | """ |
---|
| 232 | # Initialization from parent class |
---|
| 233 | super(SlitSmearer, self).__init__() |
---|
| 234 | |
---|
| 235 | ## Slit width |
---|
| 236 | self.width = 0 |
---|
| 237 | if data1D.dxw is not None and len(data1D.dxw)==len(data1D.x): |
---|
| 238 | self.width = data1D.dxw[0] |
---|
| 239 | # Sanity check |
---|
| 240 | for value in data1D.dxw: |
---|
| 241 | if value != self.width: |
---|
| 242 | raise RuntimeError, "Slit smearing parameters must be the same for all data" |
---|
| 243 | |
---|
| 244 | ## Slit height |
---|
| 245 | self.height = 0 |
---|
| 246 | if data1D.dxl is not None and len(data1D.dxl)==len(data1D.x): |
---|
| 247 | self.height = data1D.dxl[0] |
---|
| 248 | # Sanity check |
---|
| 249 | for value in data1D.dxl: |
---|
| 250 | if value != self.height: |
---|
| 251 | raise RuntimeError, "Slit smearing parameters must be the same for all data" |
---|
| 252 | |
---|
| 253 | ## Number of Q bins |
---|
| 254 | self.nbins = len(data1D.x) |
---|
| 255 | ## Minimum Q |
---|
[5859862] | 256 | self.min = min(data1D.x) |
---|
[d00f8ff] | 257 | ## Maximum |
---|
[5859862] | 258 | self.max = max(data1D.x) |
---|
| 259 | ## Q-values |
---|
| 260 | self.qvalues = data1D.x |
---|
| 261 | |
---|
[d00f8ff] | 262 | |
---|
| 263 | class _QSmearer(_BaseSmearer): |
---|
| 264 | """ |
---|
| 265 | Perform Gaussian Q smearing |
---|
| 266 | """ |
---|
| 267 | |
---|
| 268 | def __init__(self, nbins=None, width=None, min=None, max=None): |
---|
| 269 | """ |
---|
| 270 | Initialization |
---|
| 271 | |
---|
| 272 | @param nbins: number of Q bins |
---|
[c0d9981] | 273 | @param width: array standard deviation in Q [A-1] |
---|
[d00f8ff] | 274 | @param min: Q_min [A-1] |
---|
| 275 | @param max: Q_max [A-1] |
---|
| 276 | """ |
---|
[a3f8d58] | 277 | _BaseSmearer.__init__(self) |
---|
[d00f8ff] | 278 | ## Standard deviation in Q [A-1] |
---|
| 279 | self.width = width |
---|
| 280 | ## Q_min (Min Q-value for I(q)) |
---|
| 281 | self.min = min |
---|
| 282 | ## Q_max (Max Q_value for I(q)) |
---|
| 283 | self.max = max |
---|
| 284 | ## Number of Q bins |
---|
| 285 | self.nbins = nbins |
---|
| 286 | ## Smearing matrix |
---|
| 287 | self._weights = None |
---|
[65883cf] | 288 | self.qvalues = None |
---|
[d00f8ff] | 289 | |
---|
[a3f8d58] | 290 | def _initialize_smearer(self): |
---|
[d00f8ff] | 291 | """ |
---|
[a3f8d58] | 292 | Initialize the C++ smearer object. |
---|
| 293 | This method HAS to be called before smearing |
---|
[d00f8ff] | 294 | """ |
---|
[5859862] | 295 | #self._smearer = smearer.new_q_smearer(numpy.asarray(self.width), self.min, self.max, self.nbins) |
---|
| 296 | self._smearer = smearer.new_q_smearer_with_q(numpy.asarray(self.width), self.qvalues) |
---|
[a3f8d58] | 297 | self._init_complete = True |
---|
[d00f8ff] | 298 | |
---|
[5859862] | 299 | def get_unsmeared_range(self, q_min, q_max): |
---|
| 300 | """ |
---|
| 301 | Determine the range needed in unsmeared-Q to cover |
---|
| 302 | the smeared Q range |
---|
| 303 | Take 3 sigmas as the offset between smeared and unsmeared space |
---|
| 304 | """ |
---|
| 305 | # Range used for input to smearing |
---|
| 306 | _qmin_unsmeared = q_min |
---|
| 307 | _qmax_unsmeared = q_max |
---|
| 308 | try: |
---|
| 309 | offset = 3.0*max(self.width) |
---|
| 310 | _qmin_unsmeared = max([self.min, q_min-offset]) |
---|
| 311 | _qmax_unsmeared = min([self.max, q_max+offset]) |
---|
| 312 | except: |
---|
| 313 | logging.error("_QSmearer.get_bin_range: %s" % sys.exc_value) |
---|
| 314 | return _qmin_unsmeared, _qmax_unsmeared |
---|
| 315 | |
---|
| 316 | |
---|
| 317 | |
---|
[d00f8ff] | 318 | class QSmearer(_QSmearer): |
---|
| 319 | """ |
---|
| 320 | Adaptor for Gaussian Q smearing class and SANS data |
---|
| 321 | """ |
---|
| 322 | def __init__(self, data1D): |
---|
| 323 | """ |
---|
| 324 | Assumption: equally spaced bins of increasing q-values. |
---|
| 325 | |
---|
| 326 | @param data1D: data used to set the smearing parameters |
---|
| 327 | """ |
---|
| 328 | # Initialization from parent class |
---|
| 329 | super(QSmearer, self).__init__() |
---|
| 330 | |
---|
[c0d9981] | 331 | ## Resolution |
---|
[4fe4394] | 332 | self.width = numpy.zeros(len(data1D.x)) |
---|
[d00f8ff] | 333 | if data1D.dx is not None and len(data1D.dx)==len(data1D.x): |
---|
[4fe4394] | 334 | self.width = data1D.dx |
---|
[d00f8ff] | 335 | |
---|
| 336 | ## Number of Q bins |
---|
| 337 | self.nbins = len(data1D.x) |
---|
| 338 | ## Minimum Q |
---|
[5859862] | 339 | self.min = min(data1D.x) |
---|
[d00f8ff] | 340 | ## Maximum |
---|
[5859862] | 341 | self.max = max(data1D.x) |
---|
| 342 | ## Q-values |
---|
| 343 | self.qvalues = data1D.x |
---|
[d00f8ff] | 344 | |
---|
| 345 | |
---|
| 346 | if __name__ == '__main__': |
---|
| 347 | x = 0.001*numpy.arange(1,11) |
---|
| 348 | y = 12.0-numpy.arange(1,11) |
---|
| 349 | print x |
---|
| 350 | #for i in range(10): print i, 0.001 + i*0.008/9.0 |
---|
| 351 | #for i in range(100): print i, int(math.floor( (i/ (100/9.0)) )) |
---|
| 352 | |
---|
| 353 | |
---|
| 354 | s = _SlitSmearer(nbins=10, width=0.0, height=0.005, min=0.001, max=0.010) |
---|
| 355 | #s = _QSmearer(nbins=10, width=0.001, min=0.001, max=0.010) |
---|
| 356 | s._compute_matrix() |
---|
| 357 | |
---|
| 358 | sy = s(y) |
---|
| 359 | print sy |
---|
| 360 | |
---|
| 361 | if True: |
---|
| 362 | for i in range(10): |
---|
[fe2ade9] | 363 | print x[i],y[i], sy[i] |
---|
[d00f8ff] | 364 | #print q, ' : ', s.weight(q), s._compute_iq(q) |
---|
| 365 | #print q, ' : ', s(q), s._compute_iq(q) |
---|
| 366 | #s._compute_iq(q) |
---|
| 367 | |
---|
| 368 | |
---|
| 369 | |
---|
| 370 | |
---|