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 | |
---|
8 | copyright 2008, University of Tennessee |
---|
9 | """ |
---|
10 | import numpy |
---|
11 | import math |
---|
12 | import scipy.special |
---|
13 | |
---|
14 | def smear_selection(data1D): |
---|
15 | """ |
---|
16 | Creates the right type of smearer according |
---|
17 | to the data |
---|
18 | """ |
---|
19 | pass |
---|
20 | |
---|
21 | class _BaseSmearer(object): |
---|
22 | |
---|
23 | def __init__(self): |
---|
24 | self.nbins = 0 |
---|
25 | self._weights = None |
---|
26 | |
---|
27 | def _compute_matrix(self): return NotImplemented |
---|
28 | |
---|
29 | def __call__(self, iq): |
---|
30 | """ |
---|
31 | Return the smeared I(q) value at the given q. |
---|
32 | The smeared I(q) is computed using a predetermined |
---|
33 | smearing matrix for a particular binning. |
---|
34 | |
---|
35 | @param q: I(q) array |
---|
36 | @return: smeared I(q) |
---|
37 | """ |
---|
38 | # Sanity check |
---|
39 | if len(iq) != self.nbins: |
---|
40 | raise RuntimeError, "Invalid I(q) vector: inconsistent array length %s <> %s" % (len(iq), self.nbins) |
---|
41 | |
---|
42 | if self._weights == None: |
---|
43 | self._compute_matrix() |
---|
44 | |
---|
45 | iq_smeared = numpy.zeros(self.nbins) |
---|
46 | # Loop over q-values |
---|
47 | for q_i in range(self.nbins): |
---|
48 | sum = 0.0 |
---|
49 | counts = 0.0 |
---|
50 | |
---|
51 | for i in range(self.nbins): |
---|
52 | sum += iq[i] * self._weights[q_i][i] |
---|
53 | counts += self._weights[q_i][i] |
---|
54 | |
---|
55 | iq_smeared[q_i] = sum/counts |
---|
56 | |
---|
57 | return iq_smeared |
---|
58 | |
---|
59 | class _SlitSmearer(_BaseSmearer): |
---|
60 | """ |
---|
61 | Slit smearing for I(q) array |
---|
62 | """ |
---|
63 | |
---|
64 | def __init__(self, nbins=None, width=None, height=None, min=None, max=None): |
---|
65 | """ |
---|
66 | Initialization |
---|
67 | |
---|
68 | @param iq: I(q) array [cm-1] |
---|
69 | @param width: slit width [A-1] |
---|
70 | @param height: slit height [A-1] |
---|
71 | @param min: Q_min [A-1] |
---|
72 | @param max: Q_max [A-1] |
---|
73 | """ |
---|
74 | ## Slit width in Q units |
---|
75 | self.width = width |
---|
76 | ## Slit height in Q units |
---|
77 | self.height = height |
---|
78 | ## Q_min (Min Q-value for I(q)) |
---|
79 | self.min = min |
---|
80 | ## Q_max (Max Q_value for I(q)) |
---|
81 | self.max = max |
---|
82 | ## Number of Q bins |
---|
83 | self.nbins = nbins |
---|
84 | ## Number of points used in the smearing computation |
---|
85 | self.npts = 1000 |
---|
86 | ## Smearing matrix |
---|
87 | self._weights = None |
---|
88 | |
---|
89 | def _compute_matrix(self): |
---|
90 | """ |
---|
91 | Compute the smearing matrix for the current I(q) array |
---|
92 | """ |
---|
93 | weights = numpy.zeros([self.nbins, self.nbins]) |
---|
94 | |
---|
95 | # Loop over all q-values |
---|
96 | for i in range(self.nbins): |
---|
97 | q = self.min + i*(self.max-self.min)/float(self.nbins-1.0) |
---|
98 | |
---|
99 | # For each q-value, compute the weight of each other q-bin |
---|
100 | # in the I(q) array |
---|
101 | npts_h = self.npts if self.height>0 else 1 |
---|
102 | npts_w = self.npts if self.width>0 else 1 |
---|
103 | |
---|
104 | # If both height and width are great than zero, |
---|
105 | # modify the number of points in each direction so |
---|
106 | # that the total number of points is still what |
---|
107 | # the user would expect (downgrade resolution) |
---|
108 | if npts_h>1 and npts_w>1: |
---|
109 | npts_h = int(math.ceil(math.sqrt(self.npts))) |
---|
110 | npts_w = npts_h |
---|
111 | |
---|
112 | for k in range(npts_h): |
---|
113 | if npts_h==1: |
---|
114 | shift_h = 0 |
---|
115 | else: |
---|
116 | shift_h = self.height/(float(npts_h-1.0)) * k |
---|
117 | for j in range(npts_w): |
---|
118 | if npts_w==1: |
---|
119 | shift_w = 0 |
---|
120 | else: |
---|
121 | shift_w = self.width/(float(npts_w-1.0)) * j |
---|
122 | q_shifted = math.sqrt( ((q-shift_w)*(q-shift_w) + shift_h*shift_h) ) |
---|
123 | q_i = int(math.floor( (q_shifted-self.min)/((self.max-self.min)/(self.nbins -1.0)) )) |
---|
124 | |
---|
125 | # Skip the entries outside our I(q) range |
---|
126 | #TODO: be careful with edge effect |
---|
127 | if q_i<self.nbins: |
---|
128 | weights[i][q_i] = weights[i][q_i]+1.0 |
---|
129 | |
---|
130 | self._weights = weights |
---|
131 | return self._weights |
---|
132 | |
---|
133 | class SlitSmearer(_SlitSmearer): |
---|
134 | """ |
---|
135 | Adaptor for slit smearing class and SANS data |
---|
136 | """ |
---|
137 | def __init__(self, data1D): |
---|
138 | """ |
---|
139 | Assumption: equally spaced bins of increasing q-values. |
---|
140 | |
---|
141 | @param data1D: data used to set the smearing parameters |
---|
142 | """ |
---|
143 | # Initialization from parent class |
---|
144 | super(SlitSmearer, self).__init__() |
---|
145 | |
---|
146 | ## Slit width |
---|
147 | self.width = 0 |
---|
148 | if data1D.dxw is not None and len(data1D.dxw)==len(data1D.x): |
---|
149 | self.width = data1D.dxw[0] |
---|
150 | # Sanity check |
---|
151 | for value in data1D.dxw: |
---|
152 | if value != self.width: |
---|
153 | raise RuntimeError, "Slit smearing parameters must be the same for all data" |
---|
154 | |
---|
155 | ## Slit height |
---|
156 | self.height = 0 |
---|
157 | if data1D.dxl is not None and len(data1D.dxl)==len(data1D.x): |
---|
158 | self.height = data1D.dxl[0] |
---|
159 | # Sanity check |
---|
160 | for value in data1D.dxl: |
---|
161 | if value != self.height: |
---|
162 | raise RuntimeError, "Slit smearing parameters must be the same for all data" |
---|
163 | |
---|
164 | ## Number of Q bins |
---|
165 | self.nbins = len(data1D.x) |
---|
166 | ## Minimum Q |
---|
167 | self.min = data1D.x[0] |
---|
168 | ## Maximum |
---|
169 | self.max = data1D.x[len(data1D.x)-1] |
---|
170 | |
---|
171 | |
---|
172 | class _QSmearer(_BaseSmearer): |
---|
173 | """ |
---|
174 | Perform Gaussian Q smearing |
---|
175 | """ |
---|
176 | |
---|
177 | def __init__(self, nbins=None, width=None, min=None, max=None): |
---|
178 | """ |
---|
179 | Initialization |
---|
180 | |
---|
181 | @param nbins: number of Q bins |
---|
182 | @param width: standard deviation in Q [A-1] |
---|
183 | @param min: Q_min [A-1] |
---|
184 | @param max: Q_max [A-1] |
---|
185 | """ |
---|
186 | ## Standard deviation in Q [A-1] |
---|
187 | self.width = width |
---|
188 | ## Q_min (Min Q-value for I(q)) |
---|
189 | self.min = min |
---|
190 | ## Q_max (Max Q_value for I(q)) |
---|
191 | self.max = max |
---|
192 | ## Number of Q bins |
---|
193 | self.nbins = nbins |
---|
194 | ## Smearing matrix |
---|
195 | self._weights = None |
---|
196 | |
---|
197 | def _compute_matrix(self): |
---|
198 | """ |
---|
199 | Compute the smearing matrix for the current I(q) array |
---|
200 | """ |
---|
201 | weights = numpy.zeros([self.nbins, self.nbins]) |
---|
202 | |
---|
203 | # Loop over all q-values |
---|
204 | step = (self.max-self.min)/float(self.nbins-1.0) |
---|
205 | for i in range(self.nbins): |
---|
206 | q = self.min + i*step |
---|
207 | q_min = q - 0.5*step |
---|
208 | q_max = q + 0.5*step |
---|
209 | |
---|
210 | for j in range(self.nbins): |
---|
211 | q_j = self.min + j*step |
---|
212 | |
---|
213 | # Compute the fraction of the Gaussian contributing |
---|
214 | # to the q bin between q_min and q_max |
---|
215 | value = scipy.special.erf( (q_max-q_j)/(math.sqrt(2.0)*self.width) ) |
---|
216 | value -=scipy.special.erf( (q_min-q_j)/(math.sqrt(2.0)*self.width) ) |
---|
217 | |
---|
218 | weights[i][j] += value |
---|
219 | |
---|
220 | self._weights = weights |
---|
221 | return self._weights |
---|
222 | |
---|
223 | class QSmearer(_QSmearer): |
---|
224 | """ |
---|
225 | Adaptor for Gaussian Q smearing class and SANS data |
---|
226 | """ |
---|
227 | def __init__(self, data1D): |
---|
228 | """ |
---|
229 | Assumption: equally spaced bins of increasing q-values. |
---|
230 | |
---|
231 | @param data1D: data used to set the smearing parameters |
---|
232 | """ |
---|
233 | # Initialization from parent class |
---|
234 | super(QSmearer, self).__init__() |
---|
235 | |
---|
236 | ## Slit width |
---|
237 | self.width = 0 |
---|
238 | if data1D.dx is not None and len(data1D.dx)==len(data1D.x): |
---|
239 | self.width = data1D.dx[0] |
---|
240 | # Sanity check |
---|
241 | for value in data1D.dx: |
---|
242 | if value != self.width: |
---|
243 | raise RuntimeError, "dQ must be the same for all data" |
---|
244 | |
---|
245 | ## Number of Q bins |
---|
246 | self.nbins = len(data1D.x) |
---|
247 | ## Minimum Q |
---|
248 | self.min = data1D.x[0] |
---|
249 | ## Maximum |
---|
250 | self.max = data1D.x[len(data1D.x)-1] |
---|
251 | |
---|
252 | |
---|
253 | if __name__ == '__main__': |
---|
254 | x = 0.001*numpy.arange(1,11) |
---|
255 | y = 12.0-numpy.arange(1,11) |
---|
256 | print x |
---|
257 | #for i in range(10): print i, 0.001 + i*0.008/9.0 |
---|
258 | #for i in range(100): print i, int(math.floor( (i/ (100/9.0)) )) |
---|
259 | |
---|
260 | |
---|
261 | s = _SlitSmearer(nbins=10, width=0.0, height=0.005, min=0.001, max=0.010) |
---|
262 | #s = _QSmearer(nbins=10, width=0.001, min=0.001, max=0.010) |
---|
263 | s._compute_matrix() |
---|
264 | |
---|
265 | sy = s(y) |
---|
266 | print sy |
---|
267 | |
---|
268 | if True: |
---|
269 | for i in range(10): |
---|
270 | print x[i], sy[i] |
---|
271 | #print q, ' : ', s.weight(q), s._compute_iq(q) |
---|
272 | #print q, ' : ', s(q), s._compute_iq(q) |
---|
273 | #s._compute_iq(q) |
---|
274 | |
---|
275 | |
---|
276 | |
---|
277 | |
---|