1 | #TODO: Use simview to generate P(r) and I(q) pairs in sansview. |
---|
2 | # Make sure the option of saving each curve is available |
---|
3 | # Use the I(q) curve as input and compare the output to P(r) |
---|
4 | |
---|
5 | import os |
---|
6 | import wx |
---|
7 | from sans.guitools.plottables import Data1D, Theory1D |
---|
8 | from sans.guicomm.events import NewPlotEvent, StatusEvent |
---|
9 | import math, numpy |
---|
10 | from sans.pr.invertor import Invertor |
---|
11 | |
---|
12 | class Plugin: |
---|
13 | |
---|
14 | def __init__(self): |
---|
15 | ## Plug-in name |
---|
16 | self.sub_menu = "Pr inversion" |
---|
17 | |
---|
18 | ## Reference to the parent window |
---|
19 | self.parent = None |
---|
20 | |
---|
21 | ## Simulation window manager |
---|
22 | self.simview = None |
---|
23 | |
---|
24 | ## List of panels for the simulation perspective (names) |
---|
25 | self.perspective = [] |
---|
26 | |
---|
27 | ## State data |
---|
28 | self.alpha = 0.0001 |
---|
29 | self.nfunc = 10 |
---|
30 | self.max_length = 140.0 |
---|
31 | ## Remember last plottable processed |
---|
32 | self.last_data = "sphere_60_q0_2.txt" |
---|
33 | ## Time elapsed for last computation [sec] |
---|
34 | # Start with a good default |
---|
35 | self.elapsed = 0.022 |
---|
36 | |
---|
37 | ## Current invertor |
---|
38 | self.invertor = None |
---|
39 | ## Calculation thread |
---|
40 | self.calc_thread = None |
---|
41 | ## Estimation thread |
---|
42 | self.estimation_thread = None |
---|
43 | ## Result panel |
---|
44 | self.control_panel = None |
---|
45 | ## Currently views plottable |
---|
46 | self.current_plottable = None |
---|
47 | |
---|
48 | def populate_menu(self, id, owner): |
---|
49 | """ |
---|
50 | Create a menu for the plug-in |
---|
51 | """ |
---|
52 | import wx |
---|
53 | shapes = wx.Menu() |
---|
54 | |
---|
55 | id = wx.NewId() |
---|
56 | shapes.Append(id, '&Sphere test') |
---|
57 | wx.EVT_MENU(owner, id, self._fit_pr) |
---|
58 | return [(id, shapes, "P(r)")] |
---|
59 | |
---|
60 | def _fit_pr(self, evt): |
---|
61 | from sans.pr.invertor import Invertor |
---|
62 | import numpy |
---|
63 | import pylab |
---|
64 | import math |
---|
65 | from sans.guicomm.events import NewPlotEvent |
---|
66 | from sans.guitools.plottables import Data1D, Theory1D |
---|
67 | |
---|
68 | # Generate P(r) for sphere |
---|
69 | radius = 60.0 |
---|
70 | d_max = 2*radius |
---|
71 | |
---|
72 | |
---|
73 | r = pylab.arange(0.01, d_max, d_max/51.0) |
---|
74 | M = len(r) |
---|
75 | y = numpy.zeros(M) |
---|
76 | pr_err = numpy.zeros(M) |
---|
77 | |
---|
78 | sum = 0.0 |
---|
79 | for j in range(M): |
---|
80 | value = self.pr_theory(r[j], radius) |
---|
81 | sum += value |
---|
82 | y[j] = value |
---|
83 | pr_err[j] = math.sqrt(y[j]) |
---|
84 | |
---|
85 | |
---|
86 | y = y/sum*d_max/len(r) |
---|
87 | |
---|
88 | |
---|
89 | |
---|
90 | # Perform fit |
---|
91 | pr = Invertor() |
---|
92 | pr.d_max = d_max |
---|
93 | pr.alpha = 0 |
---|
94 | pr.x = r |
---|
95 | pr.y = y |
---|
96 | pr.err = pr_err |
---|
97 | out, cov = pr.pr_fit() |
---|
98 | for i in range(len(out)): |
---|
99 | print "%g +- %g" % (out[i], math.sqrt(cov[i][i])) |
---|
100 | |
---|
101 | |
---|
102 | # Show input P(r) |
---|
103 | new_plot = Data1D(pr.x, pr.y, pr.err) |
---|
104 | new_plot.name = "P_{obs}(r)" |
---|
105 | new_plot.xaxis("\\rm{r}", 'A') |
---|
106 | new_plot.yaxis("\\rm{P(r)} ","cm^{-3}") |
---|
107 | wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title="Pr")) |
---|
108 | |
---|
109 | # Show P(r) fit |
---|
110 | self.show_pr(out, pr) |
---|
111 | |
---|
112 | # Show I(q) fit |
---|
113 | q = pylab.arange(0.001, 0.1, 0.01/51.0) |
---|
114 | self.show_iq(out, pr, q) |
---|
115 | |
---|
116 | |
---|
117 | def show_shpere(self, x, radius=70.0, x_range=70.0): |
---|
118 | import numpy |
---|
119 | import pylab |
---|
120 | import math |
---|
121 | from sans.guicomm.events import NewPlotEvent |
---|
122 | from sans.guitools.plottables import Data1D, Theory1D |
---|
123 | # Show P(r) |
---|
124 | y_true = numpy.zeros(len(x)) |
---|
125 | |
---|
126 | sum_true = 0.0 |
---|
127 | for i in range(len(x)): |
---|
128 | y_true[i] = self.pr_theory(x[i], radius) |
---|
129 | sum_true += y_true[i] |
---|
130 | |
---|
131 | y_true = y_true/sum_true*x_range/len(x) |
---|
132 | |
---|
133 | # Show the theory P(r) |
---|
134 | new_plot = Theory1D(x, y_true) |
---|
135 | new_plot.name = "P_{true}(r)" |
---|
136 | new_plot.xaxis("\\rm{r}", 'A') |
---|
137 | new_plot.yaxis("\\rm{P(r)} ","cm^{-3}") |
---|
138 | |
---|
139 | |
---|
140 | #Put this call in plottables/guitools |
---|
141 | wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title="Sphere P(r)")) |
---|
142 | |
---|
143 | def show_iq(self, out, pr, q=None): |
---|
144 | import numpy |
---|
145 | import pylab |
---|
146 | import math |
---|
147 | from sans.guicomm.events import NewPlotEvent |
---|
148 | from sans.guitools.plottables import Data1D, Theory1D |
---|
149 | |
---|
150 | qtemp = pr.x |
---|
151 | if not q==None: |
---|
152 | qtemp = q |
---|
153 | |
---|
154 | # Make a plot |
---|
155 | maxq = -1 |
---|
156 | for q_i in qtemp: |
---|
157 | if q_i>maxq: |
---|
158 | maxq=q_i |
---|
159 | |
---|
160 | x = pylab.arange(0.001, maxq, maxq/301.0) |
---|
161 | y = numpy.zeros(len(x)) |
---|
162 | err = numpy.zeros(len(x)) |
---|
163 | for i in range(len(x)): |
---|
164 | value = pr.iq(out, x[i]) |
---|
165 | y[i] = value |
---|
166 | try: |
---|
167 | err[i] = math.sqrt(math.fabs(value)) |
---|
168 | except: |
---|
169 | err[i] = 1.0 |
---|
170 | print "Error getting error", value, x[i] |
---|
171 | |
---|
172 | new_plot = Theory1D(x, y) |
---|
173 | new_plot.name = "I_{fit}(q)" |
---|
174 | new_plot.xaxis("\\rm{Q}", 'A^{-1}') |
---|
175 | new_plot.yaxis("\\rm{Intensity} ","cm^{-1}") |
---|
176 | #new_plot.group_id = "test group" |
---|
177 | wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title="Iq")) |
---|
178 | |
---|
179 | |
---|
180 | |
---|
181 | |
---|
182 | def show_pr(self, out, pr, cov=None): |
---|
183 | import numpy |
---|
184 | import pylab |
---|
185 | import math |
---|
186 | from sans.guicomm.events import NewPlotEvent |
---|
187 | from sans.guitools.plottables import Data1D, Theory1D |
---|
188 | |
---|
189 | # Show P(r) |
---|
190 | x = pylab.arange(0.0, pr.d_max, pr.d_max/51.0) |
---|
191 | |
---|
192 | y = numpy.zeros(len(x)) |
---|
193 | dy = numpy.zeros(len(x)) |
---|
194 | y_true = numpy.zeros(len(x)) |
---|
195 | |
---|
196 | sum = 0.0 |
---|
197 | for i in range(len(x)): |
---|
198 | if cov==None: |
---|
199 | value = pr.pr(out, x[i]) |
---|
200 | else: |
---|
201 | (value, dy[i]) = pr.pr_err(out, cov, x[i]) |
---|
202 | sum += value |
---|
203 | y[i] = value |
---|
204 | |
---|
205 | y = y/sum*pr.d_max/len(x) |
---|
206 | dy = dy/sum*pr.d_max/len(x) |
---|
207 | |
---|
208 | if cov==None: |
---|
209 | new_plot = Theory1D(x, y) |
---|
210 | else: |
---|
211 | new_plot = Data1D(x, y, dy=dy) |
---|
212 | new_plot.name = "P_{fit}(r)" |
---|
213 | new_plot.xaxis("\\rm{r}", 'A') |
---|
214 | new_plot.yaxis("\\rm{P(r)} ","cm^{-3}") |
---|
215 | |
---|
216 | wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title="P(r) fit")) |
---|
217 | |
---|
218 | return x, pr.d_max |
---|
219 | |
---|
220 | |
---|
221 | def choose_file(self): |
---|
222 | """ |
---|
223 | |
---|
224 | """ |
---|
225 | #TODO: this should be in a common module |
---|
226 | return self.parent.choose_file() |
---|
227 | |
---|
228 | def load(self, path = "sphere_test_data.txt"): |
---|
229 | import numpy, math, sys |
---|
230 | # Read the data from the data file |
---|
231 | data_x = numpy.zeros(0) |
---|
232 | data_y = numpy.zeros(0) |
---|
233 | data_err = numpy.zeros(0) |
---|
234 | if not path == None: |
---|
235 | input_f = open(path,'r') |
---|
236 | buff = input_f.read() |
---|
237 | lines = buff.split('\n') |
---|
238 | for line in lines: |
---|
239 | try: |
---|
240 | toks = line.split() |
---|
241 | x = float(toks[0]) |
---|
242 | y = float(toks[1]) |
---|
243 | data_x = numpy.append(data_x, x) |
---|
244 | data_y = numpy.append(data_y, y) |
---|
245 | try: |
---|
246 | scale = 0.05/math.sqrt(data_x[0]) |
---|
247 | except: |
---|
248 | scale = 1.0 |
---|
249 | #data_err = numpy.append(data_err, 10.0*math.sqrt(y)+1000.0) |
---|
250 | data_err = numpy.append(data_err, scale*math.sqrt(y)) |
---|
251 | except: |
---|
252 | print "Error reading line: ", line |
---|
253 | print sys.exc_value |
---|
254 | |
---|
255 | print "Lines read:", len(data_x) |
---|
256 | return data_x, data_y, data_err |
---|
257 | |
---|
258 | def pr_theory(self, r, R): |
---|
259 | """ |
---|
260 | |
---|
261 | """ |
---|
262 | if r<=2*R: |
---|
263 | return 12.0* ((0.5*r/R)**2) * ((1.0-0.5*r/R)**2) * ( 2.0 + 0.5*r/R ) |
---|
264 | else: |
---|
265 | return 0.0 |
---|
266 | |
---|
267 | def get_context_menu(self, plot_id=None): |
---|
268 | """ |
---|
269 | Get the context menu items available for P(r) |
---|
270 | @param plot_id: Unique ID of a plot, so that we can recognize those |
---|
271 | that we created |
---|
272 | @return: a list of menu items with call-back function |
---|
273 | """ |
---|
274 | return [["Compute P(r)", "Compute P(r) from distribution", self._on_context_inversion]] |
---|
275 | |
---|
276 | |
---|
277 | def start_thread(self): |
---|
278 | from pr_thread import CalcPr |
---|
279 | from copy import deepcopy |
---|
280 | |
---|
281 | # If a thread is already started, stop it |
---|
282 | if self.calc_thread != None and self.calc_thread.isrunning(): |
---|
283 | self.calc_thread.stop() |
---|
284 | |
---|
285 | pr = self.pr.clone() |
---|
286 | self.calc_thread = CalcPr(pr, self.nfunc, error_func=self._thread_error, completefn=self._completed, updatefn=None) |
---|
287 | self.calc_thread.queue() |
---|
288 | self.calc_thread.ready(2.5) |
---|
289 | |
---|
290 | def _thread_error(self, error): |
---|
291 | wx.PostEvent(self.parent, StatusEvent(status=error)) |
---|
292 | |
---|
293 | def _estimate_completed(self, alpha, elapsed): |
---|
294 | """ |
---|
295 | Parameter estimation completed, |
---|
296 | display the results to the user |
---|
297 | @param alpha: estimated best alpha |
---|
298 | @param elapsed: computation time |
---|
299 | """ |
---|
300 | # Save useful info |
---|
301 | self.elapsed = elapsed |
---|
302 | self.control_panel.alpha_estimate = alpha |
---|
303 | |
---|
304 | def _completed(self, out, cov, pr, elapsed): |
---|
305 | """ |
---|
306 | Method called with the results when the inversion |
---|
307 | is done |
---|
308 | |
---|
309 | @param out: output coefficient for the base functions |
---|
310 | @param cov: covariance matrix |
---|
311 | @param pr: Invertor instance |
---|
312 | @param elapsed: time spent computing |
---|
313 | """ |
---|
314 | # Save useful info |
---|
315 | self.elapsed = elapsed |
---|
316 | message = "Computation completed in %g seconds [chi2=%g]" % (elapsed, pr.chi2) |
---|
317 | wx.PostEvent(self.parent, StatusEvent(status=message)) |
---|
318 | |
---|
319 | # Show result on control panel |
---|
320 | self.control_panel.chi2 = pr.chi2 |
---|
321 | self.control_panel.elapsed = elapsed |
---|
322 | self.control_panel.oscillation = pr.oscillations(out) |
---|
323 | #print "OSCILL", pr.oscillations(out) |
---|
324 | |
---|
325 | for i in range(len(out)): |
---|
326 | try: |
---|
327 | print "%d: %g +- %g" % (i, out[i], math.sqrt(math.fabs(cov[i][i]))) |
---|
328 | except: |
---|
329 | print "%d: %g +- ?" % (i, out[i]) |
---|
330 | |
---|
331 | # Make a plot of I(q) data |
---|
332 | new_plot = Data1D(self.pr.x, self.pr.y, self.pr.err) |
---|
333 | new_plot.name = "I_{obs}(q)" |
---|
334 | new_plot.xaxis("\\rm{Q}", 'A^{-1}') |
---|
335 | new_plot.yaxis("\\rm{Intensity} ","cm^{-1}") |
---|
336 | #new_plot.group_id = "test group" |
---|
337 | wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title="Iq")) |
---|
338 | |
---|
339 | # Show I(q) fit |
---|
340 | self.show_iq(out, self.pr) |
---|
341 | |
---|
342 | # Show P(r) fit |
---|
343 | x_values, x_range = self.show_pr(out, self.pr) |
---|
344 | |
---|
345 | # Popup result panel |
---|
346 | #result_panel = InversionResults(self.parent, -1, style=wx.RAISED_BORDER) |
---|
347 | |
---|
348 | def setup_plot_inversion(self, alpha, nfunc, d_max): |
---|
349 | self.alpha = alpha |
---|
350 | self.nfunc = nfunc |
---|
351 | self.max_length = d_max |
---|
352 | |
---|
353 | self._create_plot_pr() |
---|
354 | self.perform_inversion() |
---|
355 | |
---|
356 | def estimate_plot_inversion(self, alpha, nfunc, d_max): |
---|
357 | self.alpha = alpha |
---|
358 | self.nfunc = nfunc |
---|
359 | self.max_length = d_max |
---|
360 | |
---|
361 | self._create_plot_pr() |
---|
362 | self.perform_estimate() |
---|
363 | |
---|
364 | def _create_plot_pr(self): |
---|
365 | """ |
---|
366 | Create and prepare invertor instance from |
---|
367 | a plottable data set. |
---|
368 | @param path: path of the file to read in |
---|
369 | """ |
---|
370 | # Get the data from the chosen data set and perform inversion |
---|
371 | pr = Invertor() |
---|
372 | pr.d_max = self.max_length |
---|
373 | pr.alpha = self.alpha |
---|
374 | pr.x = self.current_plottable.x |
---|
375 | pr.y = self.current_plottable.y |
---|
376 | |
---|
377 | # Fill in errors if none were provided |
---|
378 | if self.current_plottable.dy == None: |
---|
379 | print "no error", self.current_plottable.name |
---|
380 | y = numpy.zeros(len(pr.y)) |
---|
381 | for i in range(len(pr.y)): |
---|
382 | y[i] = math.sqrt(pr.y[i]) |
---|
383 | pr.err = y |
---|
384 | else: |
---|
385 | pr.err = self.current_plottable.dy |
---|
386 | |
---|
387 | self.pr = pr |
---|
388 | |
---|
389 | |
---|
390 | def setup_file_inversion(self, alpha, nfunc, d_max, path): |
---|
391 | self.alpha = alpha |
---|
392 | self.nfunc = nfunc |
---|
393 | self.max_length = d_max |
---|
394 | |
---|
395 | self._create_file_pr(path) |
---|
396 | |
---|
397 | self.perform_inversion() |
---|
398 | |
---|
399 | def estimate_file_inversion(self, alpha, nfunc, d_max, path): |
---|
400 | self.alpha = alpha |
---|
401 | self.nfunc = nfunc |
---|
402 | self.max_length = d_max |
---|
403 | |
---|
404 | if self._create_file_pr(path): |
---|
405 | self.perform_estimate() |
---|
406 | |
---|
407 | def _create_file_pr(self, path): |
---|
408 | """ |
---|
409 | Create and prepare invertor instance from |
---|
410 | a file data set. |
---|
411 | @param path: path of the file to read in |
---|
412 | """ |
---|
413 | # Load data |
---|
414 | if os.path.isfile(path): |
---|
415 | x, y, err = self.load(path) |
---|
416 | |
---|
417 | # Get the data from the chosen data set and perform inversion |
---|
418 | pr = Invertor() |
---|
419 | pr.d_max = self.max_length |
---|
420 | pr.alpha = self.alpha |
---|
421 | pr.x = x |
---|
422 | pr.y = y |
---|
423 | pr.err = err |
---|
424 | |
---|
425 | self.pr = pr |
---|
426 | return True |
---|
427 | return False |
---|
428 | |
---|
429 | def perform_estimate(self): |
---|
430 | print "ESTIMATE" |
---|
431 | from pr_thread import EstimatePr |
---|
432 | from copy import deepcopy |
---|
433 | |
---|
434 | # If a thread is already started, stop it |
---|
435 | if self.estimation_thread != None and self.estimation_thread.isrunning(): |
---|
436 | self.estimation_thread.stop() |
---|
437 | |
---|
438 | pr = self.pr.clone() |
---|
439 | self.estimation_thread = EstimatePr(pr, self.nfunc, error_func=self._thread_error, |
---|
440 | completefn = self._estimate_completed, |
---|
441 | updatefn = None) |
---|
442 | self.estimation_thread.queue() |
---|
443 | self.estimation_thread.ready(2.5) |
---|
444 | |
---|
445 | def perform_inversion(self): |
---|
446 | |
---|
447 | # Time estimate |
---|
448 | #estimated = self.elapsed*self.nfunc**2 |
---|
449 | message = "Computation time may take up to %g seconds" % self.elapsed |
---|
450 | wx.PostEvent(self.parent, StatusEvent(status=message)) |
---|
451 | |
---|
452 | # Start inversion thread |
---|
453 | self.start_thread() |
---|
454 | return |
---|
455 | |
---|
456 | out, cov = self.pr.lstsq(self.nfunc) |
---|
457 | |
---|
458 | # Save useful info |
---|
459 | self.elapsed = self.pr.elapsed |
---|
460 | |
---|
461 | for i in range(len(out)): |
---|
462 | try: |
---|
463 | print "%d: %g +- %g" % (i, out[i], math.sqrt(math.fabs(cov[i][i]))) |
---|
464 | except: |
---|
465 | print "%d: %g +- ?" % (i, out[i]) |
---|
466 | |
---|
467 | |
---|
468 | |
---|
469 | # Make a plot of I(q) data |
---|
470 | new_plot = Data1D(self.pr.x, self.pr.y, dy=self.pr.err) |
---|
471 | new_plot.name = "I_{obs}(q)" |
---|
472 | new_plot.xaxis("\\rm{Q}", 'A^{-1}') |
---|
473 | new_plot.yaxis("\\rm{Intensity} ","cm^{-1}") |
---|
474 | wx.PostEvent(self.parent, NewPlotEvent(plot=new_plot, title="Iq")) |
---|
475 | |
---|
476 | # Show I(q) fit |
---|
477 | self.show_iq(out, self.pr) |
---|
478 | |
---|
479 | # Show P(r) fit |
---|
480 | x_values, x_range = self.show_pr(out, self.pr, cov=cov) |
---|
481 | |
---|
482 | |
---|
483 | |
---|
484 | |
---|
485 | def _on_context_inversion(self, event): |
---|
486 | panel = event.GetEventObject() |
---|
487 | |
---|
488 | from inversion_panel import InversionDlg |
---|
489 | |
---|
490 | # If we have more than one displayed plot, make the user choose |
---|
491 | if len(panel.plots)>1: |
---|
492 | dialog = InversionDlg(None, -1, "P(r) Inversion", panel.plots, pars=False) |
---|
493 | dialog.set_content(self.last_data, self.nfunc, self.alpha, self.max_length) |
---|
494 | if dialog.ShowModal() == wx.ID_OK: |
---|
495 | dataset = dialog.get_content() |
---|
496 | dialog.Destroy() |
---|
497 | else: |
---|
498 | dialog.Destroy() |
---|
499 | return |
---|
500 | elif len(panel.plots)==1: |
---|
501 | dataset = panel.plots.keys()[0] |
---|
502 | else: |
---|
503 | print "Error: No data is available" |
---|
504 | return |
---|
505 | |
---|
506 | # Store a reference to the current plottable |
---|
507 | self.current_plottable = panel.plots[dataset] |
---|
508 | self.control_panel.plotname = dataset |
---|
509 | self.control_panel.nfunc = self.nfunc |
---|
510 | self.control_panel.d_max = self.max_length |
---|
511 | self.control_panel.alpha = self.alpha |
---|
512 | self.parent.set_perspective(self.perspective) |
---|
513 | |
---|
514 | def get_panels(self, parent): |
---|
515 | """ |
---|
516 | Create and return a list of panel objects |
---|
517 | """ |
---|
518 | from inversion_panel import InversionControl |
---|
519 | |
---|
520 | self.parent = parent |
---|
521 | self.control_panel = InversionControl(self.parent, -1, style=wx.RAISED_BORDER) |
---|
522 | self.control_panel.set_manager(self) |
---|
523 | self.control_panel.nfunc = self.nfunc |
---|
524 | self.control_panel.d_max = self.max_length |
---|
525 | self.control_panel.alpha = self.alpha |
---|
526 | |
---|
527 | self.perspective = [] |
---|
528 | self.perspective.append(self.control_panel.window_name) |
---|
529 | return [self.control_panel] |
---|
530 | |
---|
531 | def get_perspective(self): |
---|
532 | """ |
---|
533 | Get the list of panel names for this perspective |
---|
534 | """ |
---|
535 | return self.perspective |
---|
536 | |
---|
537 | def on_perspective(self, event): |
---|
538 | """ |
---|
539 | Call back function for the perspective menu item. |
---|
540 | We notify the parent window that the perspective |
---|
541 | has changed. |
---|
542 | """ |
---|
543 | self.parent.set_perspective(self.perspective) |
---|
544 | |
---|
545 | def post_init(self): |
---|
546 | """ |
---|
547 | Post initialization call back to close the loose ends |
---|
548 | [Somehow openGL needs this call] |
---|
549 | """ |
---|
550 | pass |
---|
551 | |
---|