1 | import _asap
|
---|
2 | from asap import rcParams
|
---|
3 |
|
---|
4 | class fitter:
|
---|
5 | """
|
---|
6 | The fitting class for ASAP.
|
---|
7 | """
|
---|
8 | def _verbose(self, *args):
|
---|
9 | """
|
---|
10 | Set stdout output.
|
---|
11 | """
|
---|
12 | if type(args[0]) is bool:
|
---|
13 | self._vb = args[0]
|
---|
14 | return
|
---|
15 | elif len(args) == 0:
|
---|
16 | return self._vb
|
---|
17 |
|
---|
18 | def __init__(self):
|
---|
19 | """
|
---|
20 | Create a fitter object. No state is set.
|
---|
21 | """
|
---|
22 | self.fitter = _asap.fitter()
|
---|
23 | self.x = None
|
---|
24 | self.y = None
|
---|
25 | self.mask = None
|
---|
26 | self.fitfunc = None
|
---|
27 | self.fitted = False
|
---|
28 | self.data = None
|
---|
29 | self._p = None
|
---|
30 | self._vb = True
|
---|
31 |
|
---|
32 | def set_data(self, xdat, ydat, mask=None):
|
---|
33 | """
|
---|
34 | Set the absissa and ordinate for the fit. Also set the mask
|
---|
35 | indicationg valid points.
|
---|
36 | This can be used for data vectors retrieved from a scantable.
|
---|
37 | For scantable fitting use 'fitter.set_scan(scan, mask)'.
|
---|
38 | Parameters:
|
---|
39 | xdat: the abcissa values
|
---|
40 | ydat: the ordinate values
|
---|
41 | mask: an optional mask
|
---|
42 |
|
---|
43 | """
|
---|
44 | self.fitted = False
|
---|
45 | self.x = xdat
|
---|
46 | self.y = ydat
|
---|
47 | if mask == None:
|
---|
48 | from numarray import ones
|
---|
49 | self.mask = ones(len(xdat))
|
---|
50 | else:
|
---|
51 | self.mask = mask
|
---|
52 | return
|
---|
53 |
|
---|
54 | def set_scan(self, thescan=None, mask=None):
|
---|
55 | """
|
---|
56 | Set the 'data' (a scantable) of the fitter.
|
---|
57 | Parameters:
|
---|
58 | thescan: a scantable
|
---|
59 | mask: a msk retireved from the scantable
|
---|
60 | """
|
---|
61 | if not thescan:
|
---|
62 | print "Please give a correct scan"
|
---|
63 | self.fitted = False
|
---|
64 | self.data = thescan
|
---|
65 | if mask is None:
|
---|
66 | from numarray import ones
|
---|
67 | self.mask = ones(self.data.nchan())
|
---|
68 | else:
|
---|
69 | self.mask = mask
|
---|
70 | return
|
---|
71 |
|
---|
72 | def set_function(self, **kwargs):
|
---|
73 | """
|
---|
74 | Set the function to be fit.
|
---|
75 | Parameters:
|
---|
76 | poly: use a polynomial of the order given
|
---|
77 | gauss: fit the number of gaussian specified
|
---|
78 | Example:
|
---|
79 | fitter.set_function(gauss=2) # will fit two gaussians
|
---|
80 | fitter.set_function(poly=3) # will fit a 3rd order polynomial
|
---|
81 | """
|
---|
82 | #default poly order 0
|
---|
83 |
|
---|
84 |
|
---|
85 | if kwargs.has_key('poly'):
|
---|
86 | self.fitfunc = 'poly'
|
---|
87 | n = kwargs.get('poly')
|
---|
88 | elif kwargs.has_key('gauss'):
|
---|
89 | n = kwargs.get('gauss')
|
---|
90 | self.fitfunc = 'gauss'
|
---|
91 |
|
---|
92 | self.fitter.setexpression(self.fitfunc,n)
|
---|
93 | return
|
---|
94 |
|
---|
95 | def fit(self):
|
---|
96 | """
|
---|
97 | Execute the actual fitting process. All the state has to be set.
|
---|
98 | Parameters:
|
---|
99 | none
|
---|
100 | Example:
|
---|
101 | s= scantable('myscan.asap')
|
---|
102 | f = fitter()
|
---|
103 | f.set_scan(s)
|
---|
104 | f.set_function(poly=0)
|
---|
105 | f.fit()
|
---|
106 | """
|
---|
107 | if ((self.x is None or self.y is None) and self.data is None) \
|
---|
108 | or self.fitfunc is None:
|
---|
109 | print "Fitter not yet initialised. Please set data & fit function"
|
---|
110 | return
|
---|
111 | else:
|
---|
112 | if self.data is not None:
|
---|
113 | self.x = self.data._getabcissa()
|
---|
114 | self.y = self.data._getspectrum()
|
---|
115 | print "Fitting:"
|
---|
116 | vb = self.data._vb
|
---|
117 | self.data._vb = True
|
---|
118 | s = self.data.get_cursor()
|
---|
119 | self.data._vb = vb
|
---|
120 |
|
---|
121 | self.fitter.setdata(self.x,self.y,self.mask)
|
---|
122 | if self.fitfunc == 'gauss':
|
---|
123 | ps = self.fitter.getparameters()
|
---|
124 | if len(ps) == 0:
|
---|
125 | self.fitter.estimate()
|
---|
126 | self.fitter.fit()
|
---|
127 | self.fitted = True
|
---|
128 | return
|
---|
129 |
|
---|
130 | def set_parameters(self, params, fixed=None):
|
---|
131 | self.fitter.setparameters(params)
|
---|
132 | if fixed is not None:
|
---|
133 | self.fitter.setfixedparameters(fixed)
|
---|
134 | return
|
---|
135 |
|
---|
136 | def get_parameters(self):
|
---|
137 | """
|
---|
138 | Return the fit paramters.
|
---|
139 |
|
---|
140 | """
|
---|
141 | if not self.fitted:
|
---|
142 | print "Not yet fitted."
|
---|
143 | pars = list(self.fitter.getparameters())
|
---|
144 | fixed = list(self.fitter.getfixedparameters())
|
---|
145 | if self._vb:
|
---|
146 | print self._format_pars(pars)
|
---|
147 | return pars,fixed
|
---|
148 |
|
---|
149 | def _format_pars(self, pars):
|
---|
150 | out = ''
|
---|
151 | if self.fitfunc == 'poly':
|
---|
152 | c = 0
|
---|
153 | for i in pars:
|
---|
154 | out += ' p%d = %3.3f, ' % (c,i)
|
---|
155 | c+=1
|
---|
156 | elif self.fitfunc == 'gauss':
|
---|
157 | i = 0
|
---|
158 | c = 0
|
---|
159 | unit = ''
|
---|
160 | if self.data:
|
---|
161 | unit = self.data.get_unit()
|
---|
162 | while i < len(pars):
|
---|
163 | out += ' %d: peak = %3.3f , centre = %3.3f %s, FWHM = %3.3f %s \n' % (c,pars[i],pars[i+1],unit,pars[i+2],unit)
|
---|
164 | c+=1
|
---|
165 | i+=3
|
---|
166 | return out
|
---|
167 |
|
---|
168 | def get_estimate(self):
|
---|
169 | """
|
---|
170 | Return the paramter estimates (for non-linear functions).
|
---|
171 | """
|
---|
172 | pars = self.fitter.getestimate()
|
---|
173 | if self._vb:
|
---|
174 | print self._format_pars(pars)
|
---|
175 | return pars
|
---|
176 |
|
---|
177 |
|
---|
178 | def get_residual(self):
|
---|
179 | """
|
---|
180 | Return the residual of the fit.
|
---|
181 | """
|
---|
182 | if not self.fitted:
|
---|
183 | print "Not yet fitted."
|
---|
184 | return self.fitter.getresidual()
|
---|
185 |
|
---|
186 | def get_chi2(self):
|
---|
187 | """
|
---|
188 | Return chi^2.
|
---|
189 | """
|
---|
190 |
|
---|
191 | if not self.fitted:
|
---|
192 | print "Not yet fitted."
|
---|
193 | ch2 = self.fitter.getchi2()
|
---|
194 | if self._vb:
|
---|
195 | print 'Chi^2 = %3.3f' % (ch2)
|
---|
196 | return ch2
|
---|
197 |
|
---|
198 | def get_fit(self):
|
---|
199 | """
|
---|
200 | Return the fitted ordinate values.
|
---|
201 | """
|
---|
202 | if not self.fitted:
|
---|
203 | print "Not yet fitted."
|
---|
204 | return self.fitter.getfit()
|
---|
205 |
|
---|
206 | def commit(self):
|
---|
207 | """
|
---|
208 | Return a new scan where teh fits have been commited.
|
---|
209 | """
|
---|
210 | if not self.fitted:
|
---|
211 | print "Not yet fitted."
|
---|
212 | if self.data is not scantable:
|
---|
213 | print "Only works with scantables"
|
---|
214 | return
|
---|
215 | scan = self.data.copy()
|
---|
216 | scan._setspectrum(self.fitter.getresidual())
|
---|
217 |
|
---|
218 | def plot(self, residual=False):
|
---|
219 | """
|
---|
220 | Plot the last fit.
|
---|
221 | Parameters:
|
---|
222 | residual: an optional parameter indicating if the residual
|
---|
223 | should be plotted (default 'False')
|
---|
224 | """
|
---|
225 | if not self.fitted:
|
---|
226 | return
|
---|
227 | if not self._p:
|
---|
228 | from asap.asaplot import ASAPlot
|
---|
229 | self._p = ASAPlot()
|
---|
230 | if self._p.is_dead:
|
---|
231 | from asap.asaplot import ASAPlot
|
---|
232 | self._p = ASAPlot()
|
---|
233 | self._p.clear()
|
---|
234 | tlab = 'Spectrum'
|
---|
235 | xlab = 'Abcissa'
|
---|
236 | if self.data:
|
---|
237 | tlab = self.data._getsourcename(0)
|
---|
238 | xlab = self.data._getabcissalabel(0)
|
---|
239 | ylab = r'Flux'
|
---|
240 | m = self.data.getmask(0)
|
---|
241 | self._p.set_line(colour='blue',label='Spectrum')
|
---|
242 | self._p.plot(self.x, self.y, m)
|
---|
243 | if residual:
|
---|
244 | self._p.set_line(colour='green',label='Residual')
|
---|
245 | self._p.plot(self.x, self.get_residual(), m)
|
---|
246 | self._p.set_line(colour='red',label='Fit')
|
---|
247 | self._p.plot(self.x, self.get_fit(), m)
|
---|
248 |
|
---|
249 | self._p.set_axes('xlabel',xlab)
|
---|
250 | self._p.set_axes('ylabel',ylab)
|
---|
251 | self._p.set_axes('title',tlab)
|
---|
252 | self._p.release()
|
---|
253 |
|
---|
254 |
|
---|
255 | def auto_fit(self, insitu=None):
|
---|
256 | """
|
---|
257 | Return a scan where the function is applied to all rows for all Beams/IFs/Pols.
|
---|
258 |
|
---|
259 | """
|
---|
260 | from asap import scantable
|
---|
261 | if not isinstance(self.data,scantable) :
|
---|
262 | print "Only works with scantables"
|
---|
263 | return
|
---|
264 | if insitu is None: insitu = rcParams['insitu']
|
---|
265 | if not insitu:
|
---|
266 | scan = self.data.copy()
|
---|
267 | else:
|
---|
268 | scan = self.data
|
---|
269 | vb = scan._vb
|
---|
270 | scan._vb = False
|
---|
271 | sel = scan.get_cursor()
|
---|
272 | rows = range(scan.nrow())
|
---|
273 | for i in range(scan.nbeam()):
|
---|
274 | scan.setbeam(i)
|
---|
275 | for j in range(scan.nif()):
|
---|
276 | scan.setif(j)
|
---|
277 | for k in range(scan.npol()):
|
---|
278 | scan.setpol(k)
|
---|
279 | if self._vb:
|
---|
280 | print "Fitting:"
|
---|
281 | print 'Beam[%d], IF[%d], Pol[%d]' % (i,j,k)
|
---|
282 | for iRow in rows:
|
---|
283 | self.x = scan._getabcissa(iRow)
|
---|
284 | self.y = scan._getspectrum(iRow)
|
---|
285 | self.data = None
|
---|
286 | self.fit()
|
---|
287 | x = self.get_parameters()
|
---|
288 | scan._setspectrum(self.fitter.getresidual(),iRow)
|
---|
289 | scan.set_cursor(sel[0],sel[1],sel[2])
|
---|
290 | scan._vb = vb
|
---|
291 | if not insitu:
|
---|
292 | return scan
|
---|