3 """Force spectroscopy curves filtering of flat curves
5 Other plugin dependencies:
6 procplots.py (plot processing plugin)
9 from hooke.libhooke import WX_GOOD
12 wxversion.select(WX_GOOD)
13 import xml.dom.minidom
17 from numpy import diff
20 from .. import libpeakspot as lps
21 from .. import curve as lhc
24 class flatfiltsCommands(object):
27 #configurate convfilt variables
28 convfilt_configurator=ConvfiltConfig()
29 self.convfilt_config=convfilt_configurator.load_config('convfilt.conf')
31 def do_flatfilt(self,args):
35 Filters out flat (featureless) curves of the current playlist,
36 creating a playlist containing only the curves with potential
40 flatfilt [min_npks min_deviation]
42 min_npks = minmum number of points over the deviation
45 min_deviation = minimum signal/noise ratio
48 If called without arguments, it uses default values, that
49 should work most of the times.
58 min_deviation=int(args[1])
62 print 'Processing playlist...'
66 for item in self.current_list:
70 notflat=self.has_features(item, median_filter, min_npks, min_deviation)
71 print 'Curve',item.path, 'is',c,'of',len(self.current_list),': features are ',notflat
74 print 'Curve',item.path, 'is',c,'of',len(self.current_list),': cannot be filtered. Probably unable to retrieve force data from corrupt file.'
78 item.curve=None #empty the item object, to further avoid memory leak
79 notflat_list.append(item)
81 if len(notflat_list)==0:
82 print 'Found nothing interesting. Check your playlist, could be a bug or criteria could be too much stringent'
85 print 'Found ',len(notflat_list),' potentially interesting curves'
86 print 'Regenerating playlist...'
88 self.current_list=notflat_list
89 self.current=self.current_list[self.pointer]
92 def has_features(self,item,median_filter,min_npks,min_deviation):
94 decides if a curve is flat enough to be rejected from analysis: it sees if there
95 are at least min_npks points that are higher than min_deviation times the absolute value
98 Algorithm original idea by Francesco Musiani, with my tweaks and corrections.
102 item.identify(self.drivers)
103 #we assume the first is the plot with the force curve
104 #do the median to better resolve features from noise
105 flat_plot=self.plotmanip_median(item.curve.default_plots()[0], item, customvalue=median_filter)
106 flat_vects=flat_plot.vectors
107 item.curve.close_all()
108 #needed to avoid *big* memory leaks!
112 #absolute value of derivate
113 yretdiff=diff(flat_vects[1][1])
114 yretdiff=[abs(value) for value in yretdiff]
115 #average of derivate values
116 diffmean=numpy.mean(yretdiff)
120 for value in yretdiff:
121 if value/diffmean > min_deviation:
129 del flat_plot, flat_vects, yretdiff
133 ################################################################
134 #-----CONVFILT-------------------------------------------------
135 #-----Convolution-based peak recognition and filtering.
136 #Requires the libpeakspot.py library
138 def has_peaks(self, plot, abs_devs=None, maxpeak=True, window=10):
140 Finds peak position in a force curve.
141 FIXME: should be moved in libpeakspot.py
144 abs_devs=self.convfilt_config['mindeviation']
147 xret=plot.vectors[1][0]
148 yret=plot.vectors[1][1]
149 #Calculate convolution.
150 convoluted=lps.conv_dx(yret, self.convfilt_config['convolution'])
152 #surely cut everything before the contact point
153 cut_index=self.find_contact_point(plot)
154 #cut even more, before the blind window
155 start_x=xret[cut_index]
157 for value in xret[cut_index:]:
158 if abs((value) - (start_x)) > self.convfilt_config['blindwindow']*(10**-9):
161 cut_index+=blind_index
162 #do the dirty convolution-peak finding stuff
163 noise_level=lps.noise_absdev(convoluted[cut_index:], self.convfilt_config['positive'], self.convfilt_config['maxcut'], self.convfilt_config['stable'])
164 above=lps.abovenoise(convoluted,noise_level,cut_index,abs_devs)
165 peak_location,peak_size=lps.find_peaks(above,seedouble=self.convfilt_config['seedouble'])
167 #take the minimum or the maximum of a peak
168 for i in range(len(peak_location)):
169 peak=peak_location[i]
170 valpk=min(yret[peak-window:peak+window]) #maximum in force (near the unfolding point)
171 index_pk=yret[peak-window:peak+window].index(valpk)+(peak-window)
174 valpk=max(yret[peak:peak+window]) #minimum in force, near the baseline
175 index_pk=yret[peak:peak+window].index(valpk)+(peak)
177 # Let's explain that for the minimum. Immaging that we know that there is a peak at position/region 100 and you have found its y-value,
178 # Now you look in the array, from 100-10 to 100+10 (if the window is 10).
179 # This "100-10 to 100+10" is substancially a new array with its index. In this array you have 20
180 # elements, so the index of your y-value will be 10.
181 # Now to find the index in the TOTAL array you have to add the "position" of the "region" (that in this case
182 # correspond to 100) and also substract the window size ---> (+100-10)
184 peak_location[i]=index_pk
186 return peak_location,peak_size
189 def exec_has_peaks(self,item,abs_devs):
191 encapsulates has_peaks for the purpose of correctly treating the curve objects in the convfilt loop,
192 to avoid memory leaks
194 item.identify(self.drivers)
195 #we assume the first is the plot with the force curve
196 plot=item.curve.default_plots()[0]
198 if 'flatten' in self.config['plotmanips']:
199 #If flatten is present, use it for better recognition of peaks...
200 flatten=self._find_plotmanip('flatten') #extract flatten plot manipulator
201 plot=flatten(plot, item, customvalue=1)
203 peak_location,peak_size=self.has_peaks(plot,abs_devs)
204 #close all open files
205 item.curve.close_all()
206 #needed to avoid *big* memory leaks!
209 return peak_location, peak_size
211 #------------------------
212 #------commands----------
213 #------------------------
214 def do_peaks(self,args):
218 Test command for convolution filter / test.
220 Syntax: peaks [deviations]
221 absolute deviation = number of times the convolution signal is above the noise absolute deviation.
225 args=self.convfilt_config['mindeviation']
230 print 'Wrong argument, using config value'
231 abs_devs=float(self.convfilt_config['mindeviation'])
233 defplots=self.current.curve.default_plots()[0] #we need the raw, uncorrected plots
235 if 'flatten' in self.config['plotmanips']:
236 flatten=self._find_plotmanip('flatten') #extract flatten plot manipulator
237 defplots=flatten(defplots, self.current)
239 print 'You have the flatten plot manipulator not loaded. Enabling it could give you better results.'
241 peak_location,peak_size=self.has_peaks(defplots,abs_devs)
242 print 'Found '+str(len(peak_location))+' peaks.'
243 to_dump='peaks '+self.current.path+' '+str(len(peak_location))
244 self.outlet.push(to_dump)
247 #if no peaks, we have nothing to plot. exit.
248 if len(peak_location)==0:
251 #otherwise, we plot the peak locations.
252 xplotted_ret=self.plots[0].vectors[1][0]
253 yplotted_ret=self.plots[0].vectors[1][1]
254 xgood=[xplotted_ret[index] for index in peak_location]
255 ygood=[yplotted_ret[index] for index in peak_location]
257 recplot=self._get_displayed_plot()
258 recplot.vectors.append([xgood,ygood])
259 if recplot.styles==[]:
260 recplot.styles=[None,None,'scatter']
261 recplot.colors=[None,None,None]
263 recplot.styles+=['scatter']
264 recplot.colors+=[None]
266 self._send_plot([recplot])
268 def do_convfilt(self,args):
272 Filters out flat (featureless) curves of the current playlist,
273 creating a playlist containing only the curves with potential
277 convfilt [min_npks min_deviation]
279 min_npks = minmum number of peaks
280 (to set the default, see convfilt.conf file; CONVCONF and SETCONF commands)
282 min_deviation = minimum signal/noise ratio *in the convolution*
283 (to set the default, see convfilt.conf file; CONVCONF and SETCONF commands)
285 If called without arguments, it uses default values.
288 min_npks=self.convfilt_config['minpeaks']
289 min_deviation=self.convfilt_config['mindeviation']
293 min_npks=int(args[0])
294 min_deviation=int(args[1])
298 print 'Processing playlist...'
299 print '(Please wait)'
303 for item in self.current_list:
307 peak_location,peak_size=self.exec_has_peaks(item,min_deviation)
308 if len(peak_location)>=min_npks:
312 print 'Curve',item.path, 'is',c,'of',len(self.current_list),': found '+str(len(peak_location))+' peaks.'+isok
314 peak_location,peak_size=[],[]
315 print 'Curve',item.path, 'is',c,'of',len(self.current_list),': cannot be filtered. Probably unable to retrieve force data from corrupt file.'
317 if len(peak_location)>=min_npks:
318 item.peak_location=peak_location
319 item.peak_size=peak_size
320 item.curve=None #empty the item object, to further avoid memory leak
321 notflat_list.append(item)
323 #Warn that no flattening had been done.
324 if not ('flatten' in self.config['plotmanips']):
325 print 'Flatten manipulator was not found. Processing was done without flattening.'
326 print 'Try to enable it in your configuration file for better results.'
328 if len(notflat_list)==0:
329 print 'Found nothing interesting. Check your playlist, could be a bug or criteria could be too much stringent'
332 print 'Found ',len(notflat_list),' potentially interesting curves'
333 print 'Regenerating playlist...'
335 self.current_list=notflat_list
336 self.current=self.current_list[self.pointer]
340 def do_setconv(self,args):
344 Sets the convfilt configuration variables
346 Syntax: setconv variable value
349 #FIXME: a general "set dictionary" function has to be built
351 print self.convfilt_config
353 if not (args[0] in self.convfilt_config.keys()):
354 print 'This is not an internal convfilt variable!'
355 print 'Run "setconv" without arguments to see a list of defined variables.'
359 print self.convfilt_config[args[0]]
362 self.convfilt_config[args[0]]=eval(args[1])
363 except NameError: #we have a string argument
364 self.convfilt_config[args[0]]=args[1]
367 #########################
368 #HANDLING OF CONFIGURATION FILE
369 class ConvfiltConfig(object):
371 Handling of convfilt configuration file
373 Mostly based on the simple-yet-useful examples of the Python Library Reference
374 about xml.dom.minidom
376 FIXME: starting to look a mess, should require refactoring
383 def load_config(self, filename):
384 myconfig=file(filename)
385 #the following 3 lines are needed to strip newlines. otherwise, since newlines
386 #are XML elements too, the parser would read them (and re-save them, multiplying
388 #yes, I'm an XML n00b
389 the_file=myconfig.read()
390 the_file_lines=the_file.split('\n')
391 the_file=''.join(the_file_lines)
393 self.config_tree=xml.dom.minidom.parseString(the_file)
395 def getText(nodelist):
396 #take the text from a nodelist
397 #from Python Library Reference 13.7.2
399 for node in nodelist:
400 if node.nodeType == node.TEXT_NODE:
404 def handleConfig(config):
405 noiseabsdev_elements=config.getElementsByTagName("noise_absdev")
406 convfilt_elements=config.getElementsByTagName("convfilt")
407 handleAbsdev(noiseabsdev_elements)
408 handleConvfilt(convfilt_elements)
410 def handleAbsdev(noiseabsdev_elements):
411 for element in noiseabsdev_elements:
412 for attribute in element.attributes.keys():
413 self.config[attribute]=element.getAttribute(attribute)
415 def handleConvfilt(convfilt_elements):
416 for element in convfilt_elements:
417 for attribute in element.attributes.keys():
418 self.config[attribute]=element.getAttribute(attribute)
420 handleConfig(self.config_tree)
421 #making items in the dictionary machine-readable
422 for item in self.config.keys():
424 self.config[item]=eval(self.config[item])
425 except NameError: #if it's an unreadable string, keep it as a string