def do_CCl4_calibration(filename, showplots = True, outputfile = 'CCl4_cal.pdf'):
#this is tricky. we don't need to set the backed to PDF here before importing, because fitlib does that
import matplotlib
import matplotlib.pyplot as plt
import fitlib as fl
#instanciate a log object
log = loglib.Log(tovariable = True)
log.write(filename)
# we assume the file is OK
badfile = False
# read the file
try:
data = hl.readfile(filename, tolerate_spaces = True)
except IOError:
badfile = True
log.write('Could not read the file: ' + filename)
parameters = []
if badfile is not True:
# guess peaks
peaks = fl.guess_ES_peaks(data, 1, offset = None, limit = 1)
log.write('Result of peak guessing: ' + str(peaks))
# introduce a second peak, usually 0.8 eV higher.
secondpeak = []
secondpeak.append(peaks[0][0] + 0.8)
secondpeak.append(peaks[0][1] / 35)
peaks.append(secondpeak)
log.write('With additional peak: ' + str(peaks))
# fit it
peaksfound, p = fl.fitES(data, peaks, linear_addition = False)
log.write('Result of peak fitting: ' + str(peaksfound))
# plot to get a nice file
fl.plotES(data, 'CCl4')
fl.plot_fit(data, fl.gaussfunctions(2, False), p)
log.emptyline()
if showplots is True:
plt.show()
plt.savefig(outputfile)
return p[1], log.logcontent
def do_SF6_calibration(filelist, showplots = True, quadratic = True, outputfile = 'SF6_cal.pdf'):
#this is tricky. we don't need to set the backed to PDF here before importing, because fitlib does that
import matplotlib
import matplotlib.pyplot as plt
import fitlib as fl
#instanciate a log object
log = loglib.Log(tovariable = True)
#define empty list of peaks used for calibration
calpeaks = []
#pattern to match something like nES_SF6_bla.ee
fn_pattern = re.compile('/nES_([A-Z]{1}[a-z]?)+[1-9]*_.*\.ee$')
#set information for typical calibration peaks
#dictionary with index fragmentname
#data-structure: [[listofpeaksvalues], [actualpeakvalues], [fit parameters], [peak search offset], [peak search end]]
#actualpeakvalues are set to 0 in the beginning
frag_info = {'SF6': [[0.0], [0], [], None, [7.0]], 'SF5': [[0.1], [0], [], None, [7.0]], 'F': [[5.5, 9.0, 11.5], [0, 0, 0], [], [5.0], [16.0]], 'F2': [[4.625, 11.49], [0,0], [], None, [15.0]]}
#for plot numbering
i = 1
fitdata = []
fig1 = plt.figure()
for file in filelist:
if fn_pattern.search(file):
badfile = False
log.write(file)
#read the file
try:
data = hl.readfile(file)
except IOError:
badfile = True
log.write('Could not read the file: ' + file)
#get rid of path (basename), then split by underscores.
#we only want the second part of the filename (e.g. SF6) -> [1]
fragment = os.path.basename(file).split('_')[1]
# is that a proper fragment?
if fragment not in frag_info:
badfile = True
log.write('Unknown fragment: ' + fragment)
#fit the file if we can read it and if we know the fragment
if badfile is False:
log.write('Now dealing with fragment: ' + fragment)
#guess peaks
peaks = fl.guess_ES_peaks(data, len(frag_info[fragment][0]), offset = frag_info[fragment][3], limit = frag_info[fragment][4])
log.write('Result of peak guessing: ' + str(peaks))
#use guessed values to fit; return peak list to frag_info and add fit function parameters to frag_info
if fragment is 'F':
frag_info[fragment][1], frag_info[fragment][2] = fl.fitES(data, peaks, True)
else:
frag_info[fragment][1], frag_info[fragment][2] = fl.fitES(data, peaks)
n = 0
for peakpos in frag_info[fragment][1]:
fitdata.append([peakpos, frag_info[fragment][0][n]])
n += 1
#only plot if we actually had the file (fit parameter is still of type 'list' if this is the case)
if type(frag_info[fragment][2]) is not list:
plt.subplot(3,2,i)
fl.plotES(data, fragment)
# the /3 stems from the fact that the function takes the number of peaks, not parameters
if fragment is 'F':
fl.plot_fit(data, fl.gaussfunctions(3, True), frag_info[fragment][2])
else:
fl.plot_fit(data, fl.gaussfunctions(len(frag_info[fragment][2])/3), frag_info[fragment][2])
i = i + 1
#now we fitted all the calibration scans and have an array with peaks in frag_info
log.emptyline()
log.write('Raw data of the fits: ' + str(frag_info))
log.emptyline()
#convert to numpy array
fitdata = array(fitdata, dtype = float)
#quadratic or linear?
if quadratic is True:
fitfunc = lambda p, x: p[0] + p[1]*x + p[2]*x**2
p = [1]*3
else:
fitfunc = lambda p, x: p[0] + p[1]*x
p = [0]*2
p[0] = 2
p[1] = 1
#we fit the peak positions as they are with where they should be
print fitdata
parameters = fl.fit_function_to_data(fitdata, fitfunc, p)
log.emptyline()
log.write('Fitted and received the following parameters: ' + str(parameters))
if quadratic is True:
log.write('Fit function is therefore y = %f + %f*x + %f*x^2' % (parameters[0], parameters[1], parameters[2]))
else:
log.write('Fit function is therefore y = %f + %f*x' % (parameters[0], parameters[1]))
plt.subplot(3, 2, 5)
plt.plot(fitdata[:,0], fitdata[:,1], 'bo')
plt.xlabel('Measured')
plt.ylabel('Corrected')
plt.grid(True)
plt.title('Calibration function')
fl.plot_fit(fitdata, fitfunc, parameters)
# plt.tight_layout()
if showplots is True:
plt.show()
plt.savefig(outputfile)
return parameters, log.logcontent