Skip to content

Commit 57845ef

Browse files
SarthakJariwalaSarthakJariwala
committed
fix residual func in fit_wo_irf, change export for fit_wo_irf to include raw PL with rise
1 parent 70185c0 commit 57845ef

File tree

2 files changed

+40
-31
lines changed

2 files changed

+40
-31
lines changed

‎PythonGUI_apps/Lifetime_analysis/Fit_functions.py‎

Lines changed: 31 additions & 25 deletions
Original file line numberDiff line numberDiff line change
@@ -9,10 +9,10 @@
99
from scipy.optimize import differential_evolution
1010
from scipy.special import gamma
1111

12-
def stretch_exp_fit(TRPL, t, Tc = (0,1e5), Beta = (0,1), A = (0,1e6)):
12+
def stretch_exp_fit(TRPL, t, Tc = (0,1e5), Beta = (0,1), A = (0,1e6), noise=(0,1e6)):
1313

14-
def exp_stretch(t, tc, beta, a):
15-
return (a * np.exp(-((1.0 / tc) * t) ** beta))
14+
def exp_stretch(t, tc, beta, a, noise):
15+
return ((a * np.exp(-((1.0 / tc) * t) ** beta)) + noise)
1616

1717
def avg_tau_from_exp_stretch(tc, beta):
1818
return (tc / beta) * gamma(1.0 / beta)
@@ -25,13 +25,14 @@ def residuals(params):#params are the parameters to be adjusted by differential
2525
tc = params[0]
2626
beta = params[1]
2727
a = params[2]
28+
noise = params[3]
2829

29-
PL_sim = exp_stretch(t,tc,beta,a)
30+
PL_sim = exp_stretch(t,tc,beta,a, noise)
3031

31-
Resid= np.sqrt(np.sum(((PL_sim-TRPL)**2)/(np.sqrt(PL_sim)**2)))
32+
Resid= np.sum(((PL_sim-TRPL)**2)/(np.sqrt(TRPL)**2))
3233
return Resid #returns the difference between the PL data and simulated data
3334

34-
bounds = [Tc, Beta, A]
35+
bounds = [Tc, Beta, A, noise]
3536

3637
result = differential_evolution(residuals, bounds)
3738
return result.x
@@ -41,20 +42,21 @@ def residuals(params):#params are the parameters to be adjusted by differential
4142
tc = p[0]
4243
beta = p[1]
4344
a = p[2]
45+
noise = p[3]
4446

45-
PL_fit = exp_stretch(t,tc,beta,a)
47+
PL_fit = exp_stretch(t,tc,beta,a, noise)
4648

4749
avg_tau = avg_tau_from_exp_stretch(tc,beta)
4850

49-
return tc, beta, a, avg_tau, PL_fit
51+
return tc, beta, a, avg_tau, PL_fit, noise
5052

51-
def double_exp_fit(TRPL, t, tau1_bounds=(0,100), a1_bounds=(0,1e5), tau2_bounds=(0,100), a2_bounds=(0,1e5)):
53+
def double_exp_fit(TRPL, t, tau1_bounds=(0,1000), a1_bounds=(0,1e6), tau2_bounds=(0,10000), a2_bounds=(0,1e5), noise=(0,1e6)):
5254

5355
def single_exp(t, tau, a):
54-
return (a * np.exp(-((1.0 / tau)*t) ))
56+
return (a * np.exp(-((1.0 / tau)*t)))
5557

56-
def double_exp(t, tau1, a1, tau2, a2):
57-
return ((single_exp(t, tau1, a1)) + (single_exp(t, tau2, a2)))
58+
def double_exp(t, tau1, a1, tau2, a2, noise):
59+
return ((single_exp(t, tau1, a1)) + (single_exp(t, tau2, a2)) + noise)
5860

5961
def avg_tau_from_double_exp(tau1, a1, tau2, a2):
6062
return (((tau1*a1) + (tau2*a2))/(a1+a2))
@@ -68,13 +70,14 @@ def residuals(params):#params are the parameters to be adjusted by differential
6870
a1 = params[1]
6971
tau2 = params[2]
7072
a2 = params[3]
73+
noise = params[4]
7174

72-
PL_sim = double_exp(t,tau1, a1, tau2, a2)
75+
PL_sim = double_exp(t,tau1, a1, tau2, a2, noise)
7376

74-
Resid= np.sqrt(np.sum(((PL_sim-TRPL)**2)/(np.sqrt(PL_sim)**2)))
77+
Resid= np.sum(((PL_sim-TRPL)**2)/(np.sqrt(TRPL)**2))
7578
return Resid #returns the difference between the PL data and simulated data
7679

77-
bounds = [tau1_bounds, a1_bounds, tau2_bounds, a2_bounds]
80+
bounds = [tau1_bounds, a1_bounds, tau2_bounds, a2_bounds, noise]
7881

7982
result = differential_evolution(residuals, bounds)
8083
return result.x
@@ -85,17 +88,18 @@ def residuals(params):#params are the parameters to be adjusted by differential
8588
a1 = p[1]
8689
tau2 = p[2]
8790
a2 = p[3]
91+
noise = p[4]
8892

89-
PL_fit = double_exp(t, tau1, a1, tau2, a2)
93+
PL_fit = double_exp(t, tau1, a1, tau2, a2, noise)
9094

9195
avg_tau = avg_tau_from_double_exp(tau1, a1, tau2, a2)
9296

93-
return tau1, a1, tau2, a2, avg_tau, PL_fit
97+
return tau1, a1, tau2, a2, avg_tau, PL_fit, noise
9498

95-
def single_exp_fit(TRPL, t, tau_bounds=(0,1000), a_bounds=(0,1e5)):
99+
def single_exp_fit(TRPL, t, tau_bounds=(0,10000), a_bounds=(0,1e6), noise=(0,1e6)):
96100

97-
def single_exp(t, tau, a):
98-
return (a * np.exp(-((1.0 / tau)*t) ))
101+
def single_exp(t, tau, a, noise):
102+
return (a * np.exp(-((1.0 / tau)*t) ) + noise)
99103

100104
def Diff_Ev_Fit_singleExp(TRPL):
101105
TRPL = TRPL
@@ -104,13 +108,14 @@ def residuals(params):#params are the parameters to be adjusted by differential
104108
#Variable Rates
105109
tau = params[0]
106110
a = params[1]
111+
noise = params[2]
107112

108-
PL_sim = single_exp(t, tau, a)
113+
PL_sim = single_exp(t, tau, a, noise)
109114

110-
Resid= np.sqrt(np.sum(((PL_sim-TRPL)**2)/(np.sqrt(PL_sim)**2)))
115+
Resid= np.sum(((PL_sim-TRPL)**2)/(np.sqrt(TRPL)**2))
111116
return Resid #returns the difference between the PL data and simulated data
112117

113-
bounds = [tau_bounds, a_bounds]
118+
bounds = [tau_bounds, a_bounds, noise]
114119

115120
result = differential_evolution(residuals, bounds)
116121
return result.x
@@ -119,8 +124,9 @@ def residuals(params):#params are the parameters to be adjusted by differential
119124

120125
tau = p[0]
121126
a = p[1]
127+
noise = p[2]
122128

123-
PL_fit = single_exp(t, tau, a)
129+
PL_fit = single_exp(t, tau, a, noise)
124130

125-
return tau, a, PL_fit
131+
return tau, a, PL_fit, noise
126132

‎PythonGUI_apps/Lifetime_analysis/Lifetime_plot_fit.py‎

Lines changed: 9 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -306,7 +306,7 @@ def fit_and_plot(self):
306306
self.ui.plot.plot(t, y, clear=self.ui.clear_plot_checkBox.isChecked(), pen=pg.mkPen(self.plot_color))
307307

308308
if fit_func == "Stretched Exponential": #stretch exponential tab
309-
tc, beta, a, avg_tau, PL_fit = stretch_exp_fit(TRPL_interp, t)
309+
tc, beta, a, avg_tau, PL_fit, noise = stretch_exp_fit(TRPL_interp, t)
310310
self.out = np.empty((len(t), 3))
311311
self.out[:,0] = t #time
312312
self.out[:,1] = TRPL_interp #Raw PL
@@ -316,11 +316,12 @@ def fit_and_plot(self):
316316
"\nFit Method: " + "diff_ev" + #TODO : change when diff_ev and fmin_tnc implemented for non-irf
317317
"\nAverage Lifetime = " + str(avg_tau)+ " ns"
318318
"\nCharacteristic Tau = " + str(tc)+" ns"
319-
"\nBeta = "+str(beta))
319+
"\nBeta = "+str(beta)+
320+
"\nNoise = "+ str(noise))
320321
self.ui.average_lifetime_spinBox.setValue(avg_tau)
321322

322323
elif fit_func == "Double Exponential": #double exponential tab
323-
tau1, a1, tau2, a2, avg_tau, PL_fit = double_exp_fit(TRPL_interp, t)
324+
tau1, a1, tau2, a2, avg_tau, PL_fit, noise = double_exp_fit(TRPL_interp, t)
324325
self.out = np.empty((len(t), 3))
325326
self.out[:,0] = t #time
326327
self.out[:,1] = TRPL_interp #Raw PL
@@ -332,11 +333,12 @@ def fit_and_plot(self):
332333
"\nTau 1 = " + str(tau1)+" ns"
333334
"\nA 1 = " + str(a1)+
334335
"\nTau 2 = " + str(tau2)+" ns"
335-
"\nA 2 = " + str(a2))
336+
"\nA 2 = " + str(a2)+
337+
"\nNoise = "+ str(noise))
336338
#TODO - once tau_avg implemented, set average lifetime spinbox to tau_avg value
337339

338340
elif fit_func == "Single Exponential": #single exponential tab
339-
tau, a, PL_fit = single_exp_fit(TRPL_interp, t)
341+
tau, a, PL_fit, noise = single_exp_fit(TRPL_interp, t)
340342
self.out = np.empty((len(t), 3))
341343
self.out[:,0] = t #time
342344
self.out[:,1] = TRPL_interp #Raw PL
@@ -345,7 +347,8 @@ def fit_and_plot(self):
345347
self.ui.Result_textBrowser.setText("Fit Results:\n\nFit Function: Single Exponential"
346348
"\nFit Method: " + "diff_ev" +
347349
"\nLifetime = " + str(tau)+ " ns"
348-
"\nA = " + str(a))
350+
"\nA = " + str(a)+
351+
"\nNoise = "+ str(noise))
349352
self.ui.average_lifetime_spinBox.setValue(tau)
350353

351354
#add fit params to data_list

0 commit comments

Comments
 (0)