Fix autocorrelation_fit

examples/semiclassical_micelles.py

@@ -5,7 +5,6 @@
 
 from radicalpy.data import Molecule
 from radicalpy.estimations import (
-    autocorrelation,
     autocorrelation_fit,
     exchange_interaction_in_solution_MC,
     k_STD,
@@ -20,7 +19,7 @@
 )
 from radicalpy.relaxation import SingletTripletDephasing
 from radicalpy.simulation import Basis, SemiclassicalSimulation, State
-from radicalpy.utils import Bhalf_fit, is_fast_run, read_trajectory_files
+from radicalpy.utils import autocorrelation, Bhalf_fit, is_fast_run, read_trajectory_files
 
 
 def main(
@@ -32,24 +31,24 @@ def main(
     trp = Molecule.all_nuclei("tryptophan_cation")
     sim = SemiclassicalSimulation([flavin, trp], basis=Basis.ZEEMAN)
 
-    # trajectory_data = read_trajectory_files("./examples/data/md_fad_trp_aot")
+    # trajectory_data = read_trajectory_files("./examples/data/md_fad_trp_aot", scale=1e-10)
     # trajectory_ts = (
     #     np.linspace(0, len(trajectory_data), len(trajectory_data)) * 5e-12 * 1e9
     # )
     # j = exchange_interaction_in_solution_MC(trajectory_data[:, 1], J0=5)
-
+    #
     # plot_exchange_interaction_in_solution(trajectory_ts, trajectory_data, j)
-
+    #
     # acf_j = autocorrelation(j, factor=1)
     # zero_point_crossing_j = np.where(np.diff(np.sign(acf_j)))[0][0]
     # t_j_max = max(trajectory_ts[:zero_point_crossing_j]) * 1e-9
     # t_j = np.linspace(5e-12, t_j_max, zero_point_crossing_j)
-
+    #
     # acf_j_fit = autocorrelation_fit(t_j, j, 5e-12, t_j_max)
     # plot_autocorrelation_fit(t_j, acf_j, acf_j_fit, zero_point_crossing_j)
-
-    # kstd = k_STD(j, acf_j_fit["tau_c"])
-    kstd = 462.10479319787896
+    #
+    # kstd = k_STD(-j, acf_j_fit["tau_c"])
+    kstd = 11681368.059456564
     triplet_excited_state_quenching_rate = 5e6
     recombination_rate = 8e6
     free_radical_escape_rate = 5e5

radicalpy/estimations.py

@@ -200,7 +200,7 @@ def multiexponential(x, *params):
         p0=np.zeros(num_exp),
     )
     fit = multiexponential(ts, *acf_popt)
-    tau_c = sum(acf_popt * taus) * np.var(mT_to_MHz(trajectory)) / 1e6
+    tau_c = sum(acf_popt * taus) # * np.var(mT_to_MHz(trajectory)) / 1e6
     return {"fit": fit, "tau_c": tau_c}