Alexander/P-unit_model
2
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

add ability to fit contrast flipped cells, change error for frequency lists

Browse Source
This commit is contained in:
a.ott 2020-07-05 11:03:28 +02:00
parent 405424b6eb
commit 89ebeb7fd0
1 changed files with 19 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
Fitter.py
View File

@ -60,10 +60,19 @@ class Fitter:
self.burstiness = data_baseline.get_burstiness()
fi_curve = get_fi_curve_class(cell_data, cell_data.get_fi_contrasts(), save_dir=cell_data.get_data_path())
self.f_inf_slope = fi_curve.get_f_inf_slope()
if self.f_inf_slope < 0:
contrasts = np.array(cell_data.get_fi_contrasts()) * -1
# print("old contrasts:", cell_data.get_fi_contrasts())
# print("new contrasts:", contrasts)
contrasts = sorted(contrasts)
fi_curve = get_fi_curve_class(cell_data, contrasts)
self.fi_contrasts = fi_curve.stimulus_values
self.f_inf_values = fi_curve.f_inf_frequencies
self.f_inf_slope = fi_curve.get_f_inf_slope()
self.f_zero_values = fi_curve.f_zero_frequencies
self.f_zero_fit = fi_curve.f_zero_fit
# self.f_zero_slopes = [fi_curve.get_f_zero_fit_slope_at_stimulus_value(c) for c in self.fi_contrasts]
@ -325,12 +334,12 @@ class Fitter:
error_f_inf_slope = abs((f_infinities_slope - self.f_inf_slope) / (self.f_inf_slope/20))
error_f_inf_slope = abs((f_infinities_slope - self.f_inf_slope) / abs(self.f_inf_slope+1/20))
error_f_inf = calculate_list_error(f_infinities, self.f_inf_values)
# error_f_zero_slopes = calculate_list_error(f_zero_slopes, self.f_zero_slopes)
error_f_zero_slope_at_straight = abs(self.f_zero_slope_at_straight - f_zero_slope_at_straight) \
/ (self.f_zero_slope_at_straight / 10)
/ abs(self.f_zero_slope_at_straight+1 / 10)
error_f_zero = calculate_list_error(f_zeros, self.f_zero_values)
error_list = [error_bf, error_vs, error_sc, error_cv, error_bursty,
@ -341,6 +350,8 @@ class Fitter:
error_list[i] = error_list[i] * error_weights[i]
elif error_weights is not None:
warn("Error: weights had different length than errors and were ignored!")
if sum(error_list) < 0:
print("Error negative: ", error_list)
if np.isnan(sum(error_list)):
print("--------SOME ERROR VALUE(S) IS/ARE NaN:")
print(error_list)
@ -352,13 +363,17 @@ class Fitter:
def calculate_list_error(fit, reference):
error = 0
for i in range(len(reference)):
error += abs_freq_error(fit[i] - reference[i])
# error += abs_freq_error(fit[i] - reference[i])
error += normed_quadratic_freq_error(fit[i], reference[i])
norm_error = error / len(reference)
return norm_error
def normed_quadratic_freq_error(fit, ref, factor=2):
return (abs(fit-ref)/factor)**2 / ref
def abs_freq_error(diff, factor=10):
return abs(diff) / factor