Changing equidistant frequency axis to equidistant wavelength axis - python 3.8 - axis

Question

is there an easy way to change an equidistant frequency axis to an equidistant wavelength axis
when spectrum is given in the form spec = {ndarray: (101, fiveormoredigits)}?
Range of Spectrum:
#spectral extend in THz
v_min = (fd.frequency_0 - fd.freq_intervall / 2.0) * 1e-12
v_max = (fd.frequency_0 + fd.freq_intervall / 2.0) * 1e-12
Draw the Spectrum
fig2 = plt.figure()
bx2 = fig2.add_subplot(122)
bx2.imshow(spec,aspect='auto',interpolation='bicubic',cmap='jet',vmin=spec_max-db_range,vmax=spec_max,origin='lower',extent=[l_min,l_max,0,fd.distance])
bx2.set_xlabel('Frequenz (THz)', fontsize=20)
bx2.set_ylabel('Propagation (m)', fontsize=20)
fig2.suptitle('PM 980 XP/HP: 5cm,200nJ', fontsize=25)
matplotlib.pyplot.xticks(fontsize=15)
matplotlib.pyplot.yticks(fontsize=15)
plt.show()`
Result with code above
#spectral extend in THz
l_min = 3e8/(fd.frequency_0 - fd.freq_intervall / 2.0) * 1e9
l_max = 3e8/(fd.frequency_0 + fd.freq_intervall / 2.0) * 1e9
Result with changed extend
Problem is, that axis is equidistant but center wavelength is not correct.
With this solution for every plot axis needs to be shifted.
fig2 = plt.figure()
bx2 = fig2.add_subplot(122)
bx2.imshow(spec,aspect='auto',interpolation='bicubic',cmap='jet',vmin=spec_max-db_range,vmax=spec_max,origin='lower',extent=[l_min,l_max,0,fd.distance])
bx2.set_xlabel('Frequenz (THz)', fontsize=20)
bx2.set_ylabel('Propagation (m)', fontsize=20)
fig2.suptitle('PM 980 XP/HP: 5cm,200nJ', fontsize=25)
x_label_list = ['800','900', '1035', '1200', '1300', '1400']
bx2.set_xticks([375, 333, 289, 250,230, 214])
bx2.set_xticklabels(x_label_list)
matplotlib.pyplot.xticks(fontsize=15)
matplotlib.pyplot.yticks(fontsize=15)
plt.show()`
Result with ticks replaced
Problem is, that on one hand central wavelength is correct but axis is not equidistant.
Additional. I tryed to interpolate:
Variable: spec = {ndarray:(101, 32768)}
X = np.linspace(0, 100, 101)
Y = np.linspace(3e8/v_max, 3e8/v_min, np.size(spec, 1))
x,y = np.meshgrid(X,Y)
spec_new = interpolate.interp2d(x,y, spec, kind='linear')
So basically spec consist out of 101 rows which represent the frame and 32768 columns. Each column contains an energy value. I tryed to interpolate energy on new equidistant axis but error occurs:
raise OverflowError(msg)
OverflowError: Too many data points to interpolate
I tryed to re-scale energy and interpolate on new equidistant axis but error occurs:
spec_new = interpolate.interp2d(x,y, spec/(X*X), kind='linear')
ValueError: operands could not be broadcast together with shapes (101,8192) (8192,101)

Answer 1

I try to interpolate stepwise but this does not work:
x = np.linspace(3e8/v_max, 3e8/v_min, np.size(spec, 1))
scale = x*x
s = 0
n = 100
for i in range(1, n):
spec[i,:] = interpolate.interp1d(x, spec[i,:]*scale, 'linear')
print(spec);
Error occurs. Why?
spec[i,:] = interpolate.interp1d(x, spec[i,:], 'linear')
TypeError: float() argument must be a string or a number, not 'interp1d'
And why is this not possible?:
for i in range(1, n):
spec_neu[i,:] = interpolate.interp1d(x, spec[i,:]*scale, 'linear')
print(spec_neu);
spec_neu[i,:] = interpolate.interp1d(x, spec[i,:]*scale, 'linear')
NameError: name 'spec_neu' is not defined